Alright so I'm shopping around for a cnc mill.

First I looked at getting a Taig 2027ER(from nick), then the MicroMill DSLS 3000 from microproto(just another taig with closed loop), and the CNC Super X3 from Syil.

Cost isn't really an issue as long as it's around $3000(under is nice). The biggest problem with choosing is waiting for it. I'm just home for under 2 months for the summer. I wanted to get something pretty fast to have some time to play with it before I have to start class up. I do intend to pay the mill visits every weekend as I would only have class 2 days out of the week. :) Ok so going on.

The wait:
Nick says 4 to 8 weeks
Microproto says 4 weeks
Syil says 20 days to build and 25 days to ship

Before anything else. Which is the best machine? The Taig is proven to be a great machine and I haven't heard anything about microprotos closed system. I asked them about it and they say it can exceed feedrates of 60 IPM while the 2027 can only do about 30. Since the Super X3 is somewhat new I'm not sure if I want to just jump into that. Shipping the darn thing scares me also.

I'm tempted to just buy a taig off ebay but they pretty much go for what they cost new and I'm usually not too thrilled about the controllers.

What do I do? I don't know which one to choose. Which one would be worth the wait. etc. Advice very much appriciated.

Maybe I could build one instead. I'm up for anything.

I would get the X3. The rule is, get the biggest machine you can. For 3k, the taig is definitely not it. I have not found any pictures comparing the X3 vs a Taig, but It is much larger. Consider the weight
Something like 85 pounds for a Taig, vs 353lbs is what the Grizzly manual said about my X3.
So it will be a much more capable machine, I am very pleased with my X3.

Oh, it would be possible to get an X3 from Grizzly, and CNC it yourself if you wanted. I think theres a few kits that will mechanically put it all together for you, bolt on. Then you can buy a kit that will have all the electronics, there are plenty available. Or you could make the mechanical part yourself since you have the mill at your disposal.

If you've got the money to spend, the X3 is definitely the way to go...The Taig is its less expensive-less-"big" mill...Although you can buy the Taig with motor mounts pre-done at the factory.

I would go with the Taig cnc ready setup with the er16 spindle.
You would proly have it in 2 weeks plus you get a 2 yr waranty.
All though the taig is small and portable you can still make good size pcs with it.
As far as controller go with Xylotex they have a 4-axis setup with steppers.
The X3 seems like a big machine to start with your going to have weight issues.
The extra money you save will be needed for tooling and a bunch of other stuff.
Thanks
Frank

I'm aligned with the idea of purchasing a mill for conversion. I will try to point out a few reasons below.

1.
As has been mentioned getting the biggest mill you can afford is a good idea. Unless you are very sure that your interest are limted to micro machining. That extra capacity can make the difference, especially if you start to bolt on fixture plates or other special items. You do however have a size limitation based on your college lifestyle at the moment. In other words moving a big mill around a lot can be a pain in more ways than one. Generally the bigger the mill the better off you are at leaving it in one place setup.

2.
Going the retro fit route gets you into Milling right away. That is assuming one can be found on the shelf someplace. A retro fit done yourself can be done at your leasure. If you take your time sourcing part you can save big dollars finding the required parts on E-Bay and at Ham fests and such. The savings can be significant over buying new. Of course this puts actual CNC operation off a bit in the future in exchange for cash.

3.
Do the conversion yourself and you get exactly what you want. Or at least something that operates up to the limitations of the machine.

4.
It is the limitation of the imported machineries cast iron that is the greatest deterent to the do it yourself approach. Expect that rework and adjustment will be required.

5.
The other thing with the DIY approach is that you can build a solid CNC controller and have it available for future upgrades. If not that then you will have learned a great deal.

Thanks
Dave

I dont think the X3 is the machine that is too big to start with, its still not that large. Its a medium sized bench mill.
Wizard gives good reasons I agree with. Outside of anything else, you will learn a great deal CNCing.

Here is a kit available with ballscrews
http://www.cncfusion.com/smallmill1.html
Its not something I would buy, only because I would make everything myself.

First I just want to thank everyone for their responses. Great information. :)


You guys have me thinking of converting the Grizzly if it would be any faster.

http://www.grizzly.com/products/G0463 That's it right?




I'm not worried about size. I would actually prefer a more rigid mill. It would stay at home by the way. No way I'd operate a cnc mill in my apartment. :rolleyes:

I'm so confused! :)

If you made the stuff yourself, it may or may not be faster than ordering a Super X3 from Syil. Depends on how much time you have, and if you can order everything at once. If you just bought the CNCfusion kit and some electrical setup kit, it would certainly be faster.

That is the correct link to the X3. On a side note, Grizzly's service is superb. I placed my order on a wednesday evening, by thursday afternoon it was shipped. At this point its out of their hands anyways, but New Penn motor Freight had it to my door in 3 business days. Had I have ordered it monday, It would be at my house by Friday. And their tech help has knowledgable folks.

Well I don't have access to any machinery since I'm home for the summer so I would have to just buy the conversion kit.

Glad to hear about Grizzly's service. That's pretty much what I'm looking for.

What controller would you recommend?

Xylotex maybe?

So...

$1014 for grizzly mill Oooo I like the R8 spindle :)
$699 for conversion kit with ballscrews
$600 for xylotex driver box with steppers
$79 for Gas spring

$2392 ?

Am I missing anything? (besides some shipping costs)

What else does the Super X3 from syil have?

THe Xylotex should work, as long as you get hte ballscrew kit. You could use hte machine itself to make all the hardware necessary for a conversion.

The Super X3 has a few extra things going for it, most of which are useless for CNC operation.

It has a tapping function, useless for CNC operation.
It has a more powerful motor rated at 1kw, it is probably not 1kw though.
Tachometer, so the spindle RPM is displayed.
The head rotates 180 degrees, probably useless for CNC operation.
The thing I really like is the head casting, the motor is inside the head casting. The normal X3 has the motor sitting to a side. THis is cosmetic though, I just htink it looks neater.

The conversion from Syil is plug and play, all assembled and tested for you. Time is worth something also. From the pictures, it looks like Syil did a very good job with the conversion. Of course Ronny has not gotten it yet, so I can only make this assumption by what I saw and everybody else saw in the thread.

r8 is nice :)

Yeah I realize Syil does a very clean conversion. I would love to get one of thier mills but doing the conversion myself would be something I'd enjoy. Also the wait would eat me up because I have the time to do it myself.

I apoligize for my stupidity...I said I wouldn't have access to a mill...forgot about the grizzly...lol

I was thinking to not get the Xylotex drive box but instead to just get the 4 Axis System Kit which runs $425. I could just do the casing and setup myself.

Are there any other controller options?

Plenty of other controller options. HobbyCNC has one, but I am not keen on the service, and the steppers might be a little underpowered.

You can buy Gecko drives and make this mill scream :). For something like 114 for their low end model you get industrial quality circuitry, its really a good bang for the buck. Then all you need is a motor for each axis, and a power supply. The motors can be found on ebay, there are some good deals for motors out there. Just NEMA 23 run in bipolar parallel will do it fine.
Power supply's can be found at mpja, other electronic retailres including ebay, or you just build one.

I've been wondering about Gecko drives. How much better are they?

I'm a little confused on their products. Are their drivers separate from thier controller? Where's thier controller products?

Hmm. Ok I think i see. I'd need a step motor for each stepper? Then a G100/101/etc?

I have a Xylotex I recommend getting the geckodrive and putting your own together. For about $800 you can build a higher voltage 3-axis system and there simply wont be any comparison as far as speed and reliability are concerned but thats just my opinion based on what I have read about them and my own experience with Xylotex's $600 box.

I am going to put together my own soon possibly using a 78V torroid power supply sold on eBay (voltage drops to about 72-73 when running), 425oz steppers (undecided about 425oz tho I may choose a smaller motor), pactec DM-4 enclosure, 3 geckos and whatever wiring, fans and connectors I decide on.

Why 78V? I believe everytime the voltage is doubeld the high speed torque increases... so compared to 24V on the Xylotex those same motors would have more than 3x the power. Why gecko? Mid-band resonance dampening. Geckodrive is not the only drive that has something to fix the problems associated with mid-band resonance but it by far the most well known here on cnczone.

Gecko is industrial quality as far as steppers. I have the Gecko's ancestor drive, and it is quite a performer. Higher voltage means your motors will retain torque at higher speeds, as well as being able to reach the higher speeds anyways. The rule of thumb is to have about 25x the voltage rating of the motor in your power supply.

Ok so what parts would you guys recommend? What would I need from them?

You should be good to go with 3 gecko drives, 3 motors, and a power supply.
Of course theres the misc. stuff like wires, a case, fan, little stuff like that.

So like 3 G201s and what else? Are thier G-Rex products thier controllers? This is what I'm trying to figure out.

the step+dir signals are sent from the parallel port of your computer check out the geckodrive forum and look at the wiring diagrams that should answer alot of questions you have about how to use them.

Hm. I see. So what is the G-Rex 100/101 used for?


I'm leaning toward doing the conversion myself. Any part recommendations welcomed.

78 Volt Torroid - http://cgi.ebay.com/39-39V-800VA-Toroidal-Toroid-Power-Amp-Transformer_W0QQitemZ7540253745QQihZ017QQcategoryZ4660QQcmdZViewItem
can also get the current rectifier with capacitors here

PacTEC DM-4 Case - http://www.mouser.com/catalog/623/1355.pdf
You can get *alot* of items at mouser...

