PDA

View Full Version : It's my turn to be the noob



jmytyk
11-09-2005, 12:49 AM
Hello everyone, I have been lurking and planning for the last few months. I have drawings roughed up. I have some electronics in the works and have been bitten by the CNC bug.

The overall scope is this:

Mechanical
25" x 49" x 5" work area (1/4 sheet sizing)
Torsion box / hybrid construction (MDF, Birch Ply)
2" Pipes for X & Y
Dumpster find Linear Bearing for Z
Acme 1/2-10 drive screw, single center drive for gantry- but optional twin drive depending on need.
210 oz/in direct drive motors for starters

Electronics:
Centent 0143 Drivers X,Y,Z
Home built 56v power supply
CNC4PC breakout board.

Computer:
Mach2 or 3 control software for starters, EMC later on.
Still up in the air for CAM software.
Solidworks / Autocad for design.

Goals.
Parts cut from MDF / PLY, 2 pass for a 3/4" would be awesome.
Parts cut from foam
Parts cuts from plastic, likely acrylic, UHMW, etc.
1000in / sec - haha just kidding, 50-75in / min would be a more reasonable goal.

Small scale funiture, R&D environmental design.

OK-- Thanks for watching. I'll try to keep up with the documentation.

jmytyk
11-09-2005, 01:09 AM
Here are some quick sketches in AutoCad, there are not complete by any means.

jmytyk
11-09-2005, 01:30 AM
As you can see in the above pics the design is rough, but some things to note.
The torsion box idea is being borrowed (thanks)
The EMT idea is being borowed (thanks)
The angle with skate bearings pushing in, instead of out (thanks)

I will tie the Y-Axis (short) to the gantry walls with some threaded rod run through the center of the box. It will attach to the gantry walls passing through oversized holes so that adjustments for Z-Axis perpendicularity can be made, we'll see how well this works... The rod will run the full length inside of the box.

The X-Axis (long) will be a nice thick torsion box, since it is spanning 50+ inches. As a mechanical system I know it will pull some support from the tubes, but I would like to have the frame support itself, without needing support from the tubes.

Driving the thing:
The Y-Axis (short) center driven, acme screw. The Z-Axis is still up in the air.
The X-Axis (long) will be center driven, I know there has been conversation about racking, or using 2 drives screws- I'll see how this goes- simpler is better for first time around. The if it is a problem, I could move a pair of screws to the edges and belt drive the thing.

Where have you guys seen flex points when center driving the gantry? are they flexing in the bearing blocks? or what--- thanks...

jmytyk
11-09-2005, 02:17 AM
A question about step / direction and the 5vdc supply. The step and direction come off the breakout card, and originate from the computer. On my setup the 5vdc is coming from a wall-wart. do these have to be link somehow? would it be better to pull he 5vdc for the drive from the computer PS?

Also, concerning the enable in Mach2? pin 15? is this necessary, or can i run without it being attached??

Thanks_Jon

spalm
11-09-2005, 01:25 PM
Hey Jon, congrats on starting your own log.

Flex and backlash come from everywhere they can find a way.

Few comments:
How are you going to attach your pipe? I would recommend lengthening it and routing holes in the end plates and gantry walls for it to pass through.

I see you have two bearing adjuster plate thingies on X. I think you only need one. These are an easy way for flex to enter the system. The less the better.

I sense that you are really worried about Z-Axis perpendicularity. I would worry more about flex and not compromise the strength of the gantry to Y connection.

About twin leadscrews. Just keep it in mind so it is not designed out. Center drive is the easiest and may work for you.

Steve

CJL5585
11-09-2005, 02:19 PM
A question about step / direction and the 5vdc supply. The step and direction come off the breakout card, and originate from the computer. On my setup the 5vdc is coming from a wall-wart. do these have to be link somehow? would it be better to pull he 5vdc for the drive from the computer PS?

Also, concerning the enable in Mach2? pin 15? is this necessary, or can i run without it being attached??

Thanks_Jon

The Wallwart power goes to the breakout board as in attached drawing. The step and direction signals are outputted from the CNC4PC board as follows:

Pin # ID

2 X Step
3 X Dir

4 Y Step
5 Y Dir

6 Z Step
7 Z Dir

8 Not Used
9 Not Used

THE ABOVE WILL HAVE TO BE ENTERED INTO THE MACH SOFTWARE, AS THE PINS FOR OUTPUTTING TO THE CNC4PC BOARD.


Centent 143 Drives are very good units in my opinion. I have built several systems for others using them. They are an older version of the Gecko 202 drives, and the 202 manuals can be used for reference on the 143's as they wire and operate the same externally.

I cannot give you a DEFINATE AND CONCRETE answer on Mach II, but from memory, I BELIEVE THAT IT WILL WORK WITHOUT ANY CONNECTION.

Hope this helps,
Jerry

CJL5585
11-09-2005, 02:29 PM
I Forgot to mention that in the Vicor1. Dxf drawing the +5 / +12 volt supply is the one from CNC4PC. Anyway, the +5 Volts goes to the 5 volt connection on the breakout board and also to pin 10 on the Centent 143 drive. The 5V common ONLY goes to the common connection on the breakout board.

The breakout board common completes the circuit to the step/direction on the drives. No other common connections are necessary.

Jerry

jmytyk
11-24-2005, 06:48 AM
So my build log isn't all that intresting yet. Life outside of the CNC world has kept me very busy. I sold my truck (03' taco xtracab, v6, 4x4, TRD) :( but i found a new old one (87' 'yota x-cab truck, 4x4, 22re, 5sp) only a few blocks from where i live :) with the goal of having some more money towards this type of stuff... so now I am back.

I know you guys say "don't spin the motor before its finished" but I could help myself, I spun them!! 3 drives, 3 spinning motors. not at all optimized by any means, but atleaset they're moving.... Electronics is my weakest link, I made a deal that if I get the electronics done, then I can start the build phase.

