Should there be play in the Vexta steppers?

[danielb273]

Guys, I'm stumped. I've had my PRT96 for two years now, and I'm noticing the quality of my vgrooving is deteriorating. No matter how slowly I run my parts, I end up with "chatter", like the router bit is wobbling. After searching the forum, there are several possibilites, and having checked them, the only one left that I have no information on is the stepper motors. Here's what I've done so far:

1. Checked table for square and loose bolts. It was a bit out, so I squared it up.

2. Checked the Y axis hold down bearings. They're tight (perhaps TOO tight?)

3. The router appears ok. The chatter I'm seeing in my machined parts is severe, and gets worse with faster move speeds, which to me at least dismisses the router as the prime suspect.

4. The table is tight.

5. The pinion gears are tight, square with the table and lubricated.

6. I'm clamping my work pieces to the table, and in addition I'm tacking the corners with 1 1/4" 16 gauge brads.

7. The bit is sharp.

8. I've tightened the Y-rail bolts down (they were a bit loose), and checked the bolts holding the corners of the gantries.

So, that left me with a few possibilities, and I need some other graybeards to help ease my suspicions.

All of the Vexta steppers (Model A6497-9412KTG) have just a SLIGHT bit of play (backlash) in them when I unhook them from their racks. It's very minor, but is it normal? Should I hear just a faint "click" when I gently try turning the gear first one way and then the other?

If that's normal, then my next observation was that the X-car seems to "bounce back" during a short move at anything above 0.5 ips. It's almost as if the stepper isn't holding the position (or that the "slop" I mentioned in the previous paragraph is the culprit). I get GREAT repeatability, but by observing the router during an air cut I can see the "shaking" of the X-car at the router bit, which I believe is causing the chatter I'm seeing.

So I'm left scratching my head, asking, is it the stepper motors causing the shakiness of the X-car or is it the X-car itself (not stiff enough?)

Has anyone else had problems with vibration when making short (< 0.020") cuts? The faster I set my move speeds, the worse it gets.

[stevem]

Dan, some play in the gearbox is normal (I have .011" backlash in mine).
Check your Z axis bearings and all attachment bolts. Make sure the eccentric bearings are tight and the router is tight in it’s bracket.. Check the radial runout in the router spindle with the bit you are using and with a dowel pin in place of the bit.
Some shaking is expected with very short rapid moves. Do you get chatter in straight cuts as well?

[danielb273]

Steve

Thanks for the rapid response.

I was hesitant to change anything on the Z-axis bearings as they were set at the factory, but took that plunge last night. The bearings would turn by hand but did have resistance, so I nudged the eccentric nuts in a bit so they were tighter.

I then created a test file that moved the bit down -0.050 and ran a line from 0,0 to 1,1 at 0.8 ips. Looked pretty good, but I did see some very slight chatter. Maybe I'm being overly critical, but I see no chatter when running purely in the X or Y direction. Could the steppers be causing enough "noise" to affect the cutting quality?

I've read on other threads about the unique harmonics each PRT has, so I experimented with other move speeds, but only saw either more chatter at higher speeds (> 0.8 ips) or reduced chatter at 0.4 ips.

One other thing I did is reinstall the software. That took care of some of the chatter, but didn't remove it.

I'm doing all my testing with a v-bit on MDF. Do you see slight chatter in MDF when doing something like a 1" circle? Like I said, maybe I'm being too fussy.

For what it's worth, many of the pieces I machine are small engraved signs in hardwoods, so my lower move speed limit is right about 0.8 ips.

[danielb273]

By the way, Steve, I'm really impressed with your solution (http://www.talkshopbot.com/forum/messages/7/3031.html?1077239080) to the Y-carriage hold down bearings. I'm tempted to do that myself.

Overall, have you been pleased with the cleanliness of cutting the ShopBot provides? Not so much the accuracy, but the quality.

[gerald_d]

Dan, you seem to have checked the mechanical things pretty thoroughly, and I would suggest that you spend a little time checking the PC controller. A good fast PC running few things in the background, with the best available SB software.

And my boring old repeat.....check your grounding.

Oh, and the spring tension?

