PRS Backlash

[erik_f]

Anyone else noticed that your PRS has a fair amount of backlash? I have tried to get rid of any slop between the rack and pinion...but it seems there is always going to be some. I was a bit disappointed to find about .015" of play from pushing it in one direction and then pulling it in the other. What brought this measuring episode on was I made some small parts and they seemed to all be oversized by around .015". Could it be something else? Maybe part works for some reason? On the computer looking at the part they measure correctly...but they all came out oversized. Any ideas?

[scottbot]

Erik,
A while back I gave my X car a shove with the power on and was horrified by the amount of backlash.
I loosened the mounting bolts on the X motors and pulled them up tight and re-tightened the bolts.
That took care of it for me.
Hopefully that is all that is wrong with your machine.

Scott

[erik_f]

I am using a BT 48 so the process would be a little different...but I have pulled the motors tighter so the pinion is tighter in the rack. I ran a little test with not such pleasing results. I ran 2 sets of parts using conventional cutting and the other using climb cutting. There were 4 of the same parts in each run. I measured all the parts with a caliper and then averaged their measurements. Then I subtracted the climb average from the conventional average...here they are (ideally the parts would be 3.0" top, 2.75" bottom, and 2.5" long):

Conventional
2.98375 top
2.73575 bottom
2.488 long

Climb
3.03025 top
2.78075 bottom
2.52225 long

Difference
.0465 top
.045 bottom
.03425 long

Here is a picture of the part...it is actually going to be used as a end cap for some square aluminum tubing needed for the Buddy's vacuum hold down system.

8426

[GlenP]

I have a PRS Standard and do find backlash. I have tightened up all motors and pinions are tight. My Y still has some which is noticeable when cutting out small letters. My parts are perfectly square and to size every time. Only when cutting out small letters is it noticeable. Make sure you measure your cutter diameter as partworks toolpaths from the center of the router bit. If the cutter is bigger/smaller than in the software it will make your part the wrong size. If anyone has tips on getting this backlash to a minimum let us know.

[paco]

I have about 0.004-7" of play for all my three PRS motors mounted on a PRT Alpha mechanic/gantry. Beyond that is flex.

0.015" is way too much if you combine with all axis.

I wonder if the fact that the PRS motor mounting isn't spring load as the PRT was could be a source of free play if the pinion(s) isn't engaging properly along the entire length of the rack...

The PRS is surely less flexible than even the late PRT but the racks need to be pretty dam parallel with the rails to assume complete engagement...

[erik_f]

I forgot to mention in my last post that I had measured the bit to be dead nuts at .250". If you look at the Difference column of my measurement these are the number of the AVERAGE...this isn't even worst case. I do understand that part of the problem is the PC router. From what I have read those have a small amount of run out from the factory. I know that is where some of it comes from. I will call Shopbot and see where they say to go next. Right now I'm looking at an average worst case of around +/- 0.024".

[tkovacs]

I recently noticed some visible backlash as well. Adjusted the motor mounts to remove play between the pinon and that rack removed all that I could feel and see - did not take measurements though.

Needed to use a clamp on the motor mounts to pull things tight while I tightened the flange bolts.

[erik_f]

I just did mine by hand...but I don't want to over tighten things either. Ideally there should be very little change from climb to conventional cutting besides edge quality. I will try using clamps...but I was worried that might make things too tight and wear the rack and pinion.

[cnc_works]

PRT .003" to .004". It's gotta be in the gears.

Donn

[dubliner]

Erik, I had mine tightened with clamps( I thought tighter the better) but someone who knows better saw the machine run & said they were too tight, just need to be hand tight, but needs to be done by two people, or use a clamp as tight as a hand cold get it. Seems like our hand can only apply so much pressure but a clamp go go further.I did notice better running after.

[paco]

If they are too tight, they may run "rough" (the tool will shake like a tank on hard surface) and wear much faster.

[erik_f]

I have talked to Shopbot and am waiting for a call back. I will have to play around a little more...my girl friend asked if the point of the CNC was to tinker with it or actually make things since the only thing I've done with it so far is mostly made test cuts and adjustments.

