Another DYN4 and Acorn Losing Steps Issue

[slodat]

I am having issues with what appears to be DYN4/Acorn dropping steps, similar to this thread http://centroidcncforum.com/viewtopic.php?f=60&t=2610

I have everything installed per DMM's schematic on the supply side- noise filter on motor power, individual noise filters on each drive's logic supply, line reactor, etc. AC supply is segregated from encoder and DB25 inputs to DYN4. I purchased DMM's DB25 cables. I read a ton of threads on here and feel like I did what I was supposed to. Alas, something isn't right.

Walkaround video that shows the issue I'm having is below. Things work well at 400ipm and under. When I try to use 800 ipm rapids, that's where we get into trouble.



This video shows the issue on the Y axis, resulting in both Y axis drives faulting out.

[slodat]

Did some more testing. Moving X axis from -4 to +59 and back. First at 250ipm, then 200ipm, then 100.

Encoder counts:

Losing steps.. Suggestions?

[Richards]

I was hoping someone else would have a suggestion based on their machine and their experience. Here's what I would do. Because the video showed that slow moves were repeatable, let's rule out connections and mechanicals for now. You also said that you changed ramp/acceleration time from 0.5 seconds to 1.0 second. Try changing that to 2.0 seconds, not as a solution, but to see if any part of acceleration is causing the problem. Assuming that acceleration is not the prime factor, I would change each motor setting parameter ONE AT A TIME to half its value and then to twice its value. Run at least two tests at each setting. It's my guess that at least one parameter needs changing. Unfortunately, I'm not at my test bench, so I can't suggest any specifics. Whenever I'm troubleshooting electronics or mechanicals, I make big enough changes to clearly see the effect, limiting those changes only to keep within the safe operating parameters of the device under test.

If one parameter change gives stable results, then change that parameter up and down in finer increments until you know the "tipping point". Even if no parameter change solves the problem, you'll at least be confident that you systematically worked through the possibilities.

[cut2cut]

What is the fault "code" you are seeing ( from your second video ) ?

[Ken Rychlik]

To me, it is unusual to have the drives in the pc cabinet away from the machine. I'm using steppers, but the only thing going to my pc cabinet is the ethernet cable. I have been around quite a few machines and have never seen one set up like yours. The control box with the drivers ect... has always been on the machine.

I'm not really sure if the cable length to and from the motors matters, but it seems extreme to me.

From your machine video it seems like one of the import machines which will have diagonal gear racks. When powered up one way it will try to pull the spur gear one direction and vice versa. If the gear is being pulled back to far when you are jogging home, the gear itself could be rubbing against something when pulling faster/harder. Try to loosen the spur gear and give it a little more clearance??

On the long axis return, one motor is dropping out first causing the "racking" affect. Switch motors and see if the problem follows or reverses maybe.

Ken

[slodat]

I didn't get the fault code on that one. I don't want to continue to slam the gantry around to recreate the fault because it causes unnecessary stress on it. It doesn't fault out at lower feed rates. The X and Y axis had no problem at 800ipm rapids with the steppers I removed, so I'm not inclined to think it's a mechanical issue at this point. Things seem to engage and move smoothly. None of the rack/pinion was touched when I installed the DMM servos.

In my testing I am showing, as others have, that steps are being lost between Acorn and DYN4. After reading the other threads, and several offline conversations, it seems quite clear that there is an issue with the DYN4 drive and higher feed rates from Acorn using DIY and DMM's own purpose built cable. It seems there are two working solutions - use the DB25 connections for step and direction with the DMM DB25 cables I have, or use the DB25 with differential drivers.

[martyscncgarage]

Arrow controls did something pretty cool. It requires you to make a DB9 to go from Acorn spindle encoder port to DYN4 encoder port
Have a watch:

https://www.facebook.com/17584227911821 ... 890516984/

[slodat]

I saw that video on fb last week. In my video I show that the encoder counts are off. It’s losing steps.

[cnckeith]

hello. thanks for posting the videos and the report.zip

i took a look at the report and first thing i noticed is that the steps per rev don't make much sense.

go back to basics and follow the dyn4 setup guide, test and if you are still having issues post a fresh report.zip

DYN4 Acorn setup draft rev2.pdf

(665.63 KiB) Downloaded 160 times

also, for background reading.."don't do it like mach did it"
viewtopic.php?f=63&t=1801

[slodat]

The steps/rev are based on a GEAR setting of 4096 (x 4 = 16384) on the X and Y and 2048 (x 4 = 8192) on the Z. I did the basics, followed the setup guide provided by DMM, including the corrections I and others had to sort out. The issue I'm having isn't because of steps/rev or the "basics". There's something going on between Acorn and DYN4 when the axis speeds start getting higher. I felt like my first video demonstrated this when the X minus move doesn't return to the commanded location.

DMM's tech support asked me to test movement from within the DMM software to see if the problem is in the drive/motor or Acorn/drive. The attached video shows the axis moving much faster than the speed I was testing at when this problem appeared. I believe this demonstrates the problem is between Acorn and the drive. Other threads on here seem to have consensus that using the 5v signals on the DB25 solve these issues.

Anyway.. here's a video showing the axis moving quite nicely at very good speed, in both directions, without issue.