[elcruisr]

One other item that can be important if all the mechanicals are right is the software tuning. Circle resolution is a "biggie" for v-bit carving. Ramping values are also important in some cases as well. Don't just rely on the default settings. Learning to tune your software is an important CNC skill no matter what machine you run.

Don't be afraid to tune those z-avis bearings. We run the heck out of our 'bot one to two shifts a day. I find I need to adjust them about every six months. They should be tight enough not to turn by hand but the axis should still move freely with the motor shut off and disconnected from the control box.

Eric

[kerrazy]

Dan are you running dos or windows control software? If you are running dos, Take the plunge and upgrade to the windows version, it is like a whole new machine, less vibration and the harmonics are tottally different. My machine like higher speeds of 1.7, which I can rarely use as I like you do a lot of v carving in wood. I would also suggest reducing you speed to .70 vhen v carving, I have found similar problems in some materials when v carving. and ensure you z speed is about 60% of you move speed.
Dale

[danielb273]

I just upgraded from SB-DOS to SB-Windows. Ran the firmware upgrade, and that did markedly improve the "noisiness" of the steppers during moves (jogs have always been smooth sounding), but the chatter remains.

I stepped through a small .SBP using [FG], and noted that during M3 steps of about 0.020 in both X and Y axes that the X-car has a noticable "jiggle" (as my wife puts it), which I'm concluding is causing the chatter. Ran this file at 0.8 ips. The machine just sounds violent when it's making steps that small. So you can imagine that when I'm doing a v-carve and making all those little steps that the X-car really takes a beating, which doesn't make me very comfortable.

Gerald, I'll check the grounding on the machine. What would you suggest for additional grounding? Regarding spring tension, should I be concerned about all three axes, or just the X/Y?

The PC I'm using isn't a speed demon, but it's a respectable 333MHz PII running W2K. I just installed the OS, so it's running as lean as I can make it (no unnecessary services, screen savers, etc).

Eric, would ramping come into play for such short steps?

Do you guys think the X-car needs stiffening? If the slop in the steppers is normal, would adding stiffeners be the next step?

Finally, if anyone has the time to do a test run in some scrap MDF, I'll gladly post the SBP file. I'm starting to think I'm the only ShopBotter on the planet seeing chatter.

Thanks for the help, guys.

[danielb273]

One other note. I'm using the PC router, and when ShopBot shipped my machine I found I had to use the plastic sleeve to get the router to even come close to fitting snugly in the base. I've got the base torqued down as good as I can, but was surprised I had to utilize the sleeve.

So, I'm wondering if perhaps the chatter I'm seeing is cumlative, i.e. X-car "jiggle", small router movement, etc.

[gerald_d]

Dan, a quick question; are both your x-motors actually driving? Or has one died and the other must drag it around?

[danielb273]

Gerald

I suspected that myself, so I dropped both motors off the racks and then did a move to check. They were both spinning. I also checked their pinions and everything was tight.

All three motors do have that little bit of "play" in them when I gently turn them one direction and then the other. It's very minute, and at first I thought my pinions might be loose, but it is the actual shafts that are moving. This little bit of play seems to me could be amplified during short, quick moves. Could I have steppers that have gone bad?

[gerald_d]

Stepper motors cannot develop internal "play" or "slop", as far as I know. Particularly not all of them going bad together, when nobody has had this experience of them going "bad". When you turn a stepper by hand, they do feel "rough" - you can feel the "magnetism" of each of the 400 steps - maybe this is the "play" that you refer to?. Our SB doesn't have gearboxes, I don't know if one should be able to feel "play" in the gearboxes supplied with some SB's

As to your earlier questions:
- For grounding see pics 12-14 in this thread (http://www.talkshopbot.com/forum/messages/2/4095.html) that shows how we did it.
- On spring tension, I am only referring to the axes with turnbuckles where the assembler must adjust the tension - x & y axes only. We tighten until we can just see the spring coils moving apart.

[elcruisr]

Dan,
I would doubt that it would have much to do with it but I wouldn't count it out as a move from stationary would invlove a ramp up to moving speed. Circle resolution will be a factor during any curved move, including a direction change. I was having some troubles with right angle moves not being smooth and it was corrected by adjusting the circle resolution of all things.