GeckoDrive - www.geckodrive.com

425oz Stepper - http://cgi.ebay.com/STEPPER-MOTOR-CNC-ROUTER-TAIG-SHERLINE-NEW-MOTORS_W0QQitemZ260004905714QQihZ016QQcategoryZ78196QQrdZ1QQcmdZViewItem

Other stuff - http://www.partsexpress.com/pe/pshowdetl.cfm?&PartNumber=335-770&DID=7&raid=38&rak=335-770
another source for things like connectors and wire

I am not sure about what wire to use for the parallel port and the steppers maybe someone will have a recommendation about what to use.

Cool. Thanks.

Does anyone know what Gecko's G-Rex products are about? They appear to be a controller but does it have the drive motors built in one package?

Cool. Thanks.

Does anyone know what Gecko's G-Rex products are about? They appear to be a controller but does it have the drive motors built in one package?

G100/Grex is a hardware pulse-generator able to drive stepper or servo "step and direction" amps up to 4Mhz step pulse rate. With it you no longer rely on the pulses being sent through a PC parallel port (usually limited to 45khz if you are lucky), and in a some cases do not even need the PC at all (but that is mostly in industrial applications). It communicates over a Network or USB connection only, no parallel ports. The USB port is for flashing firmware only, network is for controlling.
It also has a load of digital I/O's and some analog I/O's too.
A PC can still be used to MDI, Jog, and load programs for the G100, so if you are used to running a mill with a program like Mach2/3, it will look almost exactly the same (but you use Mach IV (for now)).

It does not contain the drivers for the motors, so a full setup would look like this (how mine is setup):
PC running Mach IV(4) -> network connection -> G100 -> G320's -> DC servo motors.

Hope this helps some.. I know I was confused as heck when it was first being talked about. :)

Yeah I'm a little confused. What are the advantages?

You don't have to explain it to me. :)

It's not something I would need right? Unless its something that's worth the money.

Thanks

Well I ordered the grizzly and some tooling.

Ordered the conversion kit. Should have it in a week to 2 weeks.

Any more suggestions on the Gecko stuff? I plan on placing that order tomorrow but want to know if there's any other opinions on it. I'm either going with 3 G201s or G212s. I'm thinking the extra protection might be worth getting. I also hear that the G201s require heatsinks and active cooling? I planned on enclosing them with a nice 120mm fan anyways but I'm curious about the heatsinking.

I might also go ahead and order a 4th drive as I plan on getting a rotary table soon after.

I really appriciate all the help everyone. I've checked out all of the links and I think I know what direction to go now.

For those who use gecko drives what steppers are you using? Any power supply recommendation? I dont mind buying separate components and assembling it myself. Just kind of wondering what everyone else is running.

Thanks again. You guys are great.

I would not order until Monday.. Nothing can ship until then anyhow.
Before you order, talk to someone at Gecko about your questions and requirements.. They are very helpful.

Very true. :)

I emailed them yesterday. Still waiting on a reply.

Heh. Placed an order with mcmaster for some oil cans and some Vactra #2 and some spindle oil for everything else. What should the ballscrews be lubricated with?

The money I'm saving doing the conversion myself is buying me some nice tooling. :)

I think you made a good choice.

Might be worth getting the 212's. Besides the protection, it offers more choice in microstepping. You probably don't need to heatsink it if you are not drawing even half of its power handling capacity. I have the drivers that Mariss designed and the Gecko's were based off of, this is how it is with those.

Any stepper will work with the Gecko (within its amperage and voltage specs). The link that Smertrios gave to the auction on ebay has some nice steppers, I'de buy those.

As for a power supply. I believe it is recommended to use an unregulated power supply for the Geckos. To accurately calculate your power supply needs, that is not buying something more or less than you need, you should first decide what motors you are actually using. If they are the 8 wire motors in that ebay link, you need to decide if you are going to run them in Bipolar Series, bipolar parallel, or unipolar mode. Each has its own voltage and amperage requirements. I believe I read on the GEcko website that the drive will not draw more than 2/3 of the rated current, so you can factor that in when making your choice. Then add the result 4 times, because it sounds like you will be running 4 steppers for 4 axes.

As for the 4th axis. You might want to take the route Widgitmaster went in his thread. Harmonic drives are fantastic, and are capable of huge reductions. I can't speak from experience directly here, I've never used any of the Chinese rotary tables. But my guess is they have a good amount of backlash, which will not be the most consistent, and probably increase over time.

Post pictures :).

Yeah I'll have pics. Hopefully my mill comes in by this coming friday. I also ordered over $500 in tooling(measuring, cutters, chucks, arbors, cooling system, vise, etc).

I emailed gecko asking if they had thier g212s ready to ship. I can't expect them to jump on an email on the weekend I suppose. :) I think im just going to go with 4 G212s(maybe 3 to save $ for now).

Gotta order the steppers and power supply parts as well.

Sucks being a poor college student. :rolleyes:

Where did you order your cutters and all? Theres some good sellers on ebay so you know.

Yeah I just ordered a basic end mill cutter set and boring stuff from LMS. I wanted to find some ball nose cutters for once I get the cnc up.

YOu might want to look into carbide tooling, the X3 is rigid enough to take it. I just got a few carbide cutters in the mail yesterday, they perform much better than the CHinese HSS cutters.
McMaster Carr sells end mills also, including ballnose. MSC is another good source

Don,t forget to post some pics when you get your stuff in! We love pics !!!!

p.s. good choice on the x3, that or an Zay45 would have been my choice if I were you.

I promise. I love pics too :)

I just hope I can dolly the mill to the back.

Wish we still had our pallet jack. We had it for no reason so we got rid of it. :rolleyes:

Guess I could always jack it up on some pipes to roll it inch by inch. Oh boy. I can just picture myself being crushed by it rolling down the driveway.

What I did was disassemble the mill on my garage floor, and then reassemble it at its final destination (bench). You have to disassemble and clean it all up anyways, so you might as well. Plus it gets you familiar with the machine, this will be helpful when you want to CNC and modify other parts of it.

I didn't even think about that. That's what I'm going to have to do. Thanks. :)

It's gonna be hot when they deliver it. I just know it.

Words from a person who has been down this road before...

Wear a shirt you don't value very much when disassembling, you will get plenty of red grease on you. THe heaviest part is the column. To move the column and head, you will need to remove the head from the plate which actually has the dovetails on it. This assembly will be about 100 pounds.

The table and saddle is probably 80, just a guess. To remove it you will need a C clip remover, the Y axis leadscrew is retained by this. Might want to wear gloves, the burrs on the bottom edges of the table and column are sharp.

GRizzly recommends bolting the mill down to a bench. I opted to use the adjustable feet, it makes moving easier. This mill does not move at all during milling, never been a problem.

I found no lubrication inside my X3's head. Theres a few gears in there, not sure why they did not put grease in. Regardless, its a pain to remove everything. THe best thing you can do for yourself is get a can of white lithium grease that sprays, any auto store has this.

Lastly (for now), most places where there at two things bolted together have more than just bolts holding them together. There are also pins, used to align the pieces. THis iis the case for the column bolting to the base casting, the head to the dovetail plate, everything.

JL123,
Look, I just jumped on your thread. You asked a couple of good questions and I did not see where they were answered. If they were, I may have read past them.

The controller will be your software which will interface with your stepper drivers (gecko 201, 211 or whatever), which for most of us our software is Mach2 or Mach3 from artofcnc.com ($159). You computer will communicate through the 25 pin parellel printer port.

You should buy a breakout board which will break out the 25 pins of the port and give you a place to connect the STEP and DIR signals for each of your axis (X,Y,&Z). (these can cost $9.99 on ebay up to $150 for optical solated, your choice)


You can download a demo version of Mach3 for free and play with it. Now at the gecko website you can download a tutorial PDF which covers stepper motor basics (assuming that you are going with stepper motors)

There is a good artical about building your power supply, maybe in the electronics forum. They are basically a simple linear power supply, consisting of a transformer, large filtering capacitor, full way bridge rectifier, optional bleed resister, an input fuse (120 vac what ever amperage required) and fuses for each of the gecko drives. You will need heat sinks for each of the gecko, and a fan for cooling.

I bought some Keling stepper motors that seem to work fine for me, they are made in china, so they match the china machine I put them on. Now with motors, you need to pick nema 23 or nema 34 as for the physical size as well as their electrical strength which may be specified in in/oz or in nm(i believe) the convert between nm to in/oz multiply nm X 141 if memory is correct. (one of the guys reading this correct me if I wrong). Then you are going to connect the motor according to it's diagram either bipolar series or bipolar parellel and set the current output of the stepper drive to match the max spec output from the motors spec for that configuration.

The cabling to the motors, you will need probely a 6 conductor 20 or 22 guage (depending on your current draw) and figure out which connectors you want to use.

Now with your gecko drives, you will set you current limiting on each drive. This is determined by the spec'ed max current on each motor. All of this info will be in the Stepper motor basics PDF.

Now, with Mach3 (if thats the controller software you want to use) you will do through a series of setup pages which will allow you to specify which pins (on the 25 pin printer port) that the STEP and DIR for each of the axis.

To set up the controller, you are really just specifing what pins of the port are used for what function. This includes homeing switches and limit switches.

When you download the Mach3 demo, also down load the user's manual (which I think is still labeled Mach2) and it will guide you through the setup.

With mach3 you need XP operating system and you can specify whether you will use 25000 or 35000 kernal speed. Anyway, you will read about all of this kind of stuff as you get into it. It's not hard, it is just a lot of details.

Ask questions as you go.

If you have already figured out what I just wrote, it may help the other guy who is reading it also. Also, for more info, you may want to download the Tormach PDF manual (which is also based on mach2) because it is also very good for explain the controller and it's interface.