I'll post some pics of the frankestein mess in the morning. (3:46am right now) -reminds me of someone elses build log title - tablesaw, kids, 2am.

-Jon

jmytyk
11-24-2005, 09:08 PM
In addition to spinning the motors i would like to test the electronics for limit switches. I saw in the Mach 2 manual the best way to save pins is to wire the limit switches in series

x-limit with x-home
y-limit with y-home
z-limit with z-home

I see this as straighforward, with the switches in the NC (normally closed) configuration, and attach one end to a input pin and the other to ground?

That doesn't seem right, or do I use my 5v wallwart? I posted a pic from the Mach2 Manual, what wattage is the resistor, and does it have to be a 470 Ohm specifically?

Thanks

CJL5585
11-24-2005, 10:28 PM
In addition to spinning the motors i would like to test the electronics for limit switches. I saw in the Mach 2 manual the best way to save pins is to wire the limit switches in series

x-limit with x-home
y-limit with y-home
z-limit with z-home

I see this as straighforward, with the switches in the NC (normally closed) configuration, and attach one end to a input pin and the other to ground?

That doesn't seem right, or do I use my 5v wallwart? I posted a pic from the Mach2 Manual, what wattage is the resistor, and does it have to be a 470 Ohm specifically?

Thanks

The Resistor should be a 470 Ohm 1/8 watt unit. A higher wattage is fine. The actual wattage required is 0.053 watts.

The reason the resistor should be 470 ohms is to set the max amount of current that can flow to the computer port electronics with one of the limit switches in the open position. The 470 Ohm limits the current at 0.0106 Amps or (10.6 ma.). The max that the electronics can take is normally in the range of 16 ma (0.016 Amp). The lower threshold is somewhere around 8 ma (0.008 Amps).

The 5 Volts comes from the (+) side of the wallwart. The common (-) of the wallwart needs to connect to pins 18 - 25 on the breakout board. This might be a common connection already made on the breakout board. It varies from one BO Board to another.

Jerry

jmytyk
11-28-2005, 01:21 AM
Wow, it's amazing what kind of work you can get done when you have the house all to yourself :) Today, limit switches wired and tested, motors "fully tested" -found one wiring problem that I need to trace down- and learning more about Mach2 - all in all a good day. Tommorow will be spent at the DMV getting plates for my new truck, i surely am looking forward to that :)

Here are some pics of my wiring mess. it all works- mostly

next question, motor tuning. top speed and accel settings? I am going by the sound of the motors, there is a definite limit and i can hear it. I understand that these are unloaded settings, and will change depending on everything else the motors are hooked to. Is there other methods?

How hot is "hot" for a motor- is heatsinking them a requirement?
4.6a - 2.71v motors running at drive set current of ~4.5a and 56v

Thanks everyone.

CJL5585
11-28-2005, 01:36 AM
You should have the Powermax II motors wired in series if you are running them at 56 Volts DC.

They will get VERY HOT at 56 Volts in Series, and will want to melt if wired in Parallel at this voltage.

If they get too hot -- I would make some wrap-around metal heatsink fins for the motors.

I use these motors too, and plan to heatsink mine. They are wired series at 48 VDC.

Jerry

jmytyk
11-30-2005, 01:19 AM
Hey Jerry, What amperage do you have your drives set to? Is having them set to 4.5 too high? I get the unipolar bipolar parallel series amperages all mixed up. _Thanks

CJL5585
11-30-2005, 03:25 AM
Hey Jerry, What amperage do you have your drives set to? Is having them set to 4.5 too high? I get the unipolar bipolar parallel series amperages all mixed up. _Thanks

Use a resistor value about 33K on each of the drives. A lot of current is not required at higher voltages. Everything is calculated on wattage (Power).Your motors are rated at 4.6A @2.71 Volts. Power(watts) = Amps x Voltage.

Your motors Nameplate rating: 4.6 x 2.71 = 12.466 Watts

P divided by Voltage = Current (Amps)

12.466 watts / 56V = 0.222607 Amps (less than 1/4 Amp each motor) So you see, you don't need a lot of current. All the extra current is dissipated as HEAT FROM THE MOTORS.

I used resistor values of 33K - 47 K with supply voltages of 48 Volts.
Jerry.

jmytyk
12-27-2005, 11:10 PM
Ok-- I know it's been awhile, but here I go again.

I have decided to drive the gantry with 2 screws. I have a few questions that I haven't been able to sort out.

1. What gear ratio should i go with mating to 1/2 acme stepper (not servo)? 1:1, 2:1, 3:1, higher? (motor: screw)

2. Belt pitch/ spec? GT3, GT5, others?

3. Concerning driveline loss (roughly 2x the friction in the system) I was looking in the range of a 640oz/in stepper good/bad?

4. Using 2:1 for example. is it better to go 40:20, 60:30, or the other direction, 20:10 -already taking into account minimum diameter for a given belt size. is there an issue with power loss of the motor because the motor has to spin something larger, like the 60:30 example?

ok- thanks everyone, and oh yeah- I am redrawing everything in solidworks to make it look pretty...

_Jon

ViperTX
12-28-2005, 12:56 AM
Well...I would back off and use a single drive screw and directly couple your 640 oz-in stepper to it.....simplify your life.....

jmytyk
12-28-2005, 01:10 AM
hmmmm... now the debate for and against dual lead screws on a 30 in wide gantry? from what i have seen in other build logs- dual will greatly help solve the racking issue on something this wide.... also, i haven't bought any of the drive components for this design change yet- so the 640 is just a number, the pulleys are just numbers as well...

ger21
12-28-2005, 02:35 AM
Viper, I've yet to see a wooden machine with a single screw driven gantry that wouldn't rack.

