This BDS4 servo driver (Kollmorgen / Cininatti Milacron) has low torque. I am 100% confident it is not the servo itself (swapped).
At rest (idle state) the servo already has low torque so I am still able to turn it (although with resistance). So it cannot be an issue with the control signal from the computer (or a software setting) as this has to be 0 volts on C1/11 - C1/1) in idle state.
Are there any experts on here who know this device inside out?
10kHz / 320Vpp arrive on the Ma,Mb,Mc terminals, PWM duty cycle is only about 38% when idle (compared to 51% on the other 2 BDS4). So the 6 IGBTs working as they should (they also measured good). It can ONLY be a timing issue! Either PWM duty cycle too low OR something is wrong with the phase-shift between the 3 signals (hard to investigate with my equipment here at the moment, only having available a 1-ch oscilloscope).
Trim pots:
STABILITY: fully CCW (min)
CURRENT LIMIT: fully CW (max)
COMMAND SCALE: fully CW (max)
BALANCE: adjusted to 0 volts across C1/11 and C1/1
BDS4-COMP3 Board (BDS4COMP3310414W Rev. A):
DIP 1A: 1,2,3,4,5,6 - all up
DIP 5A: 4,5,8,10 - down, rest up
I am not sure what this COMP3 board is for and the meaning of the settings. It's purely passive and will probably compensate some kind of reactance, possibly inductance from wires and coils (from the servo resolver?).
Someone previously did some repair works on that PCB already, so it is possible that an adjustment has to be corrected. However, I have no documentation on those 8 trim pots inside (AZ, AGAIN, CZ, CGAIN, OSCG, R/DTR, R/DG, R/DZ) and can only guess their purpose. With the exception of R/DTR the seals (locking paint) seem to be intact. Turning this pot R/DTR (powered up state) didn't make any difference, so I turned it back to it's original position.
If something is broken on the component level I might be able to find and fix the fault. Electrolytic capacitors have already been replaced recently, with the exception of the tantalum caps.
Does anyone have a schematic for the BDS4 ? That would help me a lot.
Unfortunately, this forum is primarily for end users that don't get into the level of detail you are looking for, and experts that "know this device inside out" are people who make their living by repairing them, not telling everyone else how to repair them.
Much of the available documentation on the Kollmorgen drives has already been posted in previous threads so if you search the forum you should be able to find a lot of detailed information, but again, the bulk of this information is tailored to end users/machine repair, and not targeting component level repair of the drives. There are some posts that have more detailed discussion of the drives though (https://www.cnczone.com/forums/cinci...-software.html).
TMSW, Maver1ck is correct; but you leave out many valuable pieces of info for Tech Support and Repair.
Machine Serial Number: 70xx-Yyy-xxxx
Model Type: Exx if shipped to North America or a separate plate with year of Manufacture.
Control Diagrams: 9120xxxxyxx (Wiring Diagram Number)
Is this Drive for X, Y, or Z Axis? Z-Axis would always have high torque and kW
The Drive Model number BDS4-?????
The Motor p/n 1-6xx-yyyy
There are Standard pot Settings for each Axis and Model Type: Stability, Current Limit,
Command Scale would normally be in the middle as well as balance and Parameters in the control adjusted to get the proper Balance & Gain.
The Internal Comp Board was Designed by Cincinnati Milacron Electronic Control Division; now owned by Siemens (Mason, OH).
Cincinnati ESD also made many of the internal boards and sat on the Design Board for Kollmorgen on the BDS4 and BDS5 Drives and Motors.
Only Siemens and a few Kollmorgen repair facilities such as Kilroy Corp out of Dayton had this info; many others did not
The OEM "Cincinnati Milacron ESD" now Siemens since around yr. 2000.
When you Ohmed out the motors was it from the leads at the drive end? Have you insulation tested (Meggered) the leads disconnected from the Drive end?
Yes, that's what I already thought. This forum here is more general, often only scratching the surface when it comes to *real* faults.
Thanks also for sharing the link to the other post, but I already read it before. It doesn't go into much detail either, other than replacing caps and a few resistors in the DC BUS monitoring circuitry. My fault here is different, after having replaced all electrolytic caps and fixing corroded tracks. The other posts related to the BDS4 were not even close to this low level.
@gwood
Thank you for the information on the COMP3 board and shedding some light on the history of its development.
I don't think all those numbers are relevant to my fault in any way, with the exception of the exact model which is BDS4-210-23-414W32 and main PCB number BDS4-MC4-10A Rev. G. As you can tell from that, it's a 10A model.
We physically swapped X- and Z-servos. Now the good X-servo laying on top of the Z-axis connected to the Z-BDS4 went bad whereas the bad Z-servo with the X-BDS4 went good. That's what swapping both BDS4s already indicated, but some things didn't make sense and were rather confusing, so we decided to physically swap servos to make sure we're on the right track to rule out any other possibility. Hence it cannot be the servo itself, neither a mechanical issue, neither the cable (as X-BDS4 drives Z-servo on Z-axis although it seemed too slow which could have a different reason). Also it cannot be a setting in the software as the torque is already low in idle state of the servo when no velocity signal being put out by the software via PCA1 assembly. It must be the Z-BDS4 itself. Maybe this even was the reason someone did repair works on it previously, swapping the 3 IGBT driver ICs. I did replace them again, even put in IC sockets, but that didn't fix the issue.
I measured the servo's coil resistances from the BDS4 end of course to include the cable, plug and socket as well, ruling out contact resistance, shorts and breaks. The 3 coils were 1.7 Ohm each (incl. cable), the insulation resistance of the 3 motor wires to ground (the case) were all in the region of 600 kOhm (incl. servo, not just the cable). Just now measured the insulation of the servo resolver wires (separate shielded cable) to the case (ground) which were >30 MOhm, their resistance is 18-21 Ohms, incl. cable. All looking good.
As I can not hope for any help from the experts on here, I went ahead and analyzed the BDS4 myself. Putting in 2 days of work, I eventually managed to power it up safely on my lab bench, fully enabled, in DRIVE-UP mode, no red fault LEDs, putting out the PWM signals which drive the 6 IGBTs which in turn drive the servo. In my lab I have a 4 channel oscilloscope and all kind of equipment available for troubleshooting. So I took several measurements and already figured out a lot of interesting things about it, also something that seemed a bit odd. Will soon get another known-good one to do some more measurements for side-by-side comparison which should hopefully lead me to the fault.
The BDS4 is the "Servo" and it drives the servo Motor. The resistance of 18-21 Ohms is extremely high; but it is not clear what you are describing this measurement to be.
Measuring the motor leads connected to the motor should not be more than a few ohms at most.
I was thinking that this is called a "servo amplifier" or "servo driver", whereas the mechanical part being called the "servo" or "servo motor". So I was mistaken the whole time?! English is not my first language, so I beg your pardon for any grammar or syntactical mistakes in my posts!
Please read again carefully. I said :
1st: "The 3 coils were 1.7 Ohm each (incl. cable)"
2nd: "Just now measured the insulation of the servo resolver wires (separate shielded cable) to the case (ground) which were >30 MOhm, their resistance is 18-21 Ohms, incl. cable."
I talked about the resistance of the 2 RESOLVERS (SIN & COS, 21 Ohm each, coil or hall sensor?) and the 1 excitation coil (REF, 18 Ohm).
Btw, I received the known-good one in the meantime, will connect it in the lab in a couple of days.
BDS4 - Low Torque - Schematic required (Kollmorgen/Cininatti Milacron)
Reply with Quote