Hope I filled in some gaps. Keep asking questions.

Ron

Great write up sir. :)

I was confused about a few things but figured them out after reading through the stepper motor basics on gecko's site.

Now I need to think of what to make with the manual mill without a rotary table. :)

Theres a million and one things you can make on a mill, even without a rotary table :). Maybe start by modifying the mill itself? My next project is to replace the motor and gears with a belt driven, more powerful DC motor.

Haha I know there's a lot that can be made.

For some reason I just can't think of anything off the top of my head. It'll be different once I get it I'm sure.

Hey phantom do you have any pics of your setup?

Few pics of some of the small stuff I've been gathering. I was going to wait to buy some higher quality measuring instruments but I figured I could always just have these around to abuse.

I plan on snatching some starrett tools off ebay after the conversion.

I wanted to order some roughers and ballnose cutters as well but haven't gotten around to shopping around. I'll probably place an order for them pretty soon(or find them on ebay).

I should have the mill tomorrow. :)

I have some nice aluminum angle as my first victim. They're motor mounts for our old bass boat.

GOT IT!

I got it this morning at 10:50 and the truck driver was nice enough to drop it off in the back. I live behind a huge bingo place with a wide alley so he was able to back his rig in. I took my time and carefully got it off the pallet onto the cart without banging anything up. I really dont want to take it apart outside since it's so damn hot. The high is going to be 107F today. :eek:

I should have it setup on the bench tonight with some help from my brother and dad. I just love the look and feel of it.


I'm giving that cheap harbor freight cart a run for it's money. :)

Pics as promised.

JL123,
Congrads, it's a 325lb babby boy!! Where's my cigar?? Well what's next on the list. Picked out your steppers and drivers yet?

You have alot of neat stuff ahead of you and I know you can't wait to dive in. Take plenty of pic's and you will find these guys will bend over back wards to help. It's funny, the more pic's the more help you will recieve.

Ron

Yep, thats a scene I am very familiar with :). Mill looks exactly like mine.
I finally got to use my 1" end mills since I received my endmill holder. Keep in mind its 6 flutes and a rougher...
But the mill finally started to chatter at .5" deep. So I took .25" cuts no problem then, no need to overwork for the motor.

Its been 97 here in NH. Just think, if you take apart the mill you will sweat, and this will cool you off :)

Sorry, missed your question about pictures.
I took a few with my camera phone a week or so ago, the purpose was to show a friend the actual size of the mill.
http://home.comcast.net/~phantomcow3/X3.jpg
http://home.comcast.net/~phantomcow3/X3_2.jpg
You can just add a # in like X3_3, some of the pictures have been deleted because I could not seem to stay still while I took pictures!

Thanks everyone. I've always been a picture taker in every hobby I've had. :)

I'm placing my order with Gecko today. I'll get components off ebay for the power supply also.

Quick question. Will those double shaft steppers that Smertrios suggested pose a problem for the Z axis? Are there any advantages to having handles on dual shaft steppers? Obvious one would be some ability for manual control but is it something useful?

Joel

Advantages to having dual shaft steppers...
You can mount a handle on. THis means you can retain manual operation of your mill with a few additions. Some say its useful for setup, on the CNC machines I work I use the manual pulse generator for all that. You should be able to use the keys in your controller program fine.
Also, you can mount a rotary encoder if you wanted a DRO that way, or to make for a closed loop system.

Ditto with PhantomCow2,
It's also nice to have handles for moving the table for cleaning. Also, I bought a Targus Laptop Keypad that I use for local jogging control (got it on sale for $15 at Office Depot) you use Mach3's keygrapper program to assign key function to the individual key. This unit just plugs into your computer's usb (through a long extention cable).

Ron

Sounds good. Didn't think about the encoder. That's always a nice option. I'll go with those then.

I wanted to get a keypad. Could you actually get a joystick for manual jogging? :)

Congrats, I,m sure you,ll be happy with that puppy !!!!

Good how to build a power linear supply for your controller. I think this shows everything except a bleed resistor across the filtering capacitor.

Linear P/S design/construction
http://www.campbelldesigns.com/

Ron

Thanks everyone. I've always been a picture taker in every hobby I've had. :)

I'm placing my order with Gecko today. I'll get components off ebay for the power supply also.

Quick question. Will those double shaft steppers that Smertrios suggested pose a problem for the Z axis? Are there any advantages to having handles on dual shaft steppers? Obvious one would be some ability for manual control but is it something useful?

Joel

The CNC Fusion kit uses a size 34 motor for the Z axis.

Totally forgot about that....good thing I was in the shop til 1 in the morning cleaning the machine up and didn't buy the steppers. Thanks for bringing that to my attention. :)


Here's a few more pics. That closeup sure does show a rough cut. :rolleyes: Looks better in person.

I got it on the bench without taking anything apart by hoisting it up onto that bench(/w rollers). Used a chain hoist come along. :)

I share the shop with my brother and dad(14'x14' /w storage upstairs). Space is limited but with a little cleaning I'll have plenty of space. Once my mom and grandmother clean out thier stuff we'll get our old 12'x24' shop back. :D Hopefully I'll get it for myself and my brother and leave my dad with the other one.

Hi,

Your machine looks awesome and that cut doesn't look too bad at all - I need to save my money for something like that...hmmm...

I'm pretty sure I've missed this in the thread already, but are you going to use a CNCFusion kit with ballscrews?

Keep the photos coming!

Regards
Warren

Yeah I'm going to be using the cncfusion ballscrews. I plan on snatching some ground ballscrews off ebay(if they're cheap) later down the road depending on how good the rolled ones from cncfusion are.

Still deciding on whether or not I want to bolt the machine to the bench or just use the adjustable feet. Since the bench has rollers I'm going to have to cut some wood blocks to wedge under the bench and lift it off the rollers so it's more stable.

Any thoughts on this stepper for the Z?

http://cgi.ebay.com/NEMA34-680oz-in-Bipolar-Stepper-Motor_W0QQitemZ260008847200QQihZ016QQcategoryZ78196QQrdZ1QQcmdZViewItem

Still trying to find some nema 23 steppers for the x and y. Is there a brand I should go with? I was going to go with the ones suggested but the auction is over.

Maybe http://www.xylotex.com/StepperMotor.htm ?

Thats a lot of torque, probably much more than necessary for the Z axis if you have ballscrews. Remember ballscrews pretty much cut your torque needs by a third compared to acme. Also, it has pretty high inductance. The torque will be there, but the torque at speed will not.

Oh ok I see. I'm going to be using the gas spring on the Z as well.

Is there a shop online that sells cnc parts such as steppers, servors, etc?

http://cgi.ebay.com/NEMA34-SINGLE-SHAFT-STEPPER-MOTOR-400-oz-in_W0QQitemZ110011014588QQihZ001QQcategoryZ78196QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

Would that be any better? What range should I be looking for?

x,y http://cgi.ebay.com/NEMA-23-DUAL-SHAFT-STEPPER-MOTOR-280-OZ-IN-CNC-Router_W0QQitemZ110011012210QQihZ001QQcategoryZ78196QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

JL123,
Those 680 in/oz should do the trick. I would also recommend looking into using a gas spring on the Z axis, which will reduce the work that the Z stepper would have to do. Some of the guys will use a counter weight with a pulley while others may choose a gas spring. I think somewhere around an eighty pound would be about right. If I'm off some one correct me. There is a PIC of a dual set of gas springs on a IH machine on a recent tread"What size Z axis Motor" maybe on the IH forum.

Any way with the IH dual spring setup, the guy has it fairly well balanced to the point where he says he can move it up and down with something like a 120 in/oz stepper and a 9 volt battery. That's funny, but it give you an idea what a pair of gas strings will do for you. McMaster Carr would be a good source and someone said around $20.

With the spring, you would not need the 680 in/oz, but it would be nice to have an larger than necessary in case the gas spring went out, leaving you dead in the H2O.

Nema 23, I used a Keling 453 in/oz (China) on a lathe and was happy with it. ($54 I believe, they advertise on this website and I got them in 2 or 3 days). See what the other guys recommend.

Ron

I will be running dyno data on several Keling step motors (NEMA23, NEMA34) and a NEMA34 DC servomotor. The motors I got look very nice. Stay tuned.:-)

Mariss

Yeah I'm looking to buy from kelinginc on ebay. Like I said I'm going to be using the gas springs so I think I'll go with the 400oz stepper on the Z from kelinginc.

I'll wait for a few more comments. :)

I will be running dyno data on several Keling step motors (NEMA23, NEMA34) and a NEMA34 DC servomotor. The motors I got look very nice. Stay tuned.:-)

Mariss
Look forward to that info Mariss :wave:

I want that 425oz Keling motor myself, I hope he is relisting it.

Oh here it is,
http://cgi.ebay.com/NEMA-23-DUAL-SHAFT-STEPPER-MOTOR-425-OZ-IN-CNC-Router_W0QQitemZ110009944852QQihZ001QQcategoryZ78196QQssPageNameZWDVWQQrdZ1QQcmdZViewItem
But I am also looking into HOmeshopCNC's new motor here....
http://www.homeshopcnc.com/page4.html
The specs seem almost too good to be true

Excellent. I think I'll go with those. They'd be good for the x and y right? Then getting the 400oz nema34...I think.

Also after talking with gecko I think I might go with a G-Rex 100 and 3 G202s. I'd rather go back and forth and get the right stuff at the end than rush it and regret it. :)

Phantomcow2,

That is a nema 23, and that may have been the one I used on my minilathe, I think I said a 452in/oz., but it was a 425 in/oz

Go to his website http://www.Kelinginc.net and you will see a button (on the left) to show what stepper's in stock. Same price that he has on ebay. He is also listed on this website under advertisers or you can click on his banner to go to his site.