For what it's worth I've never been really happy with the v-carving results over 1" / sec. on smaller parts. Even with the Alpha upgrade good v-carving is still a little slow.

Eric

[stevem]

What do you guys consider good V carving results?

I have been dabbling with V carving lately and am getting smooth edges but rough at the bottom of the v. The cuts are much smoother with a core box bit. Results are generally better with MDF than solid wood. My cutting speed rarely goes below 2 IPS. Typical cutting depth is .040".

[danielb273]

Well, here's the "test cut" I'm using. When running at 0.8 ips, I get horrible chatter. If anyone has the time, I'd be eternally grateful for any feedback on the results you get from this. My tests were done in 1/2" MDF.

It's a small snippet of a much larger SBP, and I've done some editing to remove the rest of the cuts, but just to be safe, give it the "once over" before you actually run the file. It should cut a small v-groove when it's done.

Thanks in advance.

-Dan


ShopBot SBP File

3706 (3.9 k)

[danielb273]

Gerald

The play I refer to is the amount of movement in the stepper motor shafts before I feel any resistance. Thus, in my case, I have just a *slight* amount, but it's there. I can hear a light "click" as I run up against the resistance in both directions (clockwise and counter). I can't get to Oriental Motor's website so I'm unable to determine if my steppers are geared or not.

Hope this helps. Thanks for the pics of your grounding technique. I'll definitely be doing that.

[gerald_d]

Dan, I'm off to bed now, and will have a better look tomorrow, but your file seems to have far too many moves in it. You are moving in very, very tiny increments of around 0.001" for a lot of the time. Wonder if this has something to do with it..........

If your motor is one basic cube, then it has no gearbox. If it looks like a cube fixed to another cube, then the second cube (furthest from the wire entry) is the gearbox.

[danielb273]

Thanks for the input, Gerald.

This is a file that was generated by PartWizard. I had set my tolerances to the extreme low end in an effort to get rid of the chatter. I'll play around a bit with the file to see if increasing the tolerances reduces the chatter.

Thanks in advance for all your help. I'm looking forward to seeing how your results turn out.

[stevem]

Looks like constant velocity is turned off.

[gerald_d]

The motors in the pics in this thread (http://www.talkshopbot.com/forum/messages/22/3824.html) have gearboxes. In this thread (http://www.talkshopbot.com/forum/messages/2/4095.html) there are no gearboxes.

We changed our SB over to inches and did one "air-cut". It is a very short file time-wise and in the one second that it ran, nothing strange was noticed. So we dropped the z down for a real cut in scrap mdf. But then the file refused to run! Going back to millimeters, the file runs. In inches, no way. We weren't going to fiddle with any settings to get the SB to handle inch settings and so we stopped testing your file - sorry.


I think this is another of these cases where the ShopBot support staff should be helping you. They do read this Forum, and they watch us amateur users squirm and guess to try and help each other out, but they will only get involved if you call them directly, and then it will be in private. Snag is, the rest of us don't learn the solution, and so the squirming and guessing continues.

[danielb273]

Gerald

I appreciate all your efforts (and Steve, Eric and Dale's help) to help me resolve this. I'm with you, if your machine is working fine, don't fiddle with it!

I'm in agreement with you on the next step. I've contacted ShopBot, and hopefully this issue can get resolved before I wear a hole in my stomach.

I snapped a quick pic of my stepper motor this morning. From your last post (thanks for the pictorial help) I've also got geared steppers.

In your experience, have you ever ran across other people having problems with this model of stepper?

Thanks again for all your help, guys. I know I'm taking time out of your days, and you've been extremely generous with your advice and experience. Thank you.

-Dan


3707

[danielb273]

Steve

Good grief! You've done some impressive modifications to your bot (see Gerald's link above to gearboxes). I'm super impressed with your solution to the X car.

I'm guessing you were seeing some of the "jiggle" I've been referring to?

What concerns me the most at this point is this: are the geared stepper motors an improvement in the ShopBot design? Gerald appears to have an older PRT that had non-geared steppers, but you seem to have a newer version (based on the stop plates mounted to your racks; these were not available when I purchased mine back in 2002).

Do you have the exact same model of stepper motors that I have? Just curious.

