Hey Carve_one, Graham and all.....
the solsylva acme upgrade is basically done, and producing parts!! Have run for a couple of days at X=60IPM and y=65IPM to check for loose parts etc.... Finally found all my strange intermittent problems where do to a weird issue with the puter I was running Mach with. Didn't realize that was the problem til I talked to Art.. The issues almost went away with the Acme upgrade, but were still there very intermitently. No probs at all since I changed out the computer.
Now that it seems happy for a couple of days, I decided to see what it could actually do now. Just did a few air cuts of the contract job I'm doing, after bumping up the max speeds a lot. I set the Y up to ~100IPM, and the X up to 90IPM, accel at 10, and all jogs semmed happy, much to my surprise. It ran almost half way through the job before the Y coughed... WOW!!! set them both to 90 and it ran all the way through.. Gonna back off to 80 to actually run the job some more. Still a little gun shy after my weird problems...
These are the following changes from the plans after a year of use....
1) Xylotex board with 269oz steppers, 24" x 48" cutting area.
2) NEW! 1/2-10 acme lead screws from Enco on all axis, no turned ends, using all 0.5" ID hardware, including bottom Z bearing added.
3) NEW! Steppers on Y and Z coupled with short belts, not direct, with 15T pulleys.
4) NEW! DumpsterCNC AntiBacklash nuts.
5) NEW! Doubled up both of the 1x4 end plates to limit vibration, especially on stepper end.
6) 1.5" x 2.5" Steel tubing for sides and gantry. (sure glad I did that, and to think I only used those because they were free LOL)
7) Added dampers, since it really made the slower cutting speeds sound much better.
Now to slowly change the rest of the wood to metal...
Didn't go the multistart to keep upgrade price down. Didn't go ACME originally went 1/2-13 allthread, which was a disaster, since I didn't know about Enco 1/2"-10 prices, actually cheaper than allthread. LOL. I'm trying to forget the allthread trouble shooting time, cutting it out of my life.....
All that said, Does anybody out there have an extra foot of 1/2"-10 5 start, or 1/2"-8 4 start they didn't use from a longer piece? want to change the 1/2-10 on the Z axis to this to help with the moving the weight up and down issue. Some contract stuff I'm cutting has about 605 up and downs in each cut, and the Z is the weak link from a speed standpoint right now with the 1/2-10 single start.
Or else does any know anywhere on the web that will sell a 12" piece? everywhere seems to have a $50-$60 minimum order, and I only need 12", about $9 worth...
More on the upgrade and some pics maybe tomorrow, and will upload the gcode for part of the contract job, since it is a VERY good test of the machine. 55 sets of 11 holes, 605 in all, on a 2'x4' sheet of luan..

Don
www.eurekaaircraft.com (http://www.eurekaaircraft.com)

Im building a machine very similar to yours. Same screws, but 400 oz in motors w/xylotex controller. 4' x 6' table

i counted on problems with moving z axis up & down, my question to you is how do you expect to move the z axis with multiple start screw? Faster?? Don't the step motors have enough trouble just moving all that weight?

My plan's to use 400oz in steppers on the z axis with 1/2-10 screws. Would you suggest sticking to this plan?

Try Widgetmasters ebay site I know he has 6" to 18" acme screws. Go to

www.widgetmaster.com and click on his Ebay link.

Vince

Justcncit....

Going to a lower pitch thread gets more movement per rev of the stepper, so more speed will be achieved since you will get more speed at the rev that works now, ie at the same torque point of the stepper. Same reason you get more speed from 1/2-10 than you do from 5/16-18. Steppers lose torque as the speed up, so you try to get more movement at the same RPM. The stepper doesn'y know you changed the lead screw. And yes before anybody screams, it won't be the theoretical increase of 5x the speed, going from 10TPI to 2 TPI, due to intrinsic losses, but it should get me to where I want to be. I don't know if I explained this right, but the jist is right. Talked to David Steele, he has tried it with 3/8"-10 2 start, and it definitely helped. Time will tell.