I just looked and he does have a 396 NEMA 34 single and dual shaft for $79 or so.

Ron

JL123,
educate me. What's the G-REX 100 going to do for you? I have not looked into it at all.

Ron

That happens to be one of the 3 motors John at Keling was kind enough to submit for dyno and accuracy evaluation with our drives here. Too bad the motor is not drilled-out and tapped for an encoder mount. If it were I could have used a G100 to measure its cyclic positioning error as well.

I'm snowed thru this week with other things. Right now this weekend looks like a good time to take data.

Maybe I should start something like "Tom's Hardware Guide" to do testing data and reviews on motors.:-)

Mariss

Those will work for the X and Y, it is overkill though. Have you looked at the videos from CNCfusion.com?
On his forum he said he used a 166ozin motor for his X, and a 200 for the Y. This was with the stock inefficient 10tpi acme screws. With a ballscrew, you are nearly 3x as efficient, but it is twice the lead. Still, thats an improvement. He was taking some pretty good sized cuts and that mill was moving.

That happens to be one of the 3 motors John at Keling was kind enough to submit for dyno and accuracy evaluation with our drives here. Too bad the motor is not drilled-out and tapped for an encoder mount. If it were I could have used a G100 to measure its cyclic positioning error as well.

I'm snowed thru this week with other things. Right now this weekend looks like a good time to take data.

Maybe I should start something like "Tom's Hardware Guide" to do testing data and reviews on motors.:-)

Mariss

Does the G100 allow for closed loop operation? I have a few Sumtak high resolution encoders on some IDC steppers that have been sitting in idleness....

Mariss,
Your G-Rex 100, is that used to relate position from a stepper or did I not understand your post? Something to make a stepper system into a closed loop system? Pardon my ignorance.

Ron

Well I'm not the person to ask but Art from artsoft said the performance was like night and day. Pretty much instead of having mach 3 doing all of the pulse generation through the parallel port it insteads sends lines of code to the controller(G100) and the G100 controls the drives directly. Any deeper than that I can't explain. I have to read up on it myself.

Joel

JL123,

Personally, I would rather have a little more than I need. On the X and Y, I would go with the 425 nema 23, but that's me. Maybe it would work fine with the Keling 282 in/oz Nema 23. And I don't mind stating my opinion, hoping that if I am wrong, the strong fist of this forum will give me a good bopping and straighten me out. But, I would rather be overpowered and not use it, than be underpowered and need it.

OK, guys everyone show your hand and state your cases, if there is an evil to having too large of a stepper.

Ron

Ron, I feel exactly the same way. Which is why I haven't ordered anything yet. I'm torn between making sure I always have enough power and wondering if a weaker stepper would offer better performance in terms of speed.

I don't want to hijack this thread into a "the G-Rex is great" thread. I do think it's great and has an unbelievably teriffic future though.:-) Maybe a Moderator can move it elsewhere.

The G-Rex (G100 and G101) connects via Ethernet or USB and offloads the PC from having to do step pulse generation timing / counting donkey-work. A PC is not meant to do that; its purpose is to do far smarter stuff.

The G-Rex hardware has 6 axis and each axis has its own quadrature encoder feedback inputs. The hardware is designed to ultimately have closed-loop, adaptive feedback where a step motor cannot stall. Overload the motor, it slows down and backs up its speed-torque curve to find more grunt (torque). 2-D or 3-D motion would never leave the programmed path; the involved motors would just slow down. Relieve the load and all would pickup back to their former speed.

That's the goal; it's provocatively called the "unstallable motor project". It is also devilishly difficult to solve and why not? If it were easy, it would have been done before. The hardware is there; what stands between now and the goal is a lot of work and time.

Don't even ask when. Maybe 6 months, maybe more. It will happen and it will use the G-Rex.

Mariss

Don't worry about a hijack. We all want to hear it. I'm the ignorant fool that's been emailing you. :D

I think you've talked me into getting the G-rex. Just seems like there's such a good future for it. :)

Ron,

Yes, there is such a thing as an "overpowered step motor" and you will not like it should you have one. Here's why:

Most people confuse gobs of torque with power. If a 200 in-oz motor is good, a 400 in-oz must be better and so on is the thinking. Wrong.

Power is what gets things done and power (Watts) is torque times speed. Get the motor's speed-torque curve. Multiply torque times RPM. The motor that has the bigger result will do more. You can have a 200 in-oz motor that is still trucking at 1,000 RPM while a 400 in-oz motor can't get out of its own way at 500 RPM.

Think "power", not "torque".

Mariss

JL123,

You are going to have to get a little more info on the G-REX 100, sounds pretty impressive. To me it seems like a hybrid, between a stepper and a servo. And it sounds like to use it's full potential you would need encoders on each of the stepper motors. That may explain Mariss statement about using the GREX 100 as an instrumentation system for testing Keling's stepper motors.

At this point in time, find out the advantages specifically, for the incorporation of GREX 100 into your system. I guessing, that it currently could be used in a open loop configuration. As Mariss says, it will do most of the grunt work and free up your processor, and then as Mariss's developement continus you could add the encoders and have the projected unstallable stepper motor system.

Mariss, you comments fit into this forum, they are not viewed as advertising, but as necessary discussion in the design of Mr. JL123 system.

Please correct any misconceptions that I have (pardon my spelling)

Ron

I don't want to talk you into a G100 unless you need it. Mach3 is a terrific program; I deal with about 15 application phone calls a day and twice that by email. Easily 80% mention Mach3 as their CNC software so Art of Mach3 has to be doing something very right.

Most of those calls/emails deal with the parallel-port version of Mach3; there are only a few hundred out there that have a G100 and use the G100 version of Mach3.

The lesson in that is the parallel-port version serves 99% plus users very well. The G100 users are entirely at the high performance end of applications.

The G100 costs $400. That is a lot of money. You really need killer high performance to justify that cost.

Mariss

I built a counter weight system for my x3 and had used one for a while on my cnc'd x2. It is a compounded pulley system so I need only use 1/2 the weight of the head in counter weight. I use 44lbs which acts as 88lbs at the head. It is not quite enough to have the head be neutral or floating.The x3 head is pretty darn heavy.
With this system I can actually flick the crank handle for the z axis with my pinky finger and have the head move.
So certainly no need for a larger motor on z axis if you use this type of system.
I am going to use 425 oz nema 23 steppers on all 3 axis and another on a rotary table for the 4th and use the xylotex kit.
Just going to use cheap ball screws with double nuts.
It's a budget build but fully capable of making very good parts.
I'm not all caught up in the speed of the machine as some of you guys are as I just don't see much use for it for the type of work I do.
I'll begin the project in a week or so.
Here are a couple pics of the counter weight system.
http://i109.photobucket.com/albums/n48/S_J_H/1.jpg
http://i109.photobucket.com/albums/n48/S_J_H/2.jpg

Steve

Mariss, can the G100 be used in an open system? :)

If so I think I'm onboard. Would you recommend the 280oz steppers for the x and y then? What about the Z?

How do you mean "an open system"?

I can't say about Mach3 but for the G100 and our GeckoMotion firmware it is completely open. Even the source code is available and commented for your use. Keep it or change it in any way you like; incorporate it with any program you write in any way you like. It is in the public domain without any private or commercial use restrictions whatsoever.

Why? We manufacture hardware. Anything we can do to promote the ease of use of our product helps us as much as it helps you.

Mariss

Well I just mean like having a system like this

Mach 3(G100 version) > G100 > G202(or whatever) > steppers

with no encoders...

jl123-

I think you might be asking if the G100 can be used in an "open system" as in it can drive steppers without encoder feedback. If this is what you want to know, then yes it can. It can control steppers, the typical DC servo motor w/feedback (but that feedback is still tied to the servo amp ie- G320/G340, not the G100), and eventually steppers w/feedback.

Hope this helps!
Colin

Yep. I figured it would I just wanted to make sure. Thanks.

Larger motors typically have more inductance, especially ones that have a "Bang for their size", that is, larger torque with low current ratings. High inductance causes reactance, essentially blocking current as the speed increases (function of frequency). Take a look at these videos:
http://www.cncfusion.com/images/X3mounts/x3%20video/Video.html

He says he used only 166oz for the X, and 200 for the Y. Not sure what input voltage there is, but we know that is enough torque then.

hi

jl , try this guy www.cnckits.net mention my name " Ivan " im sure u will be impressed with his setup and price you can beat

cheers

Well I ordered both the 425s and the 280oz steppers. I'm curious about the 425s but I think I'm really going to use the 280s. I can never have too many steppers I guess. :)

G100 and G202s are ordered. Now going to get the power supply components, cabling, chassis, etc.

Quick questions. Some of the the wires from the steppers have to be in parallel. Is this done right outside the stepper or do you take all 8 wires to the chassis to do it there? Why do people use 6 conductor cable if there's usually 4 wires hooked up to the drivers?

I'm just trying to plan everything out before buying to give me a nice result. Thanks. You guys have been great. :)

I'll begin the project in a week or so.
Here are a couple pics of the counter weight system.
http://i109.photobucket.com/albums/n48/S_J_H/1.jpg
http://i109.photobucket.com/albums/n48/S_J_H/2.jpg

Steve[/QUOTE]


Steve,

The pics are not working for me but am interested. Are they somewhere else?

Thomas

They're working for me in your post. You're missing a [Quote] at the beginning btw.

Alright...

If I'm using these steppers.