If you're using 1/2-10, you probably don't want to gear it down, but you might want to gear it up 1:2. Direct drive would need 1000rpm to get 100ipm. An easier way than gearing would be to just get 1/2-10 2 start or 1/2-8 2 start.

jmytyk
12-29-2005, 01:54 AM
Gerry, Thanks for the reply. I have already bought the 1/2-10 screw, so i am going to stick with that one this first project. I have been looking at motor torque curves trying to figure out how to run 2:1 or even 3:1. I have come to he conclusion- I have no idea how many oz/in it takes to move this thing and on top of it- this thing isn't fully designed yet. (lets say 55lbs)

So a poll of sorts, does anyone have these setups, or do you think these number are doable?

2 screws 1/2-10 1:1 with a 640oz/in motor @450 rpm = 45 in/min
2 screws 1/2-10 2:1 with a 640oz/in motor @450 rpm = 90 in/min

or

2 screws 1/2-10 2:1 with a 1200oz/in motor @400 rpm = 80 in/min
2 screws 1/2-10 3:1 with a 1200oz/in motor @400 rpm = 100 in/min

- I chose the RPM's based on the curves- the motors have approx same output, the curve at lower RPM's allow the thing to get moving with the said gear ratios.

Thanks
_Jon

jmytyk
12-29-2005, 02:00 AM
oops, this would help...

CJL5585
12-29-2005, 03:12 AM
[QUOTE=jmytyk]I have been looking at motor torque curves trying to figure out how to run 2:1 or even 3:1. I have come to he conclusion- I have no idea how many oz/in it takes to move this thing and on top of it- this thing isn't fully designed yet. (lets say 55lbs)

So a poll of sorts, does anyone have these setups, or do you think these number are doable?

2 screws 1/2-10 1:1 with a 640oz/in motor @450 rpm = 45 in/min
2 screws 1/2-10 2:1 with a 640oz/in motor @450 rpm = 90 in/min

or

2 screws 1/2-10 2:1 with a 1200oz/in motor @400 rpm = 80 in/min
2 screws 1/2-10 3:1 with a 1200oz/in motor @400 rpm = 100 in/min

- I chose the RPM's based on the curves- the motors have approx same output, the curve at lower RPM's allow the thing to get moving with the said gear ratios.

Thanks
[QUOTE=jmytyk]
--------------------------------------------------------------------------

You need to get your bearings installed and aligned until it is satisfactory to you. Then tie a string to one end of the table and have a weight scale which you hold in your hand. (This is with the screw not attached to the table). The reading on the scale required to move the table will be the pull weight.

The pull weight will probably not be more than 5 pounds. Then you can calculate the size motor (torque) that is required.

A pull weight of 5 pounds using direct coupling between motor and screw would require approximately 150 oz/in of torque to move the table from a dead stop. Once moving, it will most likely require less torque.

Anyway:
The 640 generates approx. 425 oz/in of torque @ 450 RPM.
The 640 generates approx. 141.6 oz/in of torque @ 1500 RPM.

The 1200 generates approx. 375 - 400 in/oz of torque @ 450 RPM.

One 640 stepper direct coupled to the screw should move your table if it weighs 55 pounds, and the bearings and screw do not have excessive drag.

Maybe this will get you to the point where you can get an actual pull weight measurement.

Hope this helps.
Jerry

jmytyk
12-29-2005, 07:56 AM
Yeah I guess I was wanting to have everything planned to the "t" before I started building, but it doesn't seem that way unless I want to model everything to the "t" and do all the system specs in the computer- which will have in itself a degree of error. So- the "time" is the toss up- learn solidworks to the point of being able to virtually model this system to "some" degree of accuracy or spend that time actually building? -then do the calculations

ok- so my time is being shifted to drawing the final the design and getting to the point of building....

Thanks Jerry for the claifaction and direction. -Drawings to be posted...

ger21
12-29-2005, 07:56 AM
A lot of current is not required at higher voltages. Everything is calculated on wattage (Power).Your motors are rated at 4.6A @2.71 Volts. Power(watts) = Amps x Voltage.

Your motors Nameplate rating: 4.6 x 2.71 = 12.466 Watts

P divided by Voltage = Current (Amps)

12.466 watts / 56V = 0.222607 Amps (less than 1/4 Amp each motor) So you see, you don't need a lot of current. All the extra current is dissipated as HEAT FROM THE MOTORS.

I used resistor values of 33K - 47 K with supply voltages of 48 Volts.
Jerry.

Torque is directly proportional to the current supplied. Cut the current in half, and you'll have half the torque. It's the higher voltage that creates the excess heat.

A motor running at half the rated current and 25times the rated voltage should run much hotter than one running at the rated current and 10 times the rated voltage. Here is a quote from Gecko's stepper motor white papers:



It is recommended the motor current always be set at the rated value also to get the best microstep smoothness. Setting the current higher degrades the linearity of motor and causes microstep bunching and attendant low-speed vibration.

What comes with increased motor power with increased power supply voltage is increased motor heating; this heating increases more rapidly than output power and ultimately sets the maximum output power from the motor. That is to say, the limiting factor in how much power a motor can deliver is ultimately determined by how much heat it can safely dissipate.


The whole paper can be read on the Geckodrive Yahoo group.
http://f6.grp.yahoofs.com/v1/wM-zQzmSSS-CM_BrNCPGj5Hx9IiElb16dkdmd_wB3CHcIF6Y-OqgFWBJUvXK6T8-jLTwEu31ZHricPE2RQC9Cdn-4I6DvYS5/Gecko%20white%20papers.pdf

ger21
12-29-2005, 08:09 AM
.....I have been looking at motor torque curves trying to figure out how to run 2:1 or even 3:1. I have come to he conclusion- I have no idea how many oz/in it takes to move this thing and on top of it- this thing isn't fully designed yet. (lets say 55lbs)

So a poll of sorts, does anyone have these setups, or do you think these number are doable?