I though about making some kind of counter-weight, similar to the way they do the harbor freight/grizzly mill upgrades, but the Y axis moving opens up a real can of worms doing that, plus the space issue.

Don
www.eurekaaircraft.com

vconney, tried the link and there isn't a page there now. Do you have any links to his ebay stuff, or his exact ebay name?? thanks

rc_flyer,

I have an extra 3' length of 1/2-10 5 start. It's slightly bent, which is why it's extra. However, I'm guessing there's probably a foot of it that's straight. I'd be happy to send it to you for free so long as you pay shipping. I'm in a good mood since I got my machine moving tonight :).

Oh, and on the subject of getting higher speeds -- higher pitch ACME has significantly better efficiency (sometimes up to 70%, compared to only 20%) than single start 1/2-10. This means that you basically get your extra speed for free! Well, not quite -- you're still trading off resolution. You're also right in saying that it's good because it keeps the stepper turning slower, where it has more torque. Case in point, with the same motors, I went from 200 IPM on single start to 700 IPM (!) on 5 start.

Anyway, PM me your address or send me a note through my website and I'll hook you up with the ACME rod.

Ahren
www.cncrouterparts.com

Here are pictures of the lead nut mounts. They mount to 1" alum angle that replaces the 3/4" used with the smaller rod, to give room for the 1/2" rod. No pictures of those before mounting , but they are simply wider versions of the original, same holes and slots. Same mounting procedure. I tried to keep this simple, just slightly bigger.

The DumpsterCNC AntiBacklash nuts are used through out. They are very high quality pieces. I ordered the round flanges since I wasn't sure how to mount them till they arrived.

The mounts are made from 2" angle, about an inch wide, so they would be long enough for the screws to reach both slots, as before. The holes are 5/8". One end is cut down to match the shape of the AB nuts flange, as shown. The AB nuts flange is sanded to fit the mounts after mounting them. I used 8-32 bolts, with holes in the mounts tapped. The holes for the adjustment bolts are tapped in the mounts as before. All adjustments are as before.

The Z mount is trimmed down much further to get it to fit in the very tight space. It is barely wider than the barrel of the nut. I moved the Z axis trucks out closer to the edge of the box to get more room behind.

Other info: I used .5" ID- 1.125" OD bearings, (R8ZZ from Ebay) so new bearing mounts had to be made. I also used .5" ID pulleys from SDP and ordered Acme nuts from Enco along with the rods, so no turning of ends was needed. The new rods fit the bearings perfectly, unlike the old allthread, so no clicking.

Pics of the Y and Z stepper belt mounts, the new bearings in place, and other mods to come, as I get a chance to put them up.

If you want to see anything in particular, just let me know.

Don
www.eurekaaircraft.com

Don,

Your upgrades are looking really good. Thanks for sharing the information and I'll follow your lead as cash is available.

CarveOne

I think alot of people are following this lead. Alot of people are using .5" shafting & dumpsters nuts. The only thing yet to catch on is Ahren's bearing blocks.

justCNCit,

this is not intended to be a rant, but more of an explanation as to why I didn't buy them myself.

I looked at Ahren's site a while back and determined that the bearing blocks shown there are fine for the builders who don't have a mill or where a standardized block shape will work ok, as with the 8020 extrusion machines. The price is very reasonable for the amount of work involved in making them. I didn't see any note that Ahren will make custom shaped blocks for other machine designs.

Had there been a set available for the plans I'm building from then it would have added more than $100 plus shipping to my build cost. My bearing blocks need to be differently shaped, and I can mill them myself from the aluminum stock I already have on hand, so I didn't order them. I think that any low sales is not a quality or design issue as much as there will be a more limited market by not having available a small selection of kits for the more popular machine designs in addition to a standardardized bearing block for 8020 machines.

CarveOne

Hi guys,
I'd be interested to know if there are other "standard" designs for bearing blocks or motor mounts you all would be interested in. I made a run of the ones I have because they worked so well for my machine, but I realize they won't work for everyone. The setup time for custom blocks makes the pricing somewhat prohibitive, but I'm open to changing the design to make it more universally appealing.