280oz Ebay link (http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ih=001&item=110011012210&rd=1&sspagename=STRK%3AMEWA%3AIT&rd=1) Specs (http://www.kelinginc.net/KL23H276-30-8B.pdf)

1.9v bipolar(p)

38v-47.5v (20x - 25x)
------------------------
400oz Ebay link (http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&ih=001&item=110011014588&rd=1&sspagename=STRK%3AMEWA%3AIT&rd=1) Specs (http://www.kelinginc.net/KL34H260-42-8A.pdf)

1.6v bipolar(p)

32v-40v (20x - 25x)

So I should be aiming for a 40VDC Power supply right?

Marlin P Jones has a transformer in stock that I believe will give you 35vDC. THis will be suitable

Would it be possible to use a 31V+31V transformer and get around 44VDC. Would that be a little too much? I was thinking that once the load increases that voltage will drop to the safer 40VDC level.

Or should I just go with a 25 + 25?

44vdc is probably pushing it. It will work, but I think you will find motor heating to be excessive. Heat is just waste then. Granted steppers can get as hot as 200 degrees, almost enough to boil water. There is no need to push it
Dave Rigotti of HobbyCNC says there is hardly any noticable difference between 35v and 42v, or something close to this.
Those 400oz motors might not be able to achieve very high speed anyways.
I have a 28vDC Coutant Lambda power supply, 14A. That is what I will be using, I've used it on steppers before, it was fine.

Look what just came in. Sorry for the background. :)

Order took 13 days to get to my door. Not bad. :cool:

Marlin P Jones has a transformer in stock that I believe will give you 35vDC. THis will be suitable

hi phantomcow2

does Marlin P Jones have a website for his power supplys .

I'm in the process of ebaying some stepper motors , big suckers for one of my big lathes here at work

NEMA 34 HIGH TORQUE STEPPER MOTOR 1812 oz-in, ½” Single shaft with flat, 180 , KL34H160-62-8B ½” Dual shaft with flat: Price: $189 ( in stock)

from http://www.kelinginc.net/SMotorstock.html

I need a correct power source for these units , also does anyone know of stepper motor drivers for this size stepper ??

cheers

Look what just came in. Sorry for the background. :)

Order took 13 days to get to my door. Not bad. :cool:

Hi jl

ok ok who did u buy this kit of and do they do other kits for other machines


looks very nice , i gather thats for the sieg x3 mill ???


what sort of dollars did it cost u ???




cheers

The complete kit was 699 with all the mounts, bearings, ballscrews, etc. I got it from cncfusion.com. Their site seems to be down at the moment but yeah it's a kit for the Grizzly X3. I think it might work on other X3s if they resemble the Grizzly x3.

I'm going to wait to put it on until I get everything else.

http://www.mpja.com

For those beasts, you can use the Gecko Drives IF you wire them in series. You could wire in parallel but would not get the full torque because of the 7amp limit of the Geckos. I hope Mariss does not mind me suggesting Centent, their CN0165 drivers are good for 10 amps. Expensive though.
You are in the range of commercial drivers. Call Danaher motion, they will direct you to Pacific Scientific, they will get everything you need. Got a friend who has 42 frame steppers about the same torque as yours for his bridgeport, driven by Pasci. Drivers of this caliber (like his) are 117AC input, power supply is integrated into each driver. I hope you did not expect these to be cheap....

The kit was bought from cncfusion.com, they make CNC conversion kits for other Sieg machines, the X3 is the largest. FOr hte ballscrew kit, its 700 dollars. TO me it seems silly to buy a mechanical conversion kit when you have such a great little mill at your disposal, making hardware gives you a chance to familiarize yourself with the limitations of the mill. It sucks when you try to take a cut too deep/fast for your machine with CNC operation, you can't just stop turning a handle...

Well I would have made them myself but my tooling is pretty limited at the moment and I'm not familiar with doing the conversion from scratch. Silly for you maybe but I didn't want to make a mistake that would set me back.

Jl123

That kit looks good. I will be very interested in your progress as you fit it all up and complete the conversion.

I have ordered the kit from Syil which is $200 dearer.
Syil's ballscrews I suspect are a bit better but wouldn't bet the house on it :D

When I get mine I will be logging my experiences also so it will be interesting to compare.

Greolt

I'm not too concerned with having the best at the moment. I just want to get started. Like I said I plan on maybe upgrading these components later if it would improve performance. If not then maybe a better machine later down the road.

Everything is ordered. There are still some odds and ends I need to hunt for such as cable that goes to the steppers. Anyone have any pictures that show how steppers were wired up physically. What kind of cable should I use? I've only been able to find some suitable cable from ebay. I keep seeing "use 6 conductor" but would there not only be 4 wires going to the driver? :)

Other stuff includes cable management but I'll get that locally once I start building the power supply. I plan on building a 19inch rack for the powersupply/controller and computer. I actually have an old relay rack that could be used. It'd be nice to have both in a cabinet safe and secure.

Totally forgot about that....good thing I was in the shop til 1 in the morning cleaning the machine up and didn't buy the steppers. Thanks for bringing that to my attention. :)


Here's a few more pics. That closeup sure does show a rough cut. :rolleyes: Looks better in person.

I got it on the bench without taking anything apart by hoisting it up onto that bench(/w rollers). Used a chain hoist come along. :)

I share the shop with my brother and dad(14'x14' /w storage upstairs). Space is limited but with a little cleaning I'll have plenty of space. Once my mom and grandmother clean out thier stuff we'll get our old 12'x24' shop back. :D Hopefully I'll get it for myself and my brother and leave my dad with the other one.
the cut you took on that alu part was taken at about what feed rate plus or minus

All you need to wire up steppers is 20guage multi strand wire, I got mine at radioshack. Even then, your steppres might have wires long enough to not need anything.
Strip the housing off the very ends of the cable, stick them in the terminal, and tighten the little screw. Don't give connecting the steppers too much thought, its pretty simple.

The drives will say like A and A with a line over it. B with a line over it, and just B. So look on your spec sheet for the motor, connect the wire that is designed A into terminal A.

Here is a picture, look at where the wires interface with each driver.

Here is a picture, look at where the wires interface with each driver.

Humm interesting picture :p

JL123,

goto the keling website, just click on their advertiser's link from this forum and you can download the spec sheets from their web site for each of the 280in/oz and the 425 in/oz (i've used both of these and they worked for me). If you can't find them, let me know and I can mabe attach the spec sheets to a post or email them to you.

I think that I used the same transformer that phantomcow is speaking about from Marlin Jones. It was a 24VAC and when rectified and filtered it will give you around 35.5 VDC.

The leads from the keling stepper motors are only about 1 foot long, so you will need hook up wire, and 18 to 20 guage will do the trick, just make sure it is stranded.
Any wires that you tie together, you can do that at the motor and keep from having to buy 8 conductor cable.

Ron

Yeah I actually posted links to the spec sheets a few posts up. I went with a 28vac transformer. Rectified it would be around 39 volts and as current draw increases that voltage will go down lower.

Is there any way to get into the shell of the stepper? Would there be any room to do connections in there? I'm counting on this as I really want a clean look.

I dont think you understand.
The stepper motors will have 4 wires sticking out. You connect these wires individually to the stepper motor driver. THere are 4 wires coming from stepper motor to driver. You can get some cable enclosures from radioshack. Here is a link.
http://www.radioshack.com/product/index.jsp?productId=2062665&cp=2032058.2032237.2032318&cp=2032058.2032237&f=Taxonomy%2FRSK%2F2032318&categoryId=2032318&kwCatId=2032058&kw=cable&parentPage=search
Then it will have the illusion of one big cable, all nice looking.
This stuff will do hte job too, though larger diameter.
http://www.radioshack.com/product/index.jsp?productId=2103809&cp=2032058.2032237.2032318&cp=2032058.2032237&f=Taxonomy%2FRSK%2F2032318&categoryId=2032318&kwCatId=2032058&kw=cable&parentPage=search

Try to avoid disassembling a stepper motor, there is nothing you can do anyways. Inside are probably the most powerful magnets you have seen and they can and will pick up anything ferric. Yes, I have made this mistake before.

I would not go into the motors. I used molex connectors from mouser page 1079 on their on line catalog and used 12 inches of heat shrink to cover the wires from the motor where I put the first plug. Made a 10 ft cable which ran from the motor plug and connected into a bulkhead connector on my controller. These connectors have nylon tabs that can lock into retangular holes. You order male and female plugs and male and female pins and can solder the wires into the pins, or buy the $175 crimper.

Just to give you an idea of what I used,

Ron

Hi guys, I have been following this thread for a few days. It has been very helpful for me. I would like to go with an X3 doing my own conversion - or possibly using the cncfusion kit. My problem is that I need to have more Y travel. Could the Y axis be extended on the X3 and if so; how much and how would one go about it.

Thanks for your help

How much do you need?
You can get 6.25" of Y travel, the width of the table. Not sure how you would get more.
I think you are better off getting an RF45 with 9+ " of Y axis travel in this case.

Yeah Ron I actually was going to use those molex plugs. I think they have them at radioshack. I also have the crimper for the contacts already. :)

Heres a link to some wire that I was thinking of using on my conversion. Its 18 gauge but I think its available in 20 as well.

http://www1.mscdirect.com/CGI/NNSRIT?PMPXNO=1946121&PMT4NO=9872739

Festess,
Took a peek at your wire, could not tell if it was stranded or solid. If it is stranded, it is ok. If it is solid, it would not be a good choice. You may already know this, but sometimes we will include little tid bits of info for the guys who are reading these post over your shoulder. Stranded wire is preferred because it will hold up to the repetative motion that a cnc machine would put the cable through. Solid wire is more likely to break over time.