[stevem]

Dan, the stepper motors that Shopbot uses are good quality, very reliable and quite accurate. From the look of your cut file, I would guess that your problem is the result of a software setting. Your machine is starting and stopping in tiny increments. This is what’s causing your “jiggle”. Make sure your software is NOT set to Incremental mode and you have tuned the Shopbot version of Constant Velocity.

As for the modifications to my Bot, I like to tinker. If I see something that can be improved, I improve it. You should see what’s left (or isn’t left) of my 1975 VW Beetle.

[stevem]

3708

[danielb273]

That's a hoot! It reminds me of my brother's Landcruiser, his pet project for these many years.

You're very understated. I thought I liked to tinker, too, but given your definition of tinker, I'm only tinkling. At the moment, in the wind.

Thanks, Steve. I'll double-check my software settings. Still waiting on ShopBot to get back to me on the stepper question (how much play is acceptable; could I have a bad gearbox, etc)

[gerald_d]

" I like to tinker. If I see something that can be improved, I improve it."

[danielb273]

I've upgraded to the latest Windows version of the software. I don't see anywhere in it for setting incremental mode. It COULD be my glasses, though.

[gerald_d]

You can get the gearbox backlash specs direct from Oriental Motor (http://www.orientalmotor.com/products/stp/index.htm).

[danielb273]

Yes, Gerald, you too are a Master Tinkerer (high praise in my book). Your shop makes me envious, and your ShopBot mods are ingenious!

Here I am, trying to maintain the original ShopBot design (which is excellent, I think), and you guys are saying "Let's make it better".

Great spirit. Great attitude.

Still waiting on ShopBot to get back to me. Tim did respond to my initial post via email early this morning, but my response to him has not been answered. I'm hoping he'll just call; it'd be quicker.

[danielb273]

Gerald

Since OMC doesn't manufacture my model of stepper any more, I can't find the spec'd backlash.

The only geared steppers I found backlash info on was the AlphaStep motors, which lists backlash at 3 arc min. All the other geared steppers show no backlash data (the column is in the spec, but the value is listed as ---).

I wonder if ShopBot still has this info...

[Ted Hall]

Hi Dan,

Let me clear up the motor thing a bit. The stepper motors that you have -are- geared. They are 'low-backlash' tapered hob gearheads and the motor/gearhead is made for ShopBot by Oriental. That is why you don't see it in the catalog. It is the same tapered hob that is used on some of their alphaStep line. They are very good gearboxes, and we went to them to keep the backlash minimal (though not as minimal as more expensive planetary gears would be). It is unlikely that unless there is a case of severe use and wear, that the backlash in the gearboxes has become significant enough to cause the problem.

Tim has probably emailed you already about this (and I know he has some other ideas too), but he tested one of your V-carve files here on a very old benchtop, and it cut quite smoothly. So I'm thinking that something else must be going on other than backlash in the motors ... or anything specific about the file.

A couple of questions:

When you inspect the cut made by the v-bit when it moves in the x or y direction, is the bottom of the cut sharp or flattened?

What does a diagonal look like? What does a circle look like?

Your spring tension is 3.5 turns of the turnbuckle after the spring starts to open?

When you push hard against an X car ... say 50# ... with the motor, you are certain the pinion is not slipping? (I know you've tested this, I just want to make sure we are on the same page).

Any chance you can send me or Tim a digital picture of a typical cut problem?

Thanks. Ted Hall, ShopBot Tools

[danielb273]

Hi Ted

Here are a few pics snapped this morning of the test file I sent Grant and Tim. Per your suggestion, I've also done a diagonal, plus the "old standbys", the CR and CC. For the circle, rect and diagonals, I ran at Z-0.050, and the three axes' move speeds are the same as the part file pic (x/y @ 0.8 ips, z @ 0.5 ips).

I got Tim's email and sent him the problem.log file this morning.

Since I've asked anyone reading this thread to run my test file (it's the lower left leg of the Eiffel tower in the part file I sent Grant), I figured I'd just post the pics here for a reference.

I'll continue correspondence with you via email from here. Thanks for all your help, Ted. Looking forward to hearing from you.

