Karl,
It sounds like runaway, so have you tried inverting the motor leads?
I've setup three machines with Galil cards, AMC servo amps, and servos in the past. Last spring, I did a CNC mill and the X,Y,Z went fine and the 4rth axis failed. I didn't have time so I just got the mill running without the 4rth axis.
OK, I'm thinking bad component. So I got another servo, installed a BEI encoder, another AMC servo amp (25AE8), made up a small wire connector to go from pins 2,4,11 on the servo amp to ground, MOCMD_, and AMPEN_ on the Galil card.
Then I started swapping out components one at a time. No joy. So then I connected all the new componets to axis G (unused but available). No joy.
I just called it a day.
In WSDK, the failure has always been the same. It passes the hand spin encoder test. Then when I reset the Estop to power the servo it starts spinning and often trips the servo amp out.
What am I missing here?
karl
Karl,
It sounds like runaway, so have you tried inverting the motor leads?
First you get good, then you get fast. Then grouchiness sets in.
(Note: The opinions expressed in this post are my own and are not necessarily those of CNCzone and its management)
There are Galil commands to reverse both commanded direction and encoder feedback phasing, without reversing wires.
Try the MT command first. Either 1 or -1.
CE is one of the other ones.
Al.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
I'll try these suggestions tommorrow, but i don't think that's it. Step 2 in WSDK under "getting started" is "check motor polarity" . I've had this wrong on other motors; WSDK does this check and tells you to switch leads.
I'll try changed MT from 1 to -1 also.
Karl
If the drive works open loop, i.e. with a ±10vdc from a battery box, then usually the MT will do it, make sure you set the appropriate axis.
I have never figured out why some of the advice in the Galil manuals are a bit contradictory and do not mention phasing the encoder/motor by command.
Al.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
Further to the MT CE commands.
The Galil manuals were written with DC tachless servo motor WRT the set up procedure, this was relatively easy to just change the two armature conductors.
Now with new installations, the use of DCBL and AC servo's are prevalent.
If the motor runs away when hooked up, rather than having to swap two of the three stator wires and then making sure the commutation is corrected also, it is much easier to change the MT1 to MT-1.
Now if no runaway occurs, it is possible the count direction is wrong, so instead of changing the motor wires back and swapping the A & /A to change encoder direction.
All that is required is to change the MT-1 back to MT1 and change the encoder direction with the CE command.
It is advisable to burn the parameters when done, otherwise the parameters must be loaded by program command immediately at start up.
Al.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
More trials today. No joy. See if anything rings a bell.
MT ?,?,?,?,?,?,?,? gives all 1.0
CE ?,?,?,?,?,?,?,? gives all 0
I tried reversing motor leads, same response motor turns about 3 - 5 revs and kicks the servo out. repeated trial with other servo and other servo drive, same result.
So, open loop trials. I got a 500 ohm pot; put one lead on +5V, one on common, and the middle to the pin 4 on the servo amp. Turned pot to 0. Reset Estop, and I get same result as above!
So I disconnect +5 volt from pot, issue SH, and Reset the Estop. Servo does nothing, green light stays on. Touch +5 back to pot, nothing happens. Just barely, and I mean barely, turn the pot; servo goes about three turns and kick out. repeated trial with a different drive, same result.
I'm using an AMC analog servo amp 25A8E and the original servo on the 4rth axis is by reliance electric brushed dc servo model E660 rated at 1/3 hp. My trial motor is slightly larger, 1/2 hp, model E690.
Suggestions for further trials?
Karl
I just had another thought. For my open loop trials, I still had the encoders and AMPEN_ connected. The KO was due to the encoder moving and galil dropped AMPEN_
So, I just made a total open loop trial. No encoder, no AMPEN_. Motor turns slowly with nothing connected. Speeds up and slows down as I vary the speed pot from 0 - 5 volt.
Any more ideas?
I'm contemplating that last trial..
Could this be happening? With normal complete hookup, the servo starts turning even though its not commanded to (0 volts on MOCMD pin) so Galil KOs AMPEN when the servo is out of position?
If this is true, how do I stop motor rotation when servo is first powerred?
Karl
You need to get the drive and motor work together first before connecting to Galil.
I would either set the drive for Voltage Mode and use a pot with slider to +ref and the outer ends to +5 -5v, or set to voltage mode and use the offset switch to on, this allows the offset pot to be used to control the motor, reset the balance afterwards.
You could also test it with a dummy low resistive load and see if the voltage gradually increases when the input is increased, use a resistor value that limits the current to the drive max.
Al.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
I must have cross posted.
If the motor turns when the input is zero, it means the balance pot is mis-adjusted, it should allow adjustment to zero rpm (null).
If you want to stop rotation on power up before anything is initialized, you can issue a MO or SP0 for initial commands.
If it runs away with a MO there is something else amiss, such as wiring problem.
Al.
CNC, Mechatronics Integration and Custom Machine Design (Skype Avail).
“Logic will get you from A to B. Imagination will take you everywhere.”
Albert E.
OK, I'm reading the manual (when all else fails,RTM)
I may be miswiring here. I've got a common from pin17 (ground) on the galil 1700 to pin 2, signal ground, on the AMC servo amp. Another wire from MOCMDW(pin23) to pin4(+refin). The block diagram in section G engineering reference shows a wire from pin5(-refin) to the controller. I don't have this wire; where should I connect on the Galil 1700. (FWIW, I've got five drives working without this)
I'll keep reading, I'm sure to find more issues.
Karl
P.S. Here's a manual section on this. I think I should just tie pin5 (ref-) and pin2(signal gound) together. Correct?
P.P.S. I just found the wiring instruction in Galil 1700.pdf manual page 30 shows the wiring exactly as i have it.
3.6 Reference Input Wiring
Use of a twisted, shielded pair for the input reference wires is recommended. If the reference source can float (remain
ungrounded), connect the cable shield to both the reference source common and the amplifier's signal ground. It is
recommended that the input be connected directly to the amplifier's differential input (if applicable). Connect the reference
source "+" to "+ref” input, and the reference source "-" (or common) to "-ref” input. If the reference source ground and the
amplifier power ground are connected to the master chassis ground, leave the source end of the shield unconnected.
The servo amplifier's reference input circuit will attenuate the common mode voltage between signal source and amplifier
power grounds. In case of a single ended reference signal, connect the command signal to “+ref” and connect the
command return and “-ref” to the signal ground
Last edited by Karl_T; 11-19-2008 at 10:36 AM.