If your cable choice is stranded, it would ok for cnc application.

I look forward to seeing your conversion, make sure you post lot's of PIC's!!!

Ron

JL123,
Good sir, tell me about your crimper (I'm hoping you have an cheap source for one), the price at mouser scared me off!!!!!!

Ron

After looking a little closer at the msc catalog I couldn't tell if it was solid or stranded so I'll go with something else maybe like McMaster-Carr catalog number 9936K12. JL I've been following along with this thread and have learned quite a bit. I think I'm almost as anxious as you are to see your X3 cnc'ed. Keep up the good work.

Thanks festess :) How do you know how many conductors it has. Does the 20/4 mean 4 conductors? I would think more conductors of the same size equals more current capability...

Ron I haven't looked at those molex plugs too closely but I'm assuming they use the same type of crimp that the 3 and 4 pin molex uses.

http://www.frozencpu.com/slt-16.html

Works perfectly.
I just checked it out and it should work exactly the same. It's the best 20 dollars I've ever spent on a tool. I'm stocked with 3 and 4 pin molex hardware. Very handy for just about anything.

Yep thats it 20/4 has 4, 20 gauge multi-stranded wires in it.

That looks like the appropriate cable. It even says for cnc applications so I think I'm going to place an order.

Some more stuff came in today so I went ahead and bolted on most of the conversion kit already. One question I have is there's nothing to stop the Y axis from pulling out the ballscrew from the bearings so am I going to have to use some limit switches or am I missing something? I guess the same applies to the other two. Any recommendations on limit switches? :)

Joel

Are you using Mach3? I believe this program has limits set within the software.

Yeah I'll be using the G100 version of Mach3. So would this be enough insurance? Thanks.

Doesn't the ballscrew include a threaded section for a nut so that you can preload the bearings. Will this nut not also prevent the ballscrew from pulling out of the bearing, or have I misunderstood something.

Regards
Phil

Some more stuff came in today so I went ahead and bolted on most of the conversion kit already. One question I have is there's nothing to stop the Y axis from pulling out the ballscrew from the bearings so am I going to have to use some limit switches or am I missing something? I guess the same applies to the other two. Any recommendations on limit switches? :)

Joel

After a few(okay maybe 5) solid hours of reading about backlash and preloading ballnuts I'm finally off to bed. There are no threaded parts on the ballscrews.

Few links I found.

http://www.rockfordballscrew.com/manual.htm
http://www.kerkmotion.com/learn/kerktech/anti-backlash-overview.asp
and attachment

Not sure what I'm going to do yet. The Y axis is pretty good with little backlash. The X axis needs some work because it's so much longer I'm guessing. Looks like I'll be doing some preloading soon. I totally overlooked this when planning this out but oh well.

I'll have some pics later in the day. Time for some shut eye.

Doesn't the ballscrew include a threaded section for a nut so that you can preload the bearings. Will this nut not also prevent the ballscrew from pulling out of the bearing.............l
This is something I have wondered about. CNCfusion kit doesn't seem to have bearings included and relies on the existing X3 ones.
I tried to have a look at the instructions on CNCfusions site but it is down at the moment.
So I haven't checked how they load the bearings. No nuts.
Do they use a lock collar? Or even the coupling to hold it all together?

This is a pic of what Syil have for each axis
http://www.cnczone.com/gallery/data/500/Bearings.jpg

JL123,

Like Phil Burnam, I am also wondering, if the ball screw is not threaded, how are bearings slightly loaded, and the screw staying in place. Typically, what I've seen in the past, at least on the forums here. The ball screw is turned down to say a 12mm bearing journal that may be say 18mm long (this would be less the width of two 10mm wide angular bearing, for example) and then the next additional 1/2 inch or so is threaded, and the last 1/2 inch to 3/4 is turned down to the id size of the coupler or pulley that is to be used with a flat spot for the set screw to lock on.

What is the method that is used by the cncfusion kit? Phantomcow2, where are you and what do you think?

The Plot thickens,

Ron

Oh yeah the plot thickens....

Sigh...

Whoops?

I had to drill out a plain washer to put before the bearings on the ballscrew because the stock washer/bearing assembly would start to climb the thread because its so loose. I have to say the washer I made fits like a dream. :cool:

This doesn't look good. If you cannot preload the mounting bearings (not the ball-screw) then you will have backlash, which kinda makes using a ball-screw pointless. If you bought it as a kit ask the supplier. I can't believe that the design doesn't include a way to preload these bearings, but I can't see it in the photos.

Also I can't see tha the ball-screw shaft includes a shoulder for therbearing to be preloaded against. It looks like the ball-screw needs some additional machinng/threading in order to work.

Regards
Phil

Sigh...

Whoops?

I had to drill out a plain washer to put before the bearings on the ballscrew because the stock washer/bearing assembly would start to climb the thread because its so loose. I have to say the washer I made fits like a dream. :cool:

CNCfusion looks like it is still down, maybe permanent. Not suprised judging from the lack of design for mounting the ball-screw. Or am I to hasty.

Regards
Phil


This doesn't look good. If you cannot preload the mounting bearings (not the ball-screw) then you will have backlash, which kinda makes using a ball-screw pointless. If you bought it as a kit ask the supplier. I can't believe that the design doesn't include a way to preload these bearings, but I can't see it in the photos.

Also I can't see tha the ball-screw shaft includes a shoulder for therbearing to be preloaded against. It looks like the ball-screw needs some additional machinng/threading in order to work.

Regards
Phil

I'm new to all of this of course so I don't understand what you're talking about when you say there's no shoulder to preload the bearings. I thought the ballnuts got preloaded in the middle or is this done on the drive side?

If a new block was machined to allow 2 ballnuts in the middle to be preloaded would this work?

The site has been down before. I think you might just be jumping the gun...hopefully. :)

Good pictures jl123.. I too have a little Gecko buddy, but (s)he is green, and will have another (different color??) friend in a couple of days. :)

It looks like the ballscrew uses the two thrust bearings which are sandwiched in place using a set-screw locking collar?? I guess if you can push the assembly together hard enough while locking the collar down, it might hold OK. But as soon as a heavy cutting load is put on it it would seem that the collar could move.

Oh ok so this is where having a threaded part at the end of the ballscrew comes into play? Anyone have a picture that shows an ideal machined ballscrew?

Oh my :eek:
It looks like he uses the stock thrust bearings, which is Okay. But like Philbur points out, I see no way to preload them. At least you can preload the thrust bearings in the stock setup with the nut on the handle. Theres no thread, you have to rely on that lock collar only. This is not going to give you the best results....
A lock collar isnot for preloading. That end should be threaded and have two locknuts put on it so you can preload. Why did CNCfusion not do this when it is so simple. Cost would be raised 2 dollars and performance 200.

Other thing I see is that there is that the entire radial support is relying on the thrust bearings, thrust bearings = thrust load. There are no support bearings up front, the lovejoy just goes right up to the motor shaft. The motor manufacturers I see somewhere usually advise against using the motor bearings as support for the leadscrew.

Needless to say, I am pleased with my decision to make my own hardware ;).

Jl123-
WE are talking about axial play in the entire ballscrew itself, not backlash in the single nut. The ballnut setup is fine.
Also, could you take some pictures of the Y axis too? It would be interesting to see how that one looks.

JL123,

What Phil is saying, is if your not careful and don't get enough pressure (linear axial force) on the bearings you may end up with backlash. Simple example, would be a wheel bearing on a car, there you have a nut to tighten against the wheel bearing to get the slack out (backlash). If the bearing is not sandwiched snuggly and secured between the ball screw, the bearing mount and the locking device, slack (backlash) can occur during heavy cuts and the shaft will move back and forth and introduce linear backlash. We are used to seeing a nut doing the securing, which is easy to adjust the snuggness (preload) across the bearings, and then a drop of blue lock tight or a jam nut, to asure it stays in place.

Ron

Ron

Ahh. Ok now I see what everyone is pointing out about preloading the thrust bearings(I thought everyone was talking about preloading the ballnuts). Actually when I first threw it together without my washer I made there was so much backlash because of all the play of the stock thrust bearings and collar.

Would it be possible to somehow apply pressure on the ballscrew? Maybe a bar clamp and some ingenuity? :)

Think a die would work at threading the ballscrew?

hi

Why not use front wheel bearings of a motor vehicle?

Here in Australia when have rear wheel driven cars still and a lot of the front disc hub bearings utilize a 2 front tapped needle bearings one larger inner one and smaller outer one.

These bearings need a set preload on them and you can adjust the load on them to obtain zero backlash. There cheap and readily available.

just a case of making some mounting brackets ( housing brackets ) to support the ball screws and secure it to mill table .

The motor bearings are carrying a large amount of weight ( 6 bearings in total carrying 1700 kg = 3740 lbs of vehicle ) on them and at approximately speeds 4000 rpm and higher so i don't see any issues using them.

In these type of bearings you have load in all directions

cheers

I wondered about loading the thrust bearings when I looked at that kit and just assumed that there was some way they were doing it that wasn't obvious from the pics.

Did a quick drawing if anyone is wondering what is being talked about here. I think that jl sees whats being said.

And if I got it wrong feel free to smack me upside the back of the head.

Yep I understand fully now. Looks like the Y is going to have the same problem.

So on these lovejoy couplers I'm just relying on a single set screw? How are others secured? Don't tell me a thread or pin.... :rolleyes:

Question for Ya.
If you took just the bearing itself, and held it up against the hole in the cast frame of the machine where the ball screw is coming out, will the bearing pass through?