-Dan


3709
#1) Rect is square; distorted due to camera. Z-0.050"


3710
#2) 2" dia using CC; Z-0.050"


3711
#3) Arrows indicate passes. All cut @ Z-0.050"


3712
#4) Lower left leg of Eiffel Tower. This is what the SBP part file attached to this thread would cut.


3713
#5) Middle section of Eiffel Tower


3714
#6) Detail of Paris


3715
#7) Detail of Paris

[harold_weber]

Hey Steve M,

Could you please elaborate on your statement from July 15th:

"Looks like constant velocity is turned off"

I'm not familiar with that option.

Thanks.

[gerald_d]

Dan, if you want to check whether freeplay in your gearboxes is a factor, attach bungee cords (http://www.cargogear.com/images/HeavyDuty.jpg) to your axes, to pre-load them in one direction. Remember to have your motors switched ON while they are attached.

[danielb273]

Thank, Gerald. I'm starting to suspect (again; see one of my earlier posts) that there might be an aggregation of things that ultimately are causing the problem.

Honestly, I'm a bit hesitant to load the steppers, as I can't gauge the load with anything other than a "that's good enough" approach. Too little load would not remove the apparent oscillation, too much could cause more stress on them than I'm aware of (thus possibly causing missed steps, ozone, etc).

Tim and Grant at SB are looking into the issue, and I'm hoping they'll have the magic bullet answer sometime this week.

Guess I don't have that hutzpah great tinkerers have.

[gerald_d]

Hi Dan, those steppers are supposed to be able to drive 80 lbs. If you are nervous about it, just use your hand to push against the carriages as they are cutting - your actual cutting force is tiny.

Just for some perspective, your last 4 pics are about 1" X 1" in real life, and on my screen now they are actually 5X enlargements - is that correct? Knowing now that you have gearboxes with discernible freeplay, I have a suspicion that your cut quality is about as good as it will get...until we see pics from other guys (or SB) showing that it can be better.

[danielb273]

Gerald

The size of the images will vary with the resolution of your display adapter, as you know. The images are roughy 380x285 pixels, and optimized to 256 colors. I wanted to post larger images, but I'm limited to JPGs of no more than 25KB with a maximum dim of 400x400. There was no zoom (digital or otherwise) to enlarge the images. Quite the contrary, I had to scale the images down about 5 times to meet the maximum upload size limits of the forum.

I'm able to see movement in the X-car (motors on) of roughly .020" with a micrometer, putting as much force on it as I would on, say, the amount of force necessary to crush a pop can. Again, my concern was not to actually put so much force on the steppers to cause them to turn, but to see how much play I could get back and forth. This test is highly unscientific, but it demonstrated to me that the X-car has enough "slop" in it to possibly cause the unwanted movement during the rapid steps that occur during a part file cut.

Everyone has been great about their suggestions, and I'm left with these thoughts:

1. Stepper motors rarely (if ever) go bad. Increases in gear backlash at the steppers is extremely unusual.

2. The pinion gears wear out, or the set screws are suspect (loose, not square with the shaft). I've checked the set screws, and even went so far as to loosen them, reposition the pinion using the "wiggle-tighten-wiggle-tighten" method to assure squareness with the shaft. Then I cranked them down HARD. My definition of "hard" is when the allen wrench handle is twisted about 30 degrees CW from the head.

3. The racks and/or pinions could be worn. A visual check of the pinion and rack show no visible wear. This doesn't surprise me since I've run the machine for a total of 40-50 hours in the two years I've owned it.

4. The pinions could be out of square with the rack. The motors have never suffered any blunt trauma, and the machine is well out of any harms way.

As I was looking over my machine again this morning, one thing that I started suspecting was the pivot bolts that hold the stepper motor plates. When I assembled the machine, I tightened them down to the point that I had enough holding force to keep the pinion in the rack even with the tension spring removed, but loose enough that I can swing the motor down and out of the rack. I'd have to check the manual again, but in your experience, do you crank down these pivot bolts after the spring is tensioned?

Thanks, Gerald.

[gerald_d]

Dan, 0.02" (0.5mm) of freeplay is a lot. In our case, we have zero discernible freeplay on any motor position, except for the z (which does not have a spring tensioner). I still suggest that you run your file while applying your can-crush force and see if that isolates the problem.

