Author Topic: Retrofit problems.  (Read 4690 times)

0 Members and 1 Guest are viewing this topic.

Offline cseeley

  • Active Member
  • Posts: 5
    • View Profile
Retrofit problems.
« on: August 27, 2006, 08:06:09 PM »
I am working on a Tree Knee Mill retrofit, and am having some problems ??? .

I am currently using Gecko 320’s, and the existing motors.  The existing motors I believe are 90v Baldor’s, and they have 1000 line encoders.  The x is a timing belt on a 2-1 pulley, with the encoder on the 5TPI ball screw.  The Y is connected by a timing belt on a 2-1 pulley, with the encoder on a timing belt with a 1-1 pulley to the screw.

At first setup, x and z gives me a maximum of 115 IPM, and seem to run fine.  Y does not run at all.  The gecko just faults out, and resets.  After talking with Brian Barker,   It was decided that the issue was either the encoder, or power to the encoder.  So I then connected an external 5V supply to the encoders, and got the Y working intermittently for a while, but it randomly faults out.  After looking at the encoder pulses on the scope, we found that they were very low voltage.  We took the encoder apart, and found it full of oil!  In cleaning the oil out, I think that we ruined the encoder.  So we installed a different encoder (500 Line), with better, but still not good results.

With this encoder, the gecko only faults either when first starting, or when changing directions.  This has been repeated over and over again, with very slow speeds.  I am confident, that this is not a tuning issue, as the fault occurs even at the slowest acceleration setting, and also given the same basic setup as the x, it should work with the same settings, right?

Now I have the gecko’s current limit set all the way up, so the motors are getting all the current that they can.  This mill also has handles on the screw.  I have found that I can break the “motor lock” with very little force.  The y will break free easier than the x. 

Do you folks think that my problem is that the motors are weak?

Now as I said, these motors are 90V motors.  The geckos can only take a maximum of 80v.  I have a 50V transformer, and I’m getting about 72VDC out.  Could this be the problem?  Do you lose holding torque by lowering the voltage?  I am not sure, but I would think that when the drive is locking the motor, I should not be able to rotate it enough for the gecko to fault.  It seams that evryone hear, likes gecko's, but I am wondering if we should have gone with the rutex drives?

Based on the manual, that came with this mill, with the old drives, This mill got 250ipm+ rapids, so the PC, is the limiting factor here for speed.

Any feedback, that you folks can give is greatly appreciated.

Thanks

Chris Seeley

Offline Chip

  • Global Moderator
  • *
  • Posts: 2,057
  • Gainesville Florida USA
    • View Profile
Re: Retrofit problems.
« Reply #1 on: August 28, 2006, 06:24:31 PM »
Hi, Chris

The Geco drives will trip at 125 encoder count and Rutex at 32,000 as I remember, at 125 count your off your cut line quite a bit, if your doing artwork this may not be a problem but if your machining, to have the drive not trip till 32,000 is way out there.

I use both type's and find both work well.

I read some where that Geco and Rutex use the same Driver Mosfet's, thay just rate them differently.

It's a function of volt's and amp's on your motor's torque and all motors are little different, turn up the Gain and Dampening on the Y axis, till the force to move it is the same as the X.

Inc. the volt's to 80, if you have a heavy duty var-ack boost the X-former input to test.

Swap your X-Y servo motor's may be bad.

Hope this Helps, Chip






Offline cseeley

  • Active Member
  • Posts: 5
    • View Profile
Re: Retrofit problems.
« Reply #2 on: September 06, 2006, 09:22:49 PM »
OK.....

Well after reading the response, here, and the response on the yahoo group,as well as other emails, It was decided, that the problem was the encoder!

So we ordered a new encoder.

Installed it yesterday, and got the SAME results.  This leaves either the gecko, or the motor.

Thinking the motor was the problem, we unhooked it from the gecko, unhooked the drive belt, and connected it up with an external power supply...  Guess what it runs fine!

So now we think it is the gecko.  So I swapped the x, and y motor, and encoder connections from one gecko to the other..  Now here is the problem  It seems to runs!  Both x and y are working???  Now what should have happened, is that the problem with y should have followed, either the motor, or the drive.

Tomorrow, I will swap them back, and see what is happens.


Now here is my concern, and this is no different, now than it was before...  It takes very little force to turn the motor and fault the drive.. 