2 screws 1/2-10 1:1 with a 640oz/in motor @450 rpm = 45 in/min
2 screws 1/2-10 2:1 with a 640oz/in motor @450 rpm = 90 in/min

or

2 screws 1/2-10 2:1 with a 1200oz/in motor @400 rpm = 80 in/min
2 screws 1/2-10 3:1 with a 1200oz/in motor @400 rpm = 100 in/min

- I chose the RPM's based on the curves- the motors have approx same output, the curve at lower RPM's allow the thing to get moving with the said gear ratios.

Thanks
_Jon

I've seen people here posting that they can get 100+ipm using ~270oz motors, running a XYlotex at 24V and 1/2-8 2 start. Keep in mind this can be very machine dependant. I'd go with that and the 640oz motors. That would be 1:2.5 (motor:screw)

Buying 1/2-8 2 start from MSC, for about $30 for 6 ft would most likely be a lot cheaper than buying belts and pulleys, and much easier to install.


PS.


2 screws 1/2-10 3:1 with a 1200oz/in motor @400 rpm = 100 in/min
This should be 120ipm. ;)

ger21
12-29-2005, 08:20 AM
As you can see in the above pics the design is rough, but some things to note.
The torsion box idea is being borrowed (thanks)
The EMT idea is being borowed (thanks)
The angle with skate bearings pushing in, instead of out (thanks)


I don't know if I posted it in my thread (I think I mentioned it in Joe's), but after installing my gantry, just moving in back and forth 20-30 times by hand has worn the galvanized coating off the EMT and the bearings have worn flat spots. Only on one side, though. Apparently one pipe has a harder coating than the other side.


The X-Axis (long) will be a nice thick torsion box, since it is spanning 50+ inches. As a mechanical system I know it will pull some support from the tubes, but I would like to have the frame support itself, without needing support from the tubes.

On mine, the torsion box is much stronger than the EMT. If you think the EMT will support the table, you're incorrect. My table spans 59.5 inches, and weighs at least 50lbs, probably closer to 75. When complete with router and steppers and screws, my gantry should weigh about 75lbs. The torsion box can carry it's own weight with no flex at all, and the deflection from the gantry weight is negligable, if there is any at all.

My gantry torsion box is framed with 3/4" Baltic birch, with 1/2" BB skins. It weighs 22 lbs. I'd recommend building the whole thing with 3/8"BB, both frames and skins. It would be the same strength, and about half the weight.

ger21
12-29-2005, 09:11 AM
You need to get your bearings installed and aligned until it is satisfactory to you. Then tie a string to one end of the table and have a weight scale which you hold in your hand. (This is with the screw not attached to the table). The reading on the scale required to move the table will be the pull weight.

The pull weight will probably not be more than 5 pounds. Then you can calculate the size motor (torque) that is required.

A pull weight of 5 pounds using direct coupling between motor and screw would require approximately 150 oz/in of torque to move the table from a dead stop. Once moving, it will most likely require less torque.

My gantry will move with light finger pressure, probably less than 1 lb of force. Accelerating it rapidly, however, takes much more.

The Formula for Force from a leadscrew is
F = (Torque x 2Pi x Efficiency)/ Screw Lead
1/2-10 acme with a plastic nut is about 40% efficient.
Assume 400oz-in of torque

Force = (400 * 2 * Pi * .4) / .1 = 10053oz /16 = ~628lbs of force.

If you look at Nook's ACME catalog, http://www.nookind.com/pdf/NookInchAcmeScrew.pdf you'll see that 1/2-10 with a plastic nut needs .039 in-lbs, or .624oz-in, to move 1lb. 400/.624 = about 641 lbs of force. About the same as the other formula.

Rotational inertia will use up some of that, increasing with acceleration, but as you can see, 400 oz-in of torque will give you a lot of force. Switching to 1/2-8 2 start will still give you 400lbs of force. Using dual screws and motors, that's 800 lbs of force to move your gantry. Stay out of it's way.

jmytyk
12-29-2005, 05:11 PM
Gerry--- You Rock! Thanks for calc's

Ok- so 2 screws it is. The 640 on/in motor- that will run at ~400 oz-in through 400 rpm's

I can expect to have some driveline loss because of 1 motor and 2 screws, but 40% efficency for a plastic nut on acme thread- that sucks! No wonder why ball screws are so expensive.

What is the suggestion on gearing? 1:1 to start with, that will put the torque fall off right in range with a resonable cutting speed. I am thinking it could go 2:1 rather easily?

jmytyk
01-25-2006, 10:31 PM
I got some 1" black gas pipe, mainly for wall thickness. Got them all cut up, and sanded them down nice and smooth. Somewhere i saw someone building a lathe for sanding the pipes, so hey- this should be no problem. And ya know, it really was. Here are some pics.

1.) big-ole caster, rubber wheels worked very well...
2.) the missing link, 3/4in socket, perfect! A 5/16" bolt, grind down 2 points so it is 1/2" across the faces- instant 1/2" drive for a drill.
3.) nice and smooth sanded pipes.

jmytyk
01-25-2006, 10:36 PM
Here they are in all their glory, my own linear bearings. I need to trim the ends on the bolts- a bit excessive. Found out that a nut + 2 washers is just about right in terms of spacing. As per design they are -0.009 too close to the angle.
Works ok for me. 4 for the long axis, 2 for the short :wee:

jmytyk
03-24-2006, 03:08 AM
I was sad to see that I had to go to the second page of build logs to find my log. I missed my last month's update- a goal I was trying to keep, enough work done per month to post something reasonable. Well some news- I have made up for it, I got a whole lot of updates.