If not, Then your pre-load is coming from the "sandwiching" of the bearing between the cast frame, the bearing housing and the ball screw shoulder. Your pre-load in this case is fixed. But, adding shim to the bearing you can increase the the pre-load as required. Oddly enough you have some what done this by building that nice washer you put on the ball scew.
If your bearing is allowed to pass through the casting, build a plate that goes on the cast frame first, with a smaller opening for ball screw only to pass through. When you mount the bearing housing back to the machine with the 2 cap screws you will be loading the bearing to the ball screw.
The more shim the more pre-load, don't go crazy with it, or else your bearing will fail pre-mature amd your steppers will be working O/T.
Hope this helps

OK you posted a great pic while i was typing.
If you ad a plate between the bearing and the coupling (drill and tap into the aluminum) you can capture the bearing, and pre-load it that way.

Alright so you're saying to make a plate that goes on after the ballscrew shoulder and sandwiches the bearings against the collar? It'll register in a few minutes. :)

Here is how it should be.

Phil

Ahh. Ok now I see what everyone is pointing out about preloading the thrust bearings(I thought everyone was talking about preloading the ballnuts). Actually when I first threw it together without my washer I made there was so much backlash because of all the play of the stock thrust bearings and collar.

Would it be possible to somehow apply pressure on the ballscrew? Maybe a bar clamp and some ingenuity? :)

Think a die would work at threading the ballscrew?

If you don't have a lathe you have a bit of a problem. What machinery do you have.

Regards :violin:
Phil


Alright so you're saying to make a plate that goes on after the ballscrew shoulder and sandwiches the bearings against the collar? It'll register in a few minutes. :)

Philbur is right, that is one way to achieve pre-load, (most common) but it is also common practise to capture bearings using shim pack, the advantage to a threaded ball screw is that you can aplly the correct torque to achieve pre-load. Using shim stock you would have to traditionally use feeler gauges and measure the gap between the actual roller bearings and the outer race before pre-loading, and again after shimming to ensure the right clearance between roller balls and the outer/inner races of the bearing housing.

Trouble is I don't thing the ballscrew as picture actually has a shoulder to preload against. It needs machining

Phil (chair)

Question for Ya.

If not, Then your pre-load is coming from the "sandwiching" of the bearing between the cast frame, the bearing housing and the ball screw shoulder. Your pre-load in this case is fixed. But, adding shim to the bearing you can increase the the pre-load as required.

You will want a lathe to turn that down a little bit below 12mm, otherwise you cant really run a die over it. Good luck sir.
I'de be emailing CNCfusion with a love letter right about now...

No lathe...:(

Here's a few more pics of the X axis oppisite the drive side(right side). First picture shows the bearing washer against the ballscrew shoulder. Second is the washer I made to go first. Third that's the way everything is assembled.

JL123,

See if you can follow my thought. On your x axis, if you clamp an angle to the far end of the table that extended down with a threaded hole and screwed a bolt through said threaded hole until it contacted the dead end of the ball screw(I thought you could see the dead end of the ball screw in one of your pictures) and then took a piece of angle with a threaded hole and a screw through said second hole and against a socket which would be against the locking coller you would have a clamp that would #1 push the screw against the bearing mount and preload the 1st bearing and the other end of the clamp would independly preload the second bearing against the other side of the bearing mount through the clamping collar. I would give this a try before doing hardware fixes. Don't overtighten the bearings.

Just think about your table being the middle of a giant clamp. Use that new clampling kit you bought.

Ron

Can you give us a clear sideview of the ballscrew showing the shoulder.

Phil :bat:

No lathe...:(

Here's a few more pics of the X axis oppisite the drive side(right side). First picture shows the bearing washer against the ballscrew shoulder. Second is the washer I made to go first. Third that's the way everything is assembled.

FPV- You should be able to use those bearings without issue. You could even put an assembly on both ends of the screw, tensioning. THis is the ideal situation

Phantomcow2,

Could you follow my method for applying preload to the x-axis bearing? I feel like the clamping collor would hold even if you had to use a drop blue locktight on the shaft. I don't know if you can get to both ends of the y axis though.

Ron

Thrust bearings are not good for an offset (belt) drive arangement. Direct drive may be OK but the motor bearings will see extra radial load. All depends on how hard you intend to run it.

Phil

FPV- You should be able to use those bearings without issue. You could even put an assembly on both ends of the screw, tensioning. THis is the ideal situation

But the taper roller bearings are made for radial and axial loads, this should be okay.
Has anybody asked for pictures of the Z axis of the kit? If not, I will be the first :).

Video of ballscrew shoulder (http://home.comcast.net/~joel.pcnerd/MVI_0520.AVI) Right click save as.

It's got a high spot and low spot. The low spot doesnt give very much for anything to grab on to.

Few pics as well. I haven't installed anything for the Z but it's going to have the same exact problem. It uses the same collar but this time I'll have the enjoyment of trying to mount a pully on the ballscrew with 2 set screws.

Based on your latest photos I think without a lathe you have a bit of a pig in a poke. Don't really know where to start. CNCfusion appear to have left you high and dry :boxing: . You might get by with some ingenuity but it largely depends on what metalworking equipment you have access to.

Never say never.

Phil :cheers:

Video of ballscrew shoulder (http://home.comcast.net/~joel.pcnerd/MVI_0520.AVI) Right click save as.

It's got a high spot and low spot. The low spot doesnt give very much for anything to grab on to.

Few pics as well. I haven't installed anything for the Z but it's going to have the same exact problem. It uses the same collar but this time I'll have the enjoyment of trying to mount a pully on the ballscrew with 2 set screws.

Well they can't leave me high and dry as I'm in dallas,tx right now. They are located in houston,tx. I go to school 45 minutes away from houston as well. ;)

I'm working on Ron's suggestion and will upload more pics to see if this is what he's talking about.

Ok after seeing your vid. you will definitely need a bigger shoulder there, take it to local machine shop, and turn it down to a standard 12mm.
Replace bearings with radial ball bearings with 12mm I.D. and what ever the housing Diameter is that the bearing is held in will be o.d. of new bearing.
cncfusion should have provided a better shoulder than that!
Then carry on using shoulder and bearing housing (with shim if needed) to set your pre-load.

It's actually turned down to 11.92mm.... sigh

Problem is they should have used a bigger ballscrew right?

JL123,

See if you can follow my thought. On your x axis, if you clamp an angle to the far end of the table that extended down with a threaded hole and screwed a bolt through said threaded hole until it contacted the dead end of the ball screw(I thought you could see the dead end of the ball screw in one of your pictures) and then took a piece of angle with a threaded hole and a screw through said second hole and against a socket which would be against the locking coller you would have a clamp that would #1 push the screw against the bearing mount and preload the 1st bearing and the other end of the clamp would independly preload the second bearing against the other side of the bearing mount through the clamping collar. I would give this a try before doing hardware fixes. Don't overtighten the bearings.

Just think about your table being the middle of a giant clamp. Use that new clampling kit you bought.

Ron

I think I might get it. Here's some pics with a rough setup that I think you're trying to explain. Where the angle will have a threaded hole for a bold to apply pressure on the ballscrew with the bearing housing on the oppisite side bolted on which preloads the first bearing. What about the second bearing? I understand that I can't just put linear pressure on the ballscrew but instead have to turn it at the same time, correct?

ahhh...i just realized how I could preload the second bearing. If I shim the bearing housing and tighten it down then move the collar in, secure it, then remove everything and do what ron said to load the first bearing. I'm thinking that might work. Hmm. Headache.

Thanks for everyones effort. My gift to the community is being the guinea pig. lol. :)

trubleshtr, you wondered about the bearing housing so here are some pics of it as well.

what if you go down to 10mm I.D. bearings and turn shaft to suite.?
might give you enough shoulder to load against.
Here is a rough concept. Sorry for quality.....

As you increase shim size, you load the bearing even more against housing and frame and shoulder on screw. You kinda have this already, but, that shoulder you have looks too small to handle a heavy thrust load.

Alright I think I loaded the bearings and there was still backlash but I'm thinking this backlash has to do with the ballnut. Since it's only one I'm getting around .011" backlash on the table. (X axis)

I ended up just taking the ballscrew out and taking the ballnut off as well. Here's a picture of how I did it. The bearing housing is now tougher to turn. Theres probably too much pressure on the bearings but at least this shows that my technique works...I think. Let me know what you guys think.

I hope the balls didn't go everywhere when you removed the ball nuts! :-o

Yes with the pipe clamp like that is what I was saying if you can push the assembly together with enough force before locking the collar, it might be OK. However, there is no guarantee that the collar will not slip at some point down the road, ie if you are making a heavy cut.

So this CNCFusion is in Houston eh? I am in Houston... hmmmmm.

The balls didn't go everywhere when I removed the nut. I had the cardboard tube to hold against the ballscrew when sliding the nut out. The grease helped keep them together also. :)

Concerning the collar sliding would it be a good idea to vise it up and drill a hole on the side and tap it for a set screw? Maybe do this twice?

0.011" is a massive amount of backlash. Are you using the thrust bearings I saw in a previous photo. This type are difficult to preload as there is no "give" in the direction of the preload. A pair of angular contact bearings are best, mounted in opposing directions. However any bearings that can take both axial and radial forces would be an improvement, as a pair.

Phil

Alright I think I loaded the bearings and there was still backlash but I'm thinking this backlash has to do with the ballnut. Since it's only one I'm getting around .011" backlash on the table. (X axis)

I ended up just taking the ballscrew out and taking the ballnut off as well. Here's a picture of how I did it. The bearing housing is now tougher to turn. Theres probably too much pressure on the bearings but at least this shows that my technique works...I think. Let me know what you guys think.

