Hello Guest it is October 15, 2019, 07:58:50 AM

Author Topic: Z plunging deeper after each retraction. Not mechanical  (Read 2377 times)

0 Members and 1 Guest are viewing this topic.

Z plunging deeper after each retraction. Not mechanical
« on: July 21, 2016, 11:57:38 PM »
Hello all and thank to those who helped me with tool offsets. Machine is running great with my ATC with the exception of a new issue I discovered during a large run with many successive Z moves.

The machine specs as follows:

60x120 wood/aluminum/plastic flatbed router that, at least according to the equipment we have, is accurate to +/- 0.0006 per 2000mm,  uses C1/C2 grade linear motion components, dual 110mm frame 1.8Kw servos for X, single 110mm 1.3Kw servo for the Y, single 1Kw Z , harmonic gear reduction and very precise proximity sensors. Using an ESS, C62 R3, AC filters, dual shielded cable and RF chokes on every wire in and out of the machine. The control box is hidden under the skirt and is 1/8" steel. It contains the drivers, spindle speed control, ESS, BOB and mains contactor. The entire machine grounds to a single point and is routed to the outside of the building (it is near a back wall) to an 8' ground rod.

Now on to the issue. With each successive plunge the Z moves a bit further down. Please note - THIS IS NOT mechanical. I run two timing pulleys, one on the motor and one on the X ball screw. I have marked them both with the shaft and have checked the Z nut. All is tight and use keyways. I marked the motor shaft and put a spot of nail polish on the expected point and relation to the shaft of the motor. Did the same with the Z screw pulley and timing gear. This is the only place I am not using harmonic gear reduction as a belt to a screw done right has little to no measurable backlash.

All of those tests indicate the Z nut, pulleys, and motor are staying in sync - it is like the motor is receiving too many pulses. I have the driver set to if there is any deviance from the number of pulses relating to the amount moved the machine will throw an alarm and take you to a custom screen explaining the axis is out of alignment and to re-home the machine. This does not happen even though I know the checking system works as I "made it" miss 1 pulse intentionally and it threw an alarm.

I am pretty sure it is not noise being that I have way overdone the noise protection and the other axes move perfectly and the drivers are all in a bank.

I have tried setting the velocity and acceleration to ridiculously slow and no difference - the Z continues to get lower with each retract and plunge. It is roughly .001 per plunge but that adds up quickly on a run with 700 plunges.

Oddly I reversed the motor and ran the job in the air and the problem didn't appear on retract moves which led me back to mechanical. After a complete Z teardown I am confident this is not mechanical.

The Z is on a 20MM C2 screw with a 1" thick base plate attached to a Y that is also 1" thick aluminum . I am running 20MM sliders and again marked the screw, shaft motor and pulleys. They are all in sync just dropping too far as the program proceeds.holding a 20HP Colombo RS series ATC spindle. The VFD is not mounted in the control box and has a separate enclosure with cooling under the skirt of the machine. Even the braking resistor is enclosed so I don't thing noise is the issue.

I made a test program with simple Z retracts and plunges and the problem exists there so it is not the program I am running. My Z code would plunge into a dial indicator - pause for 3 seconds, retract and repeat. I recorded each measurement on the pauses which is how I cam to discover the issue.

I also tried another pin set for step/direction thinking the ESS or BOB may have something wrong - no joy.

I even increased the hero torque to just to the point of "humming" and backed it off just below the "hum" point (my drivers are programed via RS232 and I have leads out of the control box that make tuning quick and painless).

It simply acts like it is getting a few more pulses than it should. I have a spare ESS and c62 Rev 4 - this is the only thing I have not tried since a rewire is going to be a PITA.

Please. If anyone has experienced this let me know. I am at a complete loss here as I don't know what else to check.

Thanks in advance!

Kind regards,


Gio
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #1 on: July 22, 2016, 12:05:00 AM »
Sorry for the typos - Hero torque is supposed to be hold torque - there are many others but that was pretty off - lol
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #2 on: July 22, 2016, 09:44:09 AM »
I'd be looking at the Z motor encoder. Loose coupling or defective
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #3 on: July 22, 2016, 10:27:04 AM »
Hello again Gary!

And here I am again overlooking the obvious. Going to look now. This is true if the encoder is loose it would explain drops. Still can't figure out why retracts are dead on though.

I did reverse the pulse direction and change the driver to move opposite just to see and that didn't help either so I will go check the encoder.
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #4 on: July 22, 2016, 02:47:36 PM »
Gravity in on your side going up, and against you on the way down?
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #5 on: July 22, 2016, 04:16:41 PM »
Thank you for your reply.

I did a realtime encoder position check (after checking the encoder was secure to the shaft - in a dust free room) and checked the ABS position. The driver will give me the encoder position in real-time and it is indicating it received the number of pulse steps to place it in the position. In other words the motor is receiving too many steps. When I used the driver motor test program provided by the driver manufacturer I programed a cycle of up and down movements via RS232 and the encoder count and position were dead on. Somehow Mach/ESS is pushing too many steps is all I can figure.

I am going to try another ESS to see if this is the issue.

To give you an idea as the torque of the Z, I have a 20 HP ATC spindle and I literally stood on the Z plate and it lifted my weight (200 lbs) plus the spindle and lowered me to a precise position both up and down.

The error is accumulative. If I re-zero I can duplicate the amount it goes off almost every time.

In all the times I have run my Z test, oddly, a few times there were no extra pulses (or at least no extra movement down). This is starting to look more and more  like a pulse issue. I Just don't know why.
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #6 on: July 22, 2016, 04:20:13 PM »
Addendum - I will add that recently I woke up to find Windows 10 forced on me. I wonder if this can be a contributing factor?
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #7 on: July 22, 2016, 05:32:00 PM »
Years ago I had a programmer working for me on machine code.  Suddenly something weird would happen and I'd ask "what did you change?" and he'd always say nothing was changed.  After pressing hard I would get an answer like, well I installed the latest update of the compiler, or I just did a little housekeeping, no code changes.  Windows 10 unexpectly? Yeah I'd be real suspicious.  There are lots of programmable things that I really wish were not.  Sometimes pots and a small screw driver are a good thing.
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #8 on: July 22, 2016, 06:39:42 PM »
Update:

Replaced the ESS, reinstalled Mach3 and rebuilt my profile from scratch. So far I am 1/4 through my program without a single variance. Odd that only the Z channel was affected. More likely a connection may have developed some oxidation or something. It is VERY humid in South Texas and even stainless will rust - indoors. You breathe a glass of water a day here.

So I am not sure the ESS was bad per se but with removing components and placing them back in I may have knocked some oxidation off a connector or something. TTL is a fickle *****.

I will post back if the issue is fully resolved. This job will take a few hours to run but things are looking up.

Gary. Your insight always puts me on the tight path. I owe you a cold one should you ever be in my neck of the woods (I also go to Virginia a lot due to my defense work so my neck of the woods is rather large - lol).

I will post an update soon.

Kind regards,

Gio
Re: Z plunging deeper after each retraction. Not mechanical
« Reply #9 on: July 23, 2016, 10:17:33 AM »
Z is still a little erratic. Either too high or too low but now within +/- .001 and not accumulating. Overall not great but at least it doesn't appear to be dropping continuously.

Thinking about putting an extra 1.8Kw servo on the Z (overkill) to see if this will get rid of the final bit of fluctuation. It is the only other motor I have that will work on the drivers I have.

Again though this could just be my old friend (and sometimes enemy) Gravity. A servo will indeed hunt a little. A larger servo would be less inclined to care about gravity.

Overkill or worth a try?