Speced and ordered all the sdp-si parts, I went with the GT2 5mm size. I ordered 3 pulleys and really long belt. This was an expensive design change, but looking at the system, it will be worth it. Also- this size of pulley may be slightly overkill, but I can always them for my next machine.

Part numbers:
A 6A55-034DF1516 $15.70/Each $15.70
A 6A55-034NF1516 $14.70/Each $29.40
A 6R55M344150 $36.16/Each $36.16

I decided to go with Birch plywood instead of MDF. I have seen that MDF is much stiffer, but joining it requires more of a mechanical joint than just glue and screws. Since everything on this machine is handmade the first time around, I decided against MDF. I've went through 3 sheets of plywood so far, Likely will go through another sheet before i am finished.

Another design change that I borrowed from the group was to use Uni-Strut as a pipe base. Instead of drilling the 1/2 circles in the torsion box ribs I used pieces of uni-strut. Then, drill the pipes and bolt to the Uni-strut. I found the funky Uni-strut bolts that pull the pipe into the metal instead of the plywood box, works very well (yes i know, a picture is definitly needed)

Built:
Y-Axis Torsion box built, with 2 rails mounted
X-Axis Torsion box built, 4 rails mounted

Currently I am mocking up the dual screw drive system, I managed to get the wrong bearings for the Acme screw support. This small error will require some creative filing to a handful of washers... anyways sorry for a pictureless post, I will get some on in the morning. _Jon

jmytyk
03-25-2006, 01:39 AM
My un-knowledge of metal working has surfaced once again.

I need to 'turn down' some washers to use as spacers against the lead screw support bearings. I don't have a lathe, nor access to one. So using my drill press as a 'vertical' lathe I put 10 washers on a bolt, used a nut to compress the stack together and chucked it up into my drill press. I tried to use a file to cut the washers down, no luck- or VERY slow progress (at the slowest speed ~500rpm's faster was worse) . My next attempt will be using my dremel with grinding stone instead of the file- what do you guys think of this? I was researching on the 'zone using to do this, similar to the process of turning down the end of a ballscrew, but my setup will be handheld. One big question, doing this on a lathe, is the lathe running, or is it turned by hand?? Any other warnings you guys can think of?? Would annealing the washers be needed, or am i just being impatient?--Thanks _Jon

ger21
03-25-2006, 08:11 AM
How much do you need to take off?

spalm
03-25-2006, 10:54 AM
Turning down those washers sounds like a PITA. I had the same problem. I am using plastic spacers bought at the local hardware store. They are a bit thick and I had to round the edges a little with a sander so they wouldn’t rub, but they work OK. I have also used really short pieces of 1/2” copper pipe cut off with a pipe cutter. If you ever find a better spacer, please let me know.

Here is a pic (http://www.cnczone.com/forums/attachment.php?attachmentid=15681) of both types in use.

Steve

jmytyk
03-25-2006, 06:25 PM
I just got back from Lowes with a 5ft piece of copper pipe, cut up some spacers. Piece-o-cake. Took some pics and came in to post them, and what'd ya know- a pic of the same thing already up :) :cheers:

Gerry, turns out filing the washers didn't work out at all. I managed to make a big pile of metal dust. Although I did knock off a good amount from the diameter. The whole idea needed to be readdressed. The copper pipe space, much better solution

Here is the driven end mocked up - the plywood spacer represents the frame, wow, that take up a lot of travel... :(

jmytyk
03-25-2006, 08:00 PM
So here she is, the dual screw drive arrangement, well at least the end view of it...

First pic is the belt path.

Second pic is where I need to put an idler pulley, the number of teeth in mesh is approx 1/3 of total teeth per pulley.

The sdp-si site has an interactive belt length calulator, but only works with 2 pulleys. :( I'll see how this works- it is going to be plywood for the first go around, but i definitly think it will need to be remade from aluminium either totally or just the bearing surfaces.

jmytyk
03-25-2006, 08:11 PM
Here's the boxes, check out the Uni-strut mounting points. The rails are bolted though with cap-head allen bolts....

about those style bolts, they sure are nice. Is there a engineering reason why they are used? Stronger, better torquing / stress characteristics?

Netjams
03-29-2006, 06:35 PM
It's looking good and BIG too.

jmytyk
05-30-2006, 10:19 PM
:/ two months and no work done on the machine... however ther girlfriend is happy, i built her some funiture, i revamped by swamp cooler for the Vegas summer, finished up some sprinklers, cut a hole in the wall to mount a A/C unit, painted the shop and some other odds and ends. as soon as i get done with this funiture project, maybe tommorrow, i'll get back to work on the x-belt drive setup. i might work on the funiture tonight, or maybe clean the shop- holy god its a wreck! can't even see the floor...

so, anyways, who around here has any experience setting up a machine to run from USB? i found the gecko external pulse generator. is there any other options people know of? how about a parallel expansion card for an express34 slot??

- hopefully a real update by the end of the week...

ViperTX
05-31-2006, 12:20 AM
Yeah...those darn girlfriends.....I've been working on my CNC router for a year and have also built the gf 3 major pieces of furniture and built one for a neighbor and built 3 patio table for myself and 1 table base for a neighbor's sailboat....

jmytyk
10-06-2006, 08:22 PM
I'm back and have gotten a good amout of work done (pictureless for now - i'll post some when i get home)

I have gotten the end plates mounted for the screws to attach to. This will probably need some work once things are moving. I think there is going to be a good amout of flex involved in the design, but hey- that's the fun of this, build it once, then rebuild it.

