Hello Guest it is March 28, 2024, 09:01:43 PM

Author Topic: CS Labs CSMIO/P-A  (Read 51081 times)

0 Members and 1 Guest are viewing this topic.

Re: CS Labs CSMIO/P-A
« Reply #40 on: July 16, 2018, 03:56:03 PM »
Hi All

The wiring checks out correctly as we have previously established. I think you are better off using STP to pins 1 and 14 of the CSMIO rather than grounding locally as I think Craig was implying. The quasi-differential connection I recommend could reduce ground noise slightly, at least at low frequencies. The encoder

Your encoders are correctly wired including 0V and 5V. The 5V is indeed supplied by CSMIO to power the encoders. In any case you said that the encoders were working correctly when set for 10000 counts per revolution.

Your motor needs 33V per 1000 rpm, so you should be good for something approaching 3000 rpm maximum speed.

I wouldn't adjust the current limit at this point as it should already be set correctly. It would perhaps be worth applying 10V as Craig suggests and adjusting RV2 (command gain) to see if you can achieve anywhere near the desired speed.  As Craig says, initially do this with no load on the motor. I don't hold out much hope, as I believe you said you had applied over 9V using a battery and that you had adjusted the gain pot.

If you still have negative results, I again suggest the easiest way to confirm if the drive is dead would be to swap it out with another.


I tend to go along with Craig in that if the drive is defective, it would be better to ditch it and replace with ac servo motors and matched drives, but of course that will cost you a significant amount, as well as having to potentially modify the mechanical arrangements and wiring - altogether quite a bit of work.

Allan
Re: CS Labs CSMIO/P-A
« Reply #41 on: July 16, 2018, 04:08:17 PM »
Hi Allan/ Craig

Craig can you simplify the wiring suggestions from your last post please.

Done some more checks and the Voltage at the bus is 103v when under load. A drop of 5v so I presume this is ok.
The interesting thing is no what speed command is entered in mach the voltage at pins 1 and 14 (Command Hi Low) is always 10.50v. This is from a speed input of F10 to F1000.

Cheers
Mick


 

Re: CS Labs CSMIO/P-A
« Reply #42 on: July 16, 2018, 04:58:37 PM »
Mick

The CSMIO output is not simply proportional to speed, it is a function of position error.  You have such a high integral gain that this is likely to rapidly force the output to 10V unless the motor is stationary. I would suggest that for now you manually set all but the proportional gain to zero in the CSMIO manual tuning. Increase the Kp until the motor just starts to oscillate, then back  a little. Aim for the maximum Kp without oscillation.  If the motor starts to run away rather than oscillate reverse the sense of the encoder (tick box in the plugin).

I was just thinking that we are currently tending to suspect the drive as being potentially faulty, but could it be that it is the motor itself that is bad. Maybe loss of field magnetisation, for example. I am just thinking aloud that the motor would be a lot easier to test than the drive, as potentially you would only need to supply armature valtage at sufficient current. See what Craig thinks about this as a possibility.

Allan
 
Re: CS Labs CSMIO/P-A
« Reply #43 on: July 16, 2018, 05:20:36 PM »
Hi Allan

Thanks again for your time. I will try this again tomorrow. I am convinced this is a setup error as these motors and drives were working perfectly before an error occurred on the OT Control and shut the machine keyboard down. It's not a problem to swap the motors over so will try this if Adjusting the Kp doesn't work. If all fails I will send one of the drives to be tested. They offer a fixed fee repair on the drives for £150 so if they were both faulty it wouldn't be the end of the world.
Don't want to give up on them just yet. the machine has done very few hours. 

Cheers
Mick
Re: CS Labs CSMIO/P-A
« Reply #44 on: July 16, 2018, 05:34:39 PM »
Mick

I wholly understand your wish to use what you have if at all possible. I would be the same. The thing that makes us suspect a fault is simply that you have stated that the motors would not run at full speed when you applied 9V to the command imput and adjusted the gain pot. But nothing quite makes sense at this stage so we need to try to narrow things down.

At least we are now confident of your wiring and of the power supply to the bus.

I may not be around much over the next 3 days as we have visitors and other committments.

Allan
Re: CS Labs CSMIO/P-A
« Reply #45 on: July 16, 2018, 06:21:46 PM »
Hi Both

Couldn't resist having another go. Don't know why but disconnected the Tacho Wires from the drive and the Motors run at full speed. Fully controllable from speed commands in Mach. I think means the drives and Motors are working fine. Any suggestions why this might be happening? 

Cheers
Mick
Re: CS Labs CSMIO/P-A
« Reply #46 on: July 16, 2018, 06:39:06 PM »
Sounds as if the tacho gain may need adjustment, but will have to leave this to Craig as it is getting late for my brain. There is mention in the manual of adjusting the tacho to match it to the command voltage, but I'm not sure what this actually does and there is no circuit diagram. In any case, Craig is most likely to have experience in this area.

It would also be worth measuring the tacho output to see if it is within the expected limits.

Allan

Allan
Re: CS Labs CSMIO/P-A
« Reply #47 on: July 16, 2018, 07:02:42 PM »
Hi,

Quote
It would also be worth measuring the tacho output to see if it is within the expected limits.
Good idea, get the servo spinning as fast as you can, preferably with some means to measure the speed and
then measure the output of the tacho and compare to the spec.

When I get home tonight I will look a bit more closely, but what it amounts to, at the moment, with you having disconnected
the tacho, is that amp is now a 'voltage driven current output amplifier' also called a transconductance amplifier.

By disconnecting the tacho you have broken the feedback loop which turns the transconductance amp to
a  'voltage driven voltage output amplifier'. Note this feedback loop is within the drive and has nothing to do
with the PID dynamics of the CSMIO.

Much of my University experience with these devices was with a plain transconductance amp, ie torque mode.
Unless someone can tell me what advantage having the extra feedback loop, especially as it doesn't seem to work,
is I would not bother with it. That is a very preliminary assertion on my part and thorough investigation is required
before you adopt the idea.

It may be for instance by removing the tacho feedback you have effectively increased the command gain to where it should be.
Thus with the tacho loop reinstated AND the command gain increased to where it should be then the setup would
now work as the manufacturer intended.

Craig



'I enjoy sex at 73.....I live at 71 so its not too far to walk.'
Re: CS Labs CSMIO/P-A
« Reply #48 on: July 17, 2018, 01:10:39 AM »
Thanks Craig

I will remove the motor tonight and start again without the Tach connected. Once I have matched the voltage from the Tach to the Command I will try reconnecting. This couldn't be some kind of phasing issue? Eg swapping two wires on a three-phase motor makes the motor run opposite. Just thinking if the motors are turning opposite to what the controller thinks they are would the Tach be running the wrong way?

Mick
Re: CS Labs CSMIO/P-A
« Reply #49 on: July 17, 2018, 02:29:08 AM »
Hi,
yes it could be but usually what happens if the direction of rotation is inconsistent the axis will run away too one end not
just go slow.

Craig
'I enjoy sex at 73.....I live at 71 so its not too far to walk.'