I'm planning on releasing plans at some point for a complete 8020 machine, but I have a few details left to work out on the linear motion system (trying to keep costs down). Anyway, I don't want to hijack this thread, so feel free to PM me with ideas that you'd like to see become metal.

Here are some more pictures of the upgrade. Instead of going into a long explaination of each picture, just ask for details on any pic.



Don
www.eurekaaircraft.com (http://www.eurekaaircraft.com)

Here are some more pictures of the upgrade page # 2. Instead of going into a long explaination of each picture, just ask for details on any pic.



Don
www.eurekaaircraft.com (http://www.eurekaaircraft.com/)

Wow!! Don, did you really carve that -very- realistic looking cat?? The precision is just amazing. Details please....

How are the x axis acme rods running now? Any whipping problems? Are they running straight and is the belt tension on the pulleys causing any bowing of the rods?

CarveOne

yeah, he is the only thing that seems to get bigger the more I cut...LOL

The first thing I noticed, even with the 1/2" allthread from hell, is that the 2 bearings on each end of the X had a much more drastic effect on controlling the whip. They never seemed to help a lot on the thinner rod. There is very little whip on the x leadscrews at all any more. It varies slightly at different higher rpm's, like of a resonance thing. At most speeds it completely dissapears, and gets just a little at some random speeds, less than a 1/16" deviation at worst. Any whip that is there doesn't seem to effect top speed at all. With over 5' of rod, it will never go away completely.

I have the X belt tighter than I ever could have it with the smaller rods. There doesn't appear to be any bending cause by it. I think as I do more fine adjustments, the speed will continue to incease a bit.

The main thing I have noticed is that since the rods fit the bearings perfect now, it is critical that they are at exacly 90 degrees to the rod. There was a slight margin for error with the smaller, loose fitting rods.

Don

When you are cutting some 1/4" birch plywood, and making 90 degree turns (or less) please check the x axis rods to see if they tend to take a momentary hit as the router bit rapidly changes direction. Hopefully, the long belt minimizes any jolts to the rods other than disastrous slams into the stops.

Which brings up the point of limit switches and zero reference switches. Do you have any installed on your machine?

I would like to see a photo of a sample 1/4" birch plywood wing rib if you don't mind, as we use them in our heavy lifting RPV models. These things have been stress tested at ~70 lbs lift off weight. My cnc router is being built to cut these 20" chord wing parts and other fuselage parts. The root ribs are close to that length in some of your kits. I would like to show the picture to my boss for cut quality purposes.

CarveOne

Carve..
I guess I don't understand your question. When Mach is operating in constant velocity mode, which is how I always run it, the motors don't come to a screeching halt, then change directions. They do what I guess you wound call 'pivot' around the corners, making a rounded path, but it leaves the point just that, a sharp point. It does some kind of look ahead. I don't know how it does it, but the cutter never slams around inside or outside corners. Maybe someone with more experience can explain why that isn't a problem. (or Art is just a freaking progamming god, and it has to do with Karma :)

If that isn't what you are talking about, email me with a better explanation of the question, I'll see if I can figure it out...

As far as limit switches, I have 3 limit/home switches doing double duty at X & Y zero, and Z top. I don't have them at the other extremes, and have never regretted it so far. I'm sure they will get there eventually. At the 0 extremes, I want them in case the machine does lose steps, on the way back it could go past 0, thinking it is only going to 0. I have had that happen before, in the past, although it has been a while. I don't see it happening going the other way, since I don't think I have ever run a program that goes that close to the ends..

Don
www.eurekaaircraft.com

Thanks Don,

You answered my question well enough.

I haven't seen one of these Solsylva machines running in person, but I was told by a local hobbyist cnc laser cutter builder that this would be a problem with the 1/2" acme rods same as with the smaller rods. I was skeptical of this statement and wanted to know if you had seen this. Watching ShopBot and other videos on YouTube sure didn't show anything like this but they use rack and pinion drives.

CarveOne

Hey Carve.. Hope the pictures I sent helped.