I have the pipes paralleled, to each other, but not across. I have no idea how to get them parallel to each other on oppostie sides. - I don't think it is necessary though. I have decided to use Uni-Strut for the Gantry uprights and lower cross piece, this will give more adjustment than really necessary, but will make up for some of the sloppyiness in the build, (or make for a bigger headache. )

I have roughed in the attachement of the screws and spun them with a drill. WOW- is there a lot of whip. I have figure out what i'm going to do with this. I think it has to do with the 'temporary' nature of their attachment, and that i can't put a ton of tension on them yet. But i alreay see in the near future the need for 2 start screws. (anyone need 3 pieces of 1/2x10 6ft acme in las vegas??)

my goal is to have a rolling gantry in the next 2 weeks... maybe even with a rolling - y.

what else??? there's a ton more going on in my non-cnc world, hence the 4 month absence to this thread- oh yeah, did i mention how freaking hot the vegas summer is? more the reason for the break.

pictures tonght...

jmytyk
10-06-2006, 08:25 PM
...

jmytyk
11-30-2006, 08:31 PM
so here is the rough mounted gantry uprights, they are mountd on the on an upside down machine at the moment. -that's why the pic looks funky. i used uni-strut instead of ply / mdf so that i could adjust the geometry without having to rebuild the whole thing.

this brings up an intresting question- where should the cutter be in relation to the drive nut? directly over top if possible? it seems like the whole thing is a bit to "extended" for lack of a better word. what do you guys think??

jmytyk
11-30-2006, 08:35 PM
i got the belt drive done - well atleast version 1. i am having a belt tracking problem. i think it might be caused by the screws flexing, the gantry is not attached yet. it could also be a tension problem- how much tension is supposed to be on the belt? currently it is tight but it can be put on by hand. - so i dont think it is too tght... i knew this was going to be a 'challenge' but after having the gantry rolling on the x-axis, there is no way that this thing could be driven with one screw, way too much racking potential...

jmytyk
11-30-2006, 08:40 PM
and.. i managed to get a rolling gantry done. this turned out NOT to be a huge problem in terms of adjusting (yet) it rolls easily, no binding, no rough spots, and nothing is cranked too tightly... so all is well... uni-strut is nice to tweak. loosen the bolt, whack-it into place... next step is to flip the thing over to mount the y-axis and work on that axis...

jmytyk
11-30-2006, 08:50 PM
yep- ugly... but it works, one of the first pieces to be cut out when this thing is running...

belt drive detail... "pulley" works out okay. any notes on how much crap is coming off the belt? will this slow down, or is this a sign of the belt getting "eatten"

jmytyk
12-06-2006, 02:42 AM
Update...

hmm... where did i leave off?? oh yeah, upside down, way too much work to do...

I built a rather substantial sub frame for this thing to sit on, also with a dual purpose of not letting the dual x-axis drive screws destroy the end plates. if anyone wants to see what it looks like, i'll post a pic- but it wasn't too exciting, and I got tired of the cold by the time it was for pictures...

With the sub frame finished, I got two of my buddies to help me flip this beast over, now it's right side up. Threaded rod through the gantry torsion box, to get this thing attached to the uprights went smoothly. Next time i will drill larger holes than 1/4" maybe 5/16" - the 1/4" "felt" small. I had to build a small jig to help me hold the weight of the thing while I made sure the box was perpendicular and finished bolting it together.

I didn't run into any major problems, although I managed to waste 4 hours building a y-axis trolley that didn't work. I know it is about process, but for the love, I knew it wasn't going to work - but I kept building anyways. It got to the point where it was really wrong, then I stopped... ha.

So after the minor delay, I made a materials run, got some more Uni-Strut, chopped it up- and built this version of the Y-Trolley. It looks funky and not really all that pretty, but it works and is tight.

Alignment overall is okay, it is running downhill 0.04" on one side of the the X-axis over ~60" and left to right on the Y-Axis is 0.01" over ~36" The gantry torsion box has a twist in it, 0.0625" over the short side. This might come up latter when things are more perpendicular at one end, than at the other end... but hey- not too bad for a first shot. he left to right was MUCH worse when i checked it a second time, more like 0.125 over the ~36" - the joy of uni-strut, loosen the bolts, apply hammer, tighten bolts... I completely understand the joy of 8020... next machine...

So a question... when I put a piece of MDF as a spoil board and run a router over the surface this will take car of the uphill/downhill left/right problems??

Checkout what I found for a Z-Axis... freebie in the the dumpster...

That's all for now...

jmytyk
01-25-2007, 07:02 AM
I got a good amount of work done over the last few days. Got a Z-axis built, and got it mounted. Round one of getting it squared up. I'm sure i'll have to revisit the issue again. anyways, here's some pics.

(it is more square than it looks, there is a sticker that makes it look like light is showing through)

Mr.Chips
01-25-2007, 11:27 AM
Update...

hmm... where did i leave off?? oh yeah, upside down, way too much work to do...

I built a rather substantial sub frame for this thing to sit on, also with a dual purpose of not letting the dual x-axis drive screws destroy the end plates. if anyone wants to see what it looks like, i'll post a pic- but it wasn't too exciting, and I got tired of the cold by the time it was for pictures...

With the sub frame finished, I got two of my buddies to help me flip this beast over, now it's right side up. Threaded rod through the gantry torsion box, to get this thing attached to the uprights went smoothly. Next time i will drill larger holes than 1/4" maybe 5/16" - the 1/4" "felt" small. I had to build a small jig to help me hold the weight of the thing while I made sure the box was perpendicular and finished bolting it together.

I didn't run into any major problems, although I managed to waste 4 hours building a y-axis trolley that didn't work. I know it is about process, but for the love, I knew it wasn't going to work - but I kept building anyways. It got to the point where it was really wrong, then I stopped... ha.

So after the minor delay, I made a materials run, got some more Uni-Strut, chopped it up- and built this version of the Y-Trolley. It looks funky and not really all that pretty, but it works and is tight.