Does everyone honestly believe that this backlash is due to the bearings not being loaded?

I have the pdf for the ballscrew/nut and it says the tolerance is .004" per foot. In 2.5 ft that would be around 10 thousands off. Am I drawing a wrong conclusion?

Yes I think you are drawing a wrong conclusion. The tolerance of 0.004" per foot is nothing to do with backlash its about the precision of manufacture of the screw. Backlash is not measured in inches per foot, it's measured at a single point on the screw, when changing the direction of motion.

Phil

Does everyone honestly believe that this backlash is due to the bearings not being loaded?

I have the pdf for the ballscrew/nut and it says the tolerance is .004" per foot. In 2.5 ft that would be around 10 thousands off. Am I drawing a wrong conclusion?

eleven thou is huge this is definitly not right

What I'm trying to get to is that these rolled ballscrews and bearings are not 100% backlash free correct? What would be expected from something like this? In other words I cannot blame all of the backlash on just the thrust bearing assembly. Then again I'm only a beginner so I don't know.

Ball screw/nut spec sheet (http://www.nookindustries.com/pdf/Nook_102.pdf)


:idea: Errr...wait a minute. I hope everyone realizes that the stepper side has nothing attached to the coupler. It's just out there floating in the air. I should probably measure the backlash with the stepper attached no? :) My simple little mind can't comprehend whether or not that'd make a difference or not. I'd have to try it out tomorrow.

Haha...I just realized that little video of the ballscrew shoulder has sound. That's the TV. :D

Having the stepper connected will not change the ball-nut or mounting bearing backlash. It's irrelevant.

It should be quite possible to achieve zero backlash in the ball nut and mounting bearings. However there will always ben a degree of lost motion due to flexing of the various components under load. This flexing will look like backlash in the fully assembled system.

Fully assembled 0.001" to 0.002" apparent backlash should be possible, depending on how heavily loaded the system is.

Phil


What I'm trying to get to is that these rolled ballscrews and bearings are not 100% backlash free correct? What would be expected from something like this? In other words I cannot blame all of the backlash on just the thrust bearing assembly. Then again I'm only a beginner so I don't know.

Ball screw/nut spec sheet (http://www.nookindustries.com/pdf/Nook_102.pdf)


:idea: Errr...wait a minute. I hope everyone realizes that the stepper side has nothing attached to the coupler. It's just out there floating in the air. I should probably measure the backlash with the stepper attached no? :) My simple little mind can't comprehend whether or not that'd make a difference or not. I'd have to try it out tomorrow.

Haha...I just realized that little video of the ballscrew shoulder has sound. That's the TV. :D

JL123,

Your getting on the right track, and making progress. I know that I had .005 backlash with the z axis on my lathe project. I have a shopmaster, with a single ballnut, and it had about .015 (really bad) and all that I could isolate it down too was the ball nut.

Let me ask, how are you measuring the backlash, at this point you are not connected with the controller and steppers, so I assume that you are manually pushing against the table, and if so maybe the screw is turning slightly. Just a thought.

You are still planning on using the second ball nut and preloading the axis. I assuming that the second nut is not installed and it is not preloaded. When you preload the second nut on the axis, it will remove any backlash that is associated with the ball nut.

If you are not using a preloaded ballnut then you will measure the backlash and plug that value into the backlash compensation table of MACH 2 (or mach 3). But at this point, do try to isolate the backlash and if you feel that the bearing is snug. and the ballnut is securely mounted in the ball screw mount and the ball screw mount is securely mounted on the table, the the back lash is due to the nut or is a combination of nut backlast and measurement technique error.

You know because the effeciency of a ball screw if you push against a table, the axis that the force is applied will cause the ball screw to rotate. (As Phil has mentioned concerning having hand wheels on cnc machines and trying to use them during manualy operations)

Let me know about your technique of measuring, and double check the mounts for tightness. You are making good progress.

Ron

JL123,

Another thread I have been keeping an eye on.

This thread is located at:
http://www.cnczone.com/forums/showthread.php?t=22708&goto=newpost

Here is the message that has just been posted:
***************
Nope havent messed with the back lash settings yet. Ive got .006 in the z (quill) .012 in y (in and out) and 10 in the Z (left and right)
***************

Basically this is a shopmaster and it was purchased with the cnc already done, and this is what the machine measures from the factory with single nook ball nuts.

It's more than what we like, but this is straight from the factory.

This gives you another data point. Most of the time we expect .004 to the .005 area for a single ballnut.

Ron

My $.02:

$.01
First in defense of cncfusion, per the Climax metal Products website, a 12mm one piece clamp type shaft collar will withstand @ 780lbs STATIC axial thrust. This amount is derived from the seating torque of the clamping screw. I calculated this number with a linear equation from the Imperial data available. (see attached .pdf's) Not sure what happens to the 780lbs once the shaft starts rotating though!!

$.01
Proposed solution:
First of all take the stock thrust bearings and place them in the closest trash receptacle!! Get a set of American Made 12mm ID grooved race thrust bearings from McMaster Carr (#6681K12) and while you are at it get some 12mm ID bellville washers to pre-load the bearings with. The McMaster bearings will have an actual 12mm ID.
If you do not plan to do any heavy milling in steel, than you can probably get away with the clamp collar method. If not than thread the end of the ball screws to provide pre-load on the bearings. Cncfusion has done the hard job of machining thru the case hardening, you can turn and thread what is left there like butter. If you keep the clamp collars, just get a piece of pipe that fits over the ball screw end, and apply as much axial pressure to the clamp collar against the bearings as you can, then have someone tighten the clamp collar. Hell add a second clamp collar against the first one!!

I am attaching a pic of my X2 ball screw / nut YOKE for an idea of how the ball screws should be machined.

The devil always lies in the details!!

Ken
KDN Tool

KDN, Getting properly fitted bearings and washers is something I'm looking at. I'm probably just going to go ahead and order those but I have a small concern.

Without any bearings or washers and only the stock bearing housing on I can feel axial play in the ballscrew. I am being careful not to accidently mistake radial play for axial play. I can feel the ballscrew give back and forth.

Yesterday when I loaded the bearings with the bar clamp I had a socket over the collar pushing everything against the ballscrew shoulder. I tightened it up still in the clamp and under pressure. I just doubt all of the backlash I'm getting right now is because of the bearings. Don't get me wrong I want to upgrade the current bearings, washers, and whatever else I need but something just doesnt feel right with the ballnut.

I'll keep at it all day trying out any suggestions you guys have. I think I may just place that mcmaster order right now.

Joel

How tight of a fit is the washer you made to the bearing journal of the ball screw? It should be a almost a line fit. Also look at the shoulder where the journal meets the ball grooves. Is the transition nice and defined with a sharp corner or is thier a radius there from the lathe tool bit used to turn it down? If there is a radius, then you will need an appropriate chamfer on the washer you made. By "appropriate" I mean just enough to allow the two faces to become coincident. Too much chamfer and its back to sqaure one.

Ken

Using a telescoping guage /w caliper the ID of the washer is

.474"

OD of the journal

.469"

I see what you're getting to. 2 pages back there are some pictures and a video of the shoulder. It's defined but gets thin as the ball thread ends.

I drilled out that washer with a step bit and totally lucked out. :)

Video of play in ballnut (http://home.comcast.net/~joel.pcnerd/MVI_0540.AVI) Right click save as 6.5MB bigger than first video, wanted to show more detail

Hopefully it shows what I'm talking about. There was no radial movement. You should be able to see that. There are no bearings or collars installed. The bearing housing is just there to help me keep the ballscrew straight with the ballnut.




I know when the thrust bearings are loaded properly that play will not exsist but the play in the ballscrew will always be there. That translates into the rotational play at the coupler and collar. And I think that's where my backlash is coming from.

That may be too much clearance between the washer ID and the journal OD. It may be allowing the washer to tilt to the side that does not have much contact face.(see blue face in attached .jpg) You really need a washer with .0005" or less clearance. (preferably .0001-.0002"!!)

Let me backstep a bit. You pre-loaded the bearings with the pipe clamp, tightened the clamp collar, removed the pipe clamp, and then felt axial give of the ball screw alone in the bearing housing???

Ken

That may be too much clearance between the washer ID and the journal OD. It may be allowing the washer to tilt to the side that does not have much contact face.(see blue face in attached .jpg) You really need a washer with .0005" or less clearance. (preferably .0001-.0002"!!)

Let me backstep a bit. You pre-loaded the bearings with the pipe clamp, tightened the clamp collar, removed the pipe clamp, and then felt axial give of the ball screw alone in the bearing housing???

Ken

Yeah there was no give after preloading the bearings. Everything was tight enough it became harder to turn. Like I said I think it was too tight but at least I know it worked.

In that case you need a spring washer (bellville washer) in there to provide a constant spring force pre-load on the bearings while still maintaining rotational smoothness. Look at McMaster #94065K42 ($4 for 10pcs) You can stack them for increased spring force if necessary.

My experience with the Nook ball nut backlash has been in the .003-.006" range. Order some .125" balls from McMaster, re-load the nuts, and you will be in the .002-.003" range. McMaster #9528K11 = grade 25 chrome steel balls 1/8" @ $3 for 100.

Ken

Based on the vedeo the scew is moving and the housing is not so what you see must be backlash in the mounting bearings. This of course doen't mean you don't also have backlash in the ballscrew. Edit: (ooops should have read it all first, as usual)

Listen the KDFN Tool. Seems like he's been there before and has specific experience of the same issues.

Phil


Using a telescoping guage /w caliper the ID of the washer is

.474"