I guess I don't know where he's coming from. I would be very skeptical. I have never noticed that on either size rod. I don't think it will matter which you have, rack, ball, lead screw... A sudden change in direction will jerk and shake the table if not controlled. It's called inertia, and no control system in itself or table design will change that, without effecting accuracy. I would venture to guess that is why the CV mode is in Mach, and probably most other software, so the software can plan ahead and control velocities as needed, to help limit the problems caused by inertia.

I wonder what sofware he runs the laser with, and if they don't have the Constant Velocity mode. Maybe you should turn him on to Mach, I think it has special controls for the laser. :)

Don

Hey RCflyer,

I misspelled the link...this is the right one.


http://widgitmaster.com/

Yes Don, the pics look like the kind of cuts we are looking for. A little fuzz on some edges is expected. You should see the fuzz on the parts we cut by hand on the little Dremel scroll saw. I think I would rather listen to a 2hp router than that buzzing little demon. Nah, not really. :D

I'll ask him about the Mach3 and constant velocity next time I talk to him.

CarveOne

why would you even need limit/reference switches, the moment the stepper loses steps your work is scrap is it not

JustCNCit,

I don't think you get the concept of limit & home switches. It took me a while to use them effectively. I understood limit switches, but until I figured out fixtures, home switches didn't seem worth the effort. One of the Mach training videos explains fixtures very well.

Limit switched are for damage control. If for any reason the steppers try to move the ganrty past the end on the table, which would cause damage, the limit switches stop the movement before damage can be done.

This can be caused by a wide range of factors, from defective g-code, controler sending the machine too far for ANY reason, the machine drops steps, or the bit hangs for some reason, returning to what the controller thinks is 0 will actually bring the gantry PAST 0...

Lets say the machine errors for any reason and it looses 1" of travel on the X axis, so it is at physical location 34", but Mach thinks it is at 35", to drastic to be realistic, but moe effective to explain than .002". You click on 'return to Zero.' Mach sends the machine -35", since that is where it thinks it is. Unfortunately, moving it -35" sends it crashing into the end of the machine, causing a LOT of damage. Luckily, the limit switch stopped it after 34"....

Maybe someone else can explain this better, in fact I know some else can explain this better...LOL

Home switches allow you to start the machine from the SAME POINT every time you run a part, even after turning off the machine, or making manual adjustment on one of the axis' while the machine it off. They allow you to have fixtures if you cut the same job repeatedly. I use fixture points for 4 different jobs I run.

Also, more often that not, at least in my experience, the machine looses steps or binds (for any of a variety of reasons), on rapids. (usually trying to push the machine speed to close to the edge ;).... By returning the machine to it's known 0 points, the part can be cut from the beginning retracing it's steps, or started from a location just before it errored.. No bad part, everybody is happy...

Don

I figured somebody might post that. I am fully aware of their purpose but at this time in the design process am too cheap to plan for home/limit switches. There are no plans to run the machine at the limit of its ability yet.

Good home/limit switches cost money to offer decent repeatability. (be precise)

As far as I'm concerned, because losing a step would mean losing 1/10 of an inch(wrecking the part), if it does that limit/reference switches aren't going to save the part. And if the machine does go off limits, it won't be able to push any farther than the mechanical limits anyways, it's not like this is a metal cutting machine.

Also if it does lose steps and I catch it, and want to stop the machine and restart the program, all I need is a good datum surface on the part to touch on.

rc_flyer,

I can appreciate that the home reference switches need to have very good repeatability for the accuracy needed at what is considered to be 0/0, but do the extreme limit switches that stop the machine from frying steppers and drive board(s) in addition to bending mechanical parts need to be anything other than low cost and just reliably operate when needed?

CarveOne

The way I see it, if the work can somehow jam the router bit & cause the steppers to lose a step, then they aren't all that strong. Remember we're cutting wood/plastic. If the machine hits a mechanical stop, I don't think theres any way the stepper could pull it past the stop.

Besides, most work doesn't even cover most of the table. IF it does lose steps, chances are that unless it is constantly losing steps in the same direction that all it will do is screw up the work.