Alignment overall is okay, it is running downhill 0.04" on one side of the the X-axis over ~60" and left to right on the Y-Axis is 0.01" over ~36" The gantry torsion box has a twist in it, 0.0625" over the short side. This might come up latter when things are more perpendicular at one end, than at the other end... but hey- not too bad for a first shot. he left to right was MUCH worse when i checked it a second time, more like 0.125 over the ~36" - the joy of uni-strut, loosen the bolts, apply hammer, tighten bolts... I completely understand the joy of 8020... next machine...

So a question... when I put a piece of MDF as a spoil board and run a router over the surface this will take car of the uphill/downhill left/right problems??

Checkout what I found for a Z-Axis... freebie in the the dumpster...

That's all for now...

You are making a lot of progress.
In the second photo it shows the stepper drive pulley, you might need another idler pulley above and to the left of the drive, like you have on the right side. Currently you don't have many teeth engaged into the pulley, this would improve this. I might slip under load.

Mr.Chips
01-25-2007, 11:55 AM
I think your "Y" axis side mounts made up of 2 verticle and 2 angles back to the "Y" axis might be made more rigid. Possible add a third piece to make a triangle. This would be stronger.

Hager

joecnc2006
01-25-2007, 11:59 AM
Just a question and maybe a sugestion.

on your Y-axis top bearing block what hold it into place from shifting foward or backwards? I know in the front it apears to be against the elec. struts but the rear has a gap, on my "Joe's CNC Model 2006, mine has a front side and rear walls which the bearing slide fits snug against to keep it from shifting.

Also, on the adjustment bolts on to of this axis, you may want to add some type of footing or cap on the bolts to help distribute the pressure (al though it is very slight) over time the bolts will eat into the wood, On mine i use HDPE for the Y-Axis bearing slide so i do not have to worry about this issue.

Your gantry seems a little further back and the Z-axis will be adding some pressure to be pushing to the rear, I know we need to try to find the center of gravity so to speak for the gantry, but with your setup, it is easyier to adjust once the machine is running.

Joe

Mr.Chips
01-25-2007, 02:46 PM
In reference to the 7th photo in Post 70.
Run your Y axis all the way to the end of the X axia travle. Would your router be out past the table cutting area? It looks like it might.

If it does you might want to not have the Y axis lean back so far. This will also meke your machine more conpact and stronger.

Hager

jmytyk
01-25-2007, 03:35 PM
wow, i thought everyone left me :-)

pulleys- i have more than 1/3 in mesh, its easy enough to add another one if you guys think.

triangle the y- thanks - i didn't see that one. although the thing is extremly rigid at the moment, but i'll take a look at it.

joe- i see the what you mean about the CG - the cutter is about 3/4" behind the center of the truck. i'll watch this.

Mr. Chips, the shape of the gantry 'sides' need some work. but luckily it's 4 bolts.. we're are on the same page -- thanks

any one want to see other picture angles?

Mr.Chips
01-25-2007, 07:40 PM
wow, i thought everyone left me :-)

pulleys- i have more than 1/3 in mesh, its easy enough to add another one if you guys think.

triangle the y- thanks - i didn't see that one. although the thing is extremly rigid at the moment, but i'll take a look at it.

joe- i see the what you mean about the CG - the cutter is about 3/4" behind the center of the truck. i'll watch this.

Mr. Chips, the shape of the gantry 'sides' need some work. but luckily it's 4 bolts.. we're are on the same page -- thanks

any one want to see other picture angles?


No special pictures just keep up the progress.
Hager

jmytyk
01-25-2007, 10:10 PM
So hey, I got a router attached to the Z- it's pretty simple in method- just a pair of band clamps. It is very secure, and when i try to flex anything, the y-axis bearing trucks are the only thing that move.

so here is another question / problem, how hard should i drive the z into the table and not expect any play in the mechanism? the top bearing truck on the y-axis has a bit of flex in it. and looking at the pictures again- it really is very ugly, i think i will need to revisit the design.

Joe- to answer you question. the top y-axis bearing plate has 6 bolts pressing on it- 4 from the top, and 2 from the rear, the front is captured against the uni-stut. and thinking about it- this is probably where the my flexing problem is located. In your design is the top truck 'floating'- is there a similar problem when driving into the table.

Mr.Chips
01-25-2007, 10:38 PM
So hey, I got a router attached to the Z- it's pretty simple in method- just a pair of band clamps. It is very secure, and when i try to flex anything, the y-axis bearing trucks are the only thing that move.

so here is another question / problem, how hard should i drive the z into the table and not expect any play in the mechanism? the top bearing truck on the y-axis has a bit of flex in it. and looking at the pictures again- it really is very ugly, i think i will need to revisit the design.

Joe- to answer you question. the top y-axis bearing plate has 6 bolts pressing on it- 4 from the top, and 2 from the rear, the front is captured against the uni-stut. and thinking about it- this is probably where the my flexing problem is located. In your design is the top truck 'floating'- is there a similar problem when driving into the table.

As far as the downward pressure is concerned someone else might be able to help. Every flexes, just how much is too much. If the flexing distorts the hole, for sure it's too much.

The question might be: Using a 3/8” end mill what would the feed rate be to drill down X amount into some MDF material. While it is drilling down place an indicator on the top and see how much flex you get.

The flex will vary with the feed rate as it will be harder to drill faster than slower. But someone can give you an expected feed rate.
Think they would need to know your stepper/servo power and router HP and possibly other things.

Not a lot of solid test methods, just some rambling.

Good luck

Hager

joecnc2006
01-26-2007, 12:12 AM
So hey, I got a router attached to the Z- it's pretty simple in method- just a pair of band clamps. It is very secure, and when i try to flex anything, the y-axis bearing trucks are the only thing that move.

so here is another question / problem, how hard should i drive the z into the table and not expect any play in the mechanism? the top bearing truck on the y-axis has a bit of flex in it. and looking at the pictures again- it really is very ugly, i think i will need to revisit the design.