[danielb273]

Gerald

I tried the can-crush force on the test file I've been running and saw slight (albeit VERY slight) improvement. I looked ridiculous trying to hold pressure on the X and Y cars simultaneously, while keeping an eye on the router at the same time. It's the close as I want to come to dancing with my Bot. It definitely leads.

So, the problem continues. I've been all over my machine, checking not only the gantry bolts, but also all the table bolts as well. Everything's cranked down.

I changed out computers tonight as well. I was running the files on a Pentium 333MHz running W2K and the latest version of SB for Windows, and moved to a Winbook laptop running W98 and the DOS version of SB. Same results. So I think I've at least eliminated the PC side of things.

No word from SB yet. We'll see what transpires tomorrow. I've got my steppers completely off the machine now; checked the pinions, and they're spot on. My wife actually had a good suggestion with "Why don't you just send those motor thingys to ShopBot and have them run them on one of their machines?"

As she so often reminds me, the wife knows everything.

[Wes Day (Unregistered Guest)]

Dan,
Did you ever find out what your problem was???

[danielb273]

My SINCEREST apologies to everyone who has followed this thread for not giving an update. You've all been extremely helpful, and I can't thank you enough.

UPDATE:
After talking with tech support at ShopBot, in an effort to help them resolve the issue and to verify that my steppers were not faulty, I sent them all X and Y stepper motors, my control box, laptop computer and associated cords. (No sense leaving anything to chance!)

They were very helpful, and what they found was that my controller was, in their words, electronically "noisey". This was causing the undesirable shudder during X-Y moves. After getting my parts back with a new controller, it's a completely different machine. I still see some chatter, but not nearly what was being produced with the old controller. What chatter remains I'm attributing to the stiffness (or lack thereof) of the X gantry. It's a long span, and without additional stiffening brackets, I'll always have some unwanted movement when v-carving small, curved areas at speeds greater than 1 ips.

So, it looks like this mystery was solved. They said my steppers were fine, and even sent me a sample cut of one of my SBP files (the one posted on this thread) using my controller, motors and laptop, done right there in Durham.

I can't say enough good things about the folks at ShopBot. Grant,in particular, was very patient and willing to work with me to get the issue solved. He's "tops" in my book.

[lto]

Dan,

I just came across this discussion by searching "chatter". I was wondering what the final conclusion was as I seem to have a similar problem especially showing in diagonal moves. This is with a new alpa which should be able to handle higher cut speeds, but to achieve a fairly chatter free finish in a diagonal cut requires slowing down to .5 ipm. I'd be interested in hearing from anyone whether a diagonal cut significantly smoother than bandsaw can be expected at a moderate 2 ips?

[lto]

Correction on my last post;

Immediatly after posting I was taken to the top of the page. There was Dan's last reply to this discussion, some how this forum get's confused at times about which end of the list to post new messages. Possibly the solution to my problem may be similar to his. I have been communicating with Shopbot, I just felt we must be missing something yet. I'll still welcome any input.

I've copied Dan's post here in order that the next reader won't miss the resolution of the matter.

"Dan Brewer"
My SINCEREST apologies to everyone who has followed this thread for not giving an update. You've all been extremely helpful, and I can't thank you enough.

UPDATE:
After talking with tech support at ShopBot, in an effort to help them resolve the issue and to verify that my steppers were not faulty, I sent them all X and Y stepper motors, my control box, laptop computer and associated cords. (No sense leaving anything to chance!)

They were very helpful, and what they found was that my controller was, in their words, electronically "noisey". This was causing the undesirable shudder during X-Y moves. After getting my parts back with a new controller, it's a completely different machine. I still see some chatter, but not nearly what was being produced with the old controller. What chatter remains I'm attributing to the stiffness (or lack thereof) of the X gantry. It's a long span, and without additional stiffening brackets, I'll always have some unwanted movement when v-carving small, curved areas at speeds greater than 1 ips.

So, it looks like this mystery was solved. They said my steppers were fine, and even sent me a sample cut of one of my SBP files (the one posted on this thread) using my controller, motors and laptop, done right there in Durham.

I can't say enough good things about the folks at ShopBot. Grant,in particular, was very patient and willing to work with me to get the issue solved. He's "tops" in my book.