Of course I am not speaking from experience only knowledge.

I was also unaware you could fry a driver board by stalling a stepper. Is this true? how about xylotex models.

Stalling the stepper won't hurt the controller as the stepper motor is effectively stalled every time its holding position and not rotating. The reason is because the controller is already limiting the current to the motor all the time (moving vs holding).

With an ac or dc motor (regular type motors) the current rises as the motor slows and since the controller isn't typically controlling current things tend to get hot and fry. I think most people have experience with normal motors and tend to equate stalled with fried.

I need to read up on stepper motor theory and controllers 'cause at present my ignorance knows no bounds on this subject. :drowning: So to my dumb comments - boo and hiss as appropriate. :)

I made the assumption that Fifty_ohm mentioned. I'm concerned about what happens when the mechanical limit(s) are reached and would like to prevent any potential mechanical or electrical damage wherever possible. These are low power steppers (mine being 425 oz/in Xylotex) and they have less torque than a typical 9.6v cordless drill when the stepper is running at maximum torque. After overcoming the friction losses in the mechanics there is probably not much power left to bend or break anything during rapids. It probably just sits there missing a lot of steps.

If no electrical damage results from running into an unmoveable object then that's great.

CarveOne

Im sorry, my posts can be a bit rough towards you guys. I didnt mean to be so forward. A regular CNC machine really should have these switches installed but since mine is going to be cheap I was going to try avoid them until it is tested out and running.

I made the assumption that Fifty_ohm mentioned. I'm concerned about what happens when the mechanical limit(s) are reached and would like to prevent any potential mechanical or electrical damage wherever possible. These are low power steppers (mine being 425 oz/in Xylotex) and they have less torque than a typical 9.6v cordless drill when the stepper is running at maximum torque. After overcoming the friction losses in the mechanics there is probably not much power left to bend or break anything during rapids. It probably just sits there missing a lot of steps.

If no electrical damage results from running into an unmoveable object then that's great.

CarveOne

Crashing your machine and stalling the steppers won't hurt the steppers, or the drives. I've actually heard of people homing their machines against hard stops without any home switches.

When Mach is operating in constant velocity mode, which is how I always run it, the motors don't come to a screeching halt, then change directions. They do what I guess you wound call 'pivot' around the corners, making a rounded path, but it leaves the point just that, a sharp point. It does some kind of look ahead. I don't know how it does it, but the cutter never slams around inside or outside corners. Maybe someone with more experience can explain why that isn't a problem.

Two possible answers.

1) You're using cutter comp, G41/G42, which will automatically give you the rounded corner toolpath you're talking about.

2) Your CAM software is creating the offset toolpath with the rounded corners. You can do this yourself in your CAD program. Ususally, most people start with the part outline and then offset it by the tool radius. On all outside corners, you can add a radius equal to the tool radius, and you'll get square corners.


Machs CV mode isn't whats creating your round cornered toopaths. Without them, CV would be rounding your corners.

ANother option is to try running the beta version of Machs succesor, Quantum. It's only difference at this point is S-Curve accel and decel, which should give you much smoother direction changes. It's recently been updated to match the latest vesrions of Mach3, so there shouldn't be much difference. Both can co exist on the same PC.

r

Oh, and on the subject of getting higher speeds -- higher pitch ACME has significantly better efficiency (sometimes up to 70%, compared to only 20%) than single start 1/2-10. This means that you basically get your extra speed for free! Well, not quite -- you're still trading off resolution.

Not quite free. A single start screw has 5 times the mechanical advantage of a 5 start screw, so at a given torque, the single start screw can deliver 5 times more torque. Factor in efficiency and it's actually about 3 times.

If you download the acme catalog at www.nookind.com it will tell you how much torque is required to move 1 lb with all of their screws. A little math and a torque curve chart will tell you how much force you can deliver at a given speed with the motor and screw of your choice.

JustCNCit... no prob, I just chalk it up to impersonal computer words, instead of human contact. These posts rarely convey what they really meant...