Joe- to answer you question. the top y-axis bearing plate has 6 bolts pressing on it- 4 from the top, and 2 from the rear, the front is captured against the uni-stut. and thinking about it- this is probably where the my flexing problem is located. In your design is the top truck 'floating'- is there a similar problem when driving into the table.

In my design the y-axis bearing block fits snug front, rear and sides, and is held down with 4 5/16" bolts no movement is detected in that area.

jmytyk
01-26-2007, 10:06 PM
i got some work done last night, mainly re-did the y-axis bearing blocks and attach the y to the screw. i have no idea what i was thinking with the ghetto looking ones that were on there, totally useless.

i got a close up of the top y-axis top bearing block. I noticed the flex is not coming from here- although it was a bit loose. When i cranked down on the bolts, the flex then transferred down to the long x-rail pips. again i think my pre-load is the problem, but like joe pointed out earlier, i have nothing to dissipate the force- one more thing to add to the 'fix-it' list

jmytyk
01-31-2007, 10:34 PM
i know picture-less posts are not as much fun, but i do have a log update.

I worked on making new bearing blocks for the x-axis. 4 blocks in total- for 2 screws. they are all mounted up, as well as brackets to keep the 2 dumpster AB nuts in place on the lower gantry. I also got the belt drive all nice and neat, the 2 sides are as parallel as i can measure them - somewhere in the 0.002 range. my measurements were causing me headaches to get it closer than that- so anything close is going to have to wait until i get the machine under power and am able to get some baselines scribed into the surface.

if you guys are familiar with the dual screw drive i built, i ran into a belt tracking problem. when the belt is tensioned- the very ends of the screws are bent inwards due to the belt tension. this tilting is causing the belt not to track correctly and run itself right of the pulley. i think i am going to try to design something that will support the very end of the screw. - any ideas? i have room for 1 bearing- and the attachment to the frame is going to be quite a challenge.

also the z-axis screw is bent, so i have to make some new bearing blocks for that, as well as get them mounted to the z-axis. oh well, it's seeming to be a never ending project- but it is getting ever closer to the end.

-J

jmytyk
02-09-2007, 02:04 AM
maybe i'll get some pics up later tonight or tomorrow.

i got the y-axis motor mounted. nothing special, mainly a copy of what everyone else was doing.

got the x-axis fixed up, i ended up tying the 2 ends of the screws together to get them to push off each other to counter act the inward bowing that was happening and making the belt track all funky. it still doesn't track perfectly, but at least it doesn't run itself right off the pulleys.

i rebuilt the z-screw so that it is now 1/2-10 like everything else- and fixing my bent screw problem.

and cleaned the shop.

so i have some question about wiring 8-wire motors. i have a 56v PS. (more details later about the specific motors.)

-J

jmytyk
02-20-2007, 05:32 AM
here's some pics of the things i've gotten done.

the x-axis 'fix' is okay- this whole drive end needs revistiing once the machine is up and moving, i might upgrade to 2 start screw sooner than later.

I also put the electronics in a metal box. this box only contains the logic- the power supply is going to go in another box, more on that later.

jmytyk
07-04-2007, 01:11 AM
Hi guys, it's been a long while and i hope that i could get some help. I wired up my electronics boxes nice and neat, inside of metal enclosures. When i flipped the switch, i blew 2 of the 3 fuses to the drives. I think i sorted that problem out, i pulled everything out of the boxes and am trying to get the motors tuned while sitting on a bench. Here is a run down of what i have.

-Mach 3
-Home built PS metered at 53v - in the 10amp range.
-Centent drivers CN0143 - (10microstep) (Mach 3 is set to 20k steps/unit)

-cnc4pc BOB, basic model, 5v wall wart.

-win2k, stripped down, runs frequency check with no big blips. it's dual xenon 2.8ghz, no hyper-threading

-i have the motor wired in bipolar series, set at 3amps now, but did try it at 7amps for a short period. locks up great when i turn the power on.

At best, i could get a motor to spin at a mid to high speed- no low speed, and not necessarily in the correct direction. Using manual control in Mach3, the motor will spin, but sometimes it will go in the correct direction, sometimes it won't. Playing with the speed and accel in motor tuning has yet to help. I have all combinations of active hi/low on the step and direction. I have tried playing with the pulse duration all the way up to 15uS - 2,5,7,10,12,13,15 i tried higher on the step, lower on the direction, and vice-versa.

I had shielded wire on the step / direction signal wires, the shield was hooked to the BOB ground- so that it would ground to the PC. just to try i did disconnect the shield- so it's just hanging there right now.
(i'm going to try to use just normal wire)

there was one thing that was strange, i picked up a motor when it was powered and it twitched... maybe 1 step...

and here is where i'm stuck... this is the same across 2 drives, so unless i blew up both... what are the symptoms of a bad stepper?? i have 3 more, but they are all mounted to the machine- i don't want to pull them unless i have to.

i work in a place with lots of technology- i know there is an o-scope around somewhere, but i don't know how to use one. to me it seems like software.

help.... please...

thanks
jon

jmytyk
07-04-2007, 01:43 AM
just a quick update. i pulled the shielded wire off the step/direction to the drive- swapped out with some normal, non-shielded stuff... seems to work better.

I still have the problem of the motor not turning in the correct direction all the time... press left, goes CCW, press right goes CCW, sometimes CW...

jmytyk
07-04-2007, 08:41 PM
3 dead drives...

here is some background, post 17 onward.
http://www.cnczone.com/forums/showthread.php?t=22205&highlight=cold+chisel

broke the cases open and there was a visibly fried transistor in each drive.

i am in Las Vegas, and there is a possibility that i will be moving overseas, anyone interested in a partially finished machine? if not i'll start parting it out.

:(

jmytyk
08-12-2007, 08:11 PM
machine is completely disassembled... :-(

all the uni-strut parts are free if you are in Vegas, i'll sell off anything else that you might need...