Now here are my thoughts;

The gecko, is supposed to be a 20amp drive?
When the gecko is on, and there is no movement commanded, The motor should be under power going back and forth from one encoder count to another.

If you grab a hold of the motor shaft, or in my case the lead screw handle (connected to the motor with a 2:1 pulley) and try to turn the motor, then the gecko should be putting more power to the motor (proportional to the amount of error) so that it can return to it's position!  It's resistance to my turning the handle should increase!!

So here is my question...  Does the gecko do that?  Does the rutex do that?


Is the problem with the drive? or the motor?

Another point...  The gecko is limited to 80 Volts.  I was told to use 72V for safety!  The motors are supposed to get 90V.  How much torque am I losing because of the voltage drop? Does 20% less voltage= 20% less torque?

Tomorrow, I will try to measure the actual torque required to turn the motor with the gecko powering it.


Thanks in advance for any and all responses!!!


Chris

Offline Chip

  • Global Moderator
  • *
  • Posts: 2,057
  • Gainesville Florida USA
    • View Profile
Re: Retrofit problems.
« Reply #3 on: September 06, 2006, 10:43:33 PM »
Hi, Chris

Have you compaired the X Y torque with the servos off ?

Have you turned up the Gain and Dampening on the Y axis, till the force to move it is the same as the X.

Is there any slop if you hold the encoder and turn the servo back and forth.

Encoders work better if thay are on the motor direct, If your encoders are 1000 count encoders thats 4000 steps per turn, 5 turns per inch, it wouldn't take much slop  between encoder and motor to fault the drive.

In Config Motor Tuning what are the settings for both X Y,  Step's Per______Velocity______ Acceleration______ Stp., Dir. Pulse

Hope this Helps, Chip

Offline cseeley

  • Active Member
  • Posts: 5
    • View Profile
Re: Retrofit problems.
« Reply #4 on: September 07, 2006, 08:00:16 AM »
I set up x first, because of the amount of travel...

So the values that we arived at for x are 20000, 115, .5, 0, 0 (acceleration seemed way to slow, but it works right now.)

When setting up the y I started with the same values!  (same motor, same drive, why not?) NO Go!

When switchin Y to X the values that are on X are working with Y????

Also when I switched x for y, holding torque didn't change either.

Yes I did try to increase the gain and damping to get more holding torque, and it made no noticable change.

This afternoon, will try to get a torque rench, and actually measure how much it takes, and compare.

Thanks for the input!

Chris

Offline Chip

  • Global Moderator
  • *
  • Posts: 2,057
  • Gainesville Florida USA
    • View Profile
Re: Retrofit problems.
« Reply #5 on: September 07, 2006, 03:32:28 PM »
Hi, Chris

At this point you should swap the servo motors around or at least check the ohm's value between the + - motor leads with them disconnected from the control may be bad motor or brushes is it oily to.

The reason the drives are slow is the 1000 count encoder this gives you .00005 per Mach step.

If you use a 500 count encoder it will give you .0001 per Mach step and pickup allot of speed.

If your encoders are remotely mounted not mounted  on the motor, any backlash between them can cause problems.

You didn't state the values your using on the Y axis ?

Just some more input, Chip

Offline cseeley

  • Active Member
  • Posts: 5
    • View Profile
Re: Retrofit problems.
« Reply #6 on: September 08, 2006, 02:48:08 PM »
Chip,

Thats what I did...  I swapped x for y, and they both worked!!!

Ok so here is what I did last night.

I swapped them back, and guess what, now Y dosn't work again.  At this poin all that is different is the gecko tuning.  So I guess the "tune by ear" method does not work for me.

I wrote a program to feed Y back and forth at 100 IPM, I lowered the gain just a hair, and bamb. now it works.

I guess the gain is VERY touchy.

Ran it for a couple of hours without trouble, just rapiding back and fourth, doing circles ect.


Thanks for all the help.


Chris

Offline Chip

  • Global Moderator
  • *
  • Posts: 2,057
  • Gainesville Florida USA
    • View Profile
Re: Retrofit problems.
« Reply #7 on: September 08, 2006, 03:17:21 PM »
Hi, Chris

Glad you got it working.

Thank's Chip

Offline shawn_clmachine

  • Active Member
  • Posts: 10
    • View Profile
Re: Retrofit problems.
« Reply #8 on: September 18, 2006, 01:14:36 AM »
I agree, the Gain adjustment is very touchy.