From personal experience, during testing is when you NEED limit switches.. You tend to make a lot of mistakes during the learning curve. Bumpimg axis the wrong way when moving, hitting the left arrow when you meat the up arrow, a minus sign added or missed, etc. Don't ask how I know... The home is important once you get going, for repeatability.

Some one on the DIY CNC group asked for offset for the holes for the lead screws when using 1/2" rod.. I went 5/8" over from the X rail support, and 1" down from the top. As shown by the pictures I put up, that far isn't needed, but the extra clearance can't hurt.

You might be able to get by with the original dimensions is you trim the AB nut enough, but it will be tight. I know i went down further than needed. I plugged the original holes with dowels and started over.

BTW the upgraded router was very happilly cutting 1/8" lauan at 55IPM, with x rapids at 70IPM, and Y rapids at 80IPM, with plenty of headroom. Should get the 5 start rod for the Z tomorrow. Can't wait...

And everything they say about feeds and speeds is right!!! Go Figure ;) Now that I can run at faster speeds, the cuts are cleaner, and the bits are lasting longer, with much less heat coming off of the cutter during cutting. You can feel the difference during cutting, since I don't have a variable speed router mounted. That will be the next investment.

Don
www.eurekaaircraft.com

Cave_one and all..

I may be wrong, but these machines are stonger than you give them credit for. If they can move a cutter through wood with no problems, at least usually.. you don't want to know what happens if a carbide blade comes off one side of a bit when it hit a very hard knot. I'm sure it was a defective bit, but accidents do happen :(

They can also bend or break parts, a lot of which are wood or mounted to wood, the same thing you are cutting, when they collide with a 'mechanical stop' as it is called. No different than a bit getting stuck because of too aggressive cutting, a bit tip breaking, or what happened above. Something has to give, and don't rely on the steppers to stall before something is damaged. If they stalled that easy, you wouldn't get any work done.. JMHO...

Don
www.eurekaaircraft.com

Ger,
You're right -- you don't get your extra speed exactly for free with high pitch screws. After looking up the relative efficiencies of a 1/2-10 vs. 1/2-2 (the difference RC Flyer is seeing), it appears the difference is 40% vs 75%. So you have roughly a 2:1 improvement, but this is tempered by the 5:1 mechanical advantage you lose.

However, the fact that steppers have more torque at lower speeds means that so long as you still have sufficient torque to move and cut with your machine, much more of your stepper motor power will be going into moving things than into friction. Additionally, you'll have mcuh less trouble with whip, since the screw doesn't need to turn as fast to get the same linear speed. From practical experience, I can say that high pitch ACME was a significant improvement for my machine.

Ahren

I forgot you guys are making machines out of wood. My personal preference though, is metal.

Just got the 1/2"-10 5 start acme rod (Thanks, Ahren!!) and found a 10" straight section. Installed it to replace the 1/2"-10 acme on my Z axis, which was only a 30min job, all I had to do is trim the DumpsterCNC AB nut to fit my bracket, and put it back together.

The rapids on the Z were previously about 40IPM do to the added vertical weight of the router, 1/2 of the other 2 axis, plenty of speed unless you are doing a job with 605 vertical moves, and need to do a LOT of them.

With the new rod, I am getting a conservatively set 120 IPM on the Z. I tested it higher with no problems, but sim tests in Mach showed negligible production time decrease above 100IPM, so I didn't set it ant higher.

My cut time for the job went down from 38 minutes, to about 27 minutes, with that change. Considering the job, and the next jobs from the same client take about 45 runs, that is a big difference, very worth the upgrade.

In the probably not to distant future, it looks like these rods will be on the other 2 axis as well... That will cut the production time in more than half again, but have to catch up on backlog cutting before I do it.

Don
www.eurekaaircraft.com (http://www.eurekaaircraft.com)

Don,
Glad the ACME is working out for you. It's amazing what a difference the 5 start can make.

Ahren
www.cncrouterparts.com
NEMA 23 / 34 motor mounts and end bearings for 1/2" ACME