[dubliner]

Is your GF related to my wife? It takes a while to get "comfortable" and making mistakes is a very good way to learn. It took me 6 months to put it together ! Other people look at your "mistakes" and dont notice them. Tinkering is a really good way to learn to run it well. If you paid for a Komo, while it would run like a top, when it started nattering at you you'd have to bring in the $1000 a day guy. So in reality this is much cheaper. OK try that one her - its been holding my wife off for a few months now :-)

[erik_f]

I talked to shopbot and they said I need to try and get the rack and pinion seated better. Also they said I might need to grind some of the slot for the motor mount if I don't have enough travel. I let you guys know how it turns out...thanks everyone.

[harryball]

I'm not familiar with the PRS, but there is no way the pinions could be slipping on the shaft? Are the shafts keyed like the 7.2:1 motors on the PRTAlpha upgrade?

/RB

[Gary Campbell]

Eric...
I also ran into some minor offsized parts and poor cut quality due to chattering. Using a dial indicator on a drill rod and found that the "slop" was in most directions. I was able to cure all my machines woes by adjusting and tightening wheels, rollers, pinions,motors, etc., just basically tuning up the the OEM specs. A hundred hour service, if you will. Many of these metal to metal surfaces need to mate or seat in, and thats where my problem was.
Gary

[tkovacs]

FWIW - I used wood clamps with rubber protectors so hopefully with the give of the rubber it did not get too tight. I could not apply enough hand pressure and tighten the bolts at the same time.

[expressm]

I had a similar problem with the x axis on a brand new machine. I talked to tech support without resolution. My machine already had a slot ground out of the frame to allow for extra adjustment of the motor/pinion mount. It was adjusted all the way to its limit and still the gear and track did not mesh properly. I installed a solid 1/8" aluminum shim between the table and the rack. This corrected the problem and put the motor/pinion adjustment back to where it should have been originally. I now have plenty of adjustment room in both directions and the backlash is gone.

[erik_f]

I ran some more parts last night after tightening the y motor up as well as trying to get the Y motor to have zero slop(I used Terry's clamp idea)...There was plenty of travel for the Y motor...but for X I am going to have to take a look at the motor mount and how it its...like brian said. At first the motor mount was hitting in the front where the threaded rod comes in...but now, after I ground that area, the back of the mount is hitting, not allowing me to get things tight enough. If the mounting hole was only up another 1/8 to 1/4" I think I would have things running quite nice. I ran a second batch of parts last night and was able to get acceptable results using conventional cutting...but when using climb cutting parts are still measuring up to .033 off the design to large. I am going to try a brand new bit tonight to see that will help things along.
Brian...I was thinking along those lines last night...I don't think there is enough room to drill a new hole in the same location to move the motor mount up...and moving it to a new location would limit x travel in one direction .Your solution maybe the only one...unless I can get shopbot to send me a blank cross member. I'm really trying to contain my anger about all of this. My BT48 certainly hasn't been my buddy so far...and ready to run is pretty far from the truth in my case. Its frustrating to find mistake after mistake in construction/engineering on a brand new machine. Adjusting is one thing...modifying a brand new machine so it will work properly is obseard. Thanks for everyones input on this. I just want to get the most out of this over stuffed Tonka toy!

[paco]

Just a thought on adjustment slots; the motor's shaft is off center, couldn't it be turn to accommodate a better reach?

[expressm]

Erik, I shared some of the same fustration. The X-backlash is the only problem I have had so far but wondered how it ever made it out of manufacturing with the backlash. I also wondered why Shopbot would grind into the frame for more adjustment instead of researching the real problem and fixing. I was also very disapointed in the documentation. There really is very little documentation on the buddy. The documentation is geared toward the larger machines. I am very thankful for this forum and it has helped me get going.

Back to the backlash. The 1/8" shim was the best solution for me. This put everything in alignment and the small channel that was ground into the frame is not necessary at all with plenty of room for adjustment both directions. I just picked up a 1/8" thick aluminum strip at Menards. I cut it to length to match the table, drilled holes so the mounting screws could pass through and installed. Let me know if I help in any way. Good luck.

[erik_f]

I didn't know that Paco...I will see the orientation of the motor and maybe i can rotate it for a closer fit.

[erik_f]

I was able to tighten the pinions a bit and with the advice of Brady slow things down and make more passes to reduse deflection and chip load. With conventional cutting I was within a few thousanths of my cad drawing. I guess it makes sense now looking back at it but even with the new prs it seems you need to be pretty conservitive when shooting for max accurecy along with using mostly conventinal cutting method. Excuse spelling and grammer as I'm sending this from my phone