Triac 200 Y axis following error

[peter watt]

We have an early Triac 200 CNC centre. Inboard rack PC has dongle on parallel port and 3 x BL30 drives. Running Fanucms software.
Recently the Z axis stopped working altogether and the Z axis motor (ML2340 A-10) has been replaced satisfactorily and we are now working again. (Our longterm solution will likely be to change all - from PC to motors)

However ever since the time of installation the machine occasionally and randomly throws up a Y axis following error.
To date this has been dealt with with a workaround involving breaking up code into smaller blocks and resuming machining. Oil pump is working and all lead screws are turn freely. Feed rates are usually low.

We have no information on the BL30 drives and indeed on the whole installation and no spares to swap in and out.
Info from any in the forum would be appreciated.

Peter Watt
Clinical Engineering
Liverpool University

[peter watt]

Further:
Obtained a BL Brushless Servo Drives User Guide from web.
Next will follow "Tuning the Drive without a positioner" as per page 29 with an external signal source and motor out of machine.
PW 4/NOV/2009

[Denford Admin]

"Tuning the Drive without a positioner" as per page 29 with an external signal source and motor out of machine.

After reading the 1st post, that's what I was going to suggest you do next.
Although nothing has changed on the Y axis, and you have change the Z motor, it woul be well worth setting the motors up to run smoothly and without any analogue drift/offset.
I guess this guide involves running the drive from a 9v battery + potentiometer box, and monitoring the tacho feedback ?
Too high or low a gain setting, or too much analogue offset when stopped, could cause the occasional positioning errors.
Getting the motor speed correct to match the commanded analogue input is also very important when it comes to linear/circular interpolation in >1 axes
I'm not familiar with this machine at all, closed loop systems are all the same... where does positional feedback come from ? Is it encoders on the motors, or a linear scale ?
If it's a scale, I would also be looking at checking the backlash in the Y axes - this would cause the motor to really struggle to get into position - especially if it is overshooting, as backing off to pull into position will be running the motor through the backlash, where the system won't see any movement from the scale.

HTH

[peter watt]

Thank you for your post.
Our machine uses ML 2340 A-10 motors. (2 phase servo with shaft encoder feedback) Photo.

Since my last post I have done the following:
I have been running the Y axis motor (out of its axis) using a function generator input to the user I/O with scope attached to link1 (velocity signal). PC control unplugged.

The function generator was set to produce a 1Hz sine/square/triangle wave of appropriate amplitude to exercise the motor and tweak controls. Sine/triangle wave input produced smooth reversing turning motions whereas square wave produced sharp direction changes allowing damping control to be optimised.

During two exercise sessions of approx 40 mins each the motor was once observed to go into "jittery spasms". These During rotations these jitters appeared to pause the motor momentarily in a certain positions with extra noises. Without stopping the exercise session these jitters stopped without any intervention after a minute or so.

My conclusion is that such irregularity would be instantly interpreted as a following error with the drive under fanuc software control.

When the machine was installed a following error problem was apparant and a motor was exchanged but the drive board not. Apparantly the problem was cured temporarily, but the symptom returned and we have been dealing with a intermittant following error since.

Next; I will remove the Z axis motor and plug it in to the Y drive and see if the problem feature still remains.
If so I could do with a spare BL30 to swap in and test.

Not sure about the effect of Tach Gain control adjustment. The User Guide says this sets the Velocity Amp sensitivity and results in adjustment of the speed of the motor for a given input voltage.
However it would seems that the encoders must have the final say about speed and position.

Comments appreciated.
Peter Watt

[Denford Admin]

If so I could do with a spare BL30 to swap in and test.

Again, not sure of your machine model, but usually the Triacs have a bigger Z motor and drive (probably because they weren't counter balanced) - I'd be aiming to swap the X motor if it's just as easy to remove?

Not sure about the effect of Tach Gain control adjustment. The User Guide says this sets the Velocity Amp sensitivity and results in adjustment of the speed of the motor for a given input voltage.
However it would seems that the encoders must have the final say about speed and position.

The encoder does have the final say, but if your velocity loop (ie the bit you are in the process of tuning now) is rubbish, then the control will struggle to get the motor to go where it wants it to.

Another thing that occured to me was whether the drive has accel/decel ramps in it ? They should be set either off, or very high so that the position-loop ramps can be applied without hindrance. You may still want to tune the drives with a ramp from the function generator as instant velocity commands make it difficult to tune the oscillation out of any system. It's been a while since I did any of this (digital drives pretty much set themselves up) but it's coming back to me now

[peter watt]

Hi
Thank you for that prompt reply!
Our Z axis has the counterbalance and all three motors are the same. Photo.
Z of course is new.

Accessible controls as follows: re pdf file on this post
Balance, Tach gain, Time Constant and Damping

Drive and board are currently back in and the workshop engineer is proceeding with caution.
I do have one question:
The User guide page 29 step 6 says "Set the velocity amplifier sensitivity by applying a signal of 20% of maximum input (2V for ±10V operation ) and adjusting the Tach Gain 20-turn pot to give 20% of maximum speed."
How would you do this?
My tach gain control is now fully CW and turning ccw increases motor speed. This must be a coarse control because it is done on three axes and each separate adjustment will be slightly different.
Anyway where do we find out what maximum speed is?

Regards
Peter Watt

[Denford Admin]

Anyway where do we find out what maximum speed is?

There is no easy answer to this, as Max speed will be determined by the maximum (ie rapid) feed of the machine. (bear in mind that the rapid rate may be increased by 120 or 150% by the feed override)
Usually the motors will be a chosen rating of say 3000 RPM and only run at maybe 2500 RPM to get 10 Metres/min rapid through the gearing to the ballscrew.
eg, 2500RPM at motor > 2.5 to 1 gearbox > 1000RPM of 10mm pitch ballscrew = 10,000mm/min
You'd have to work out what this maximum motor speed should be for the full 10V analogue command to give you the maximum traverse rate and set the Tach gain accordingly. eg, apply 20% (2Volts) command and adjust to get (2500*0.2)=500RPM at the motor (in my example that is)

[peter watt]

UPDATE: Further to the above and with the aim to fix the above on limited budget:

The Triac 200 installation has its original PC (in rack form) and the original PSU (60Watts). It has a 3.5" floppy drive.

During its life an internal hard drive was added but without uprating the PSU. This week I have temporarily rigged in a 250W PC PSU to the Triac200 PC unit (with HDD removed) and the CNC is generally behaving better.
So it appears some CNC anomalies have likely been generated as a result of poor +5V +12V -12V supply regulation.

The Y axis is will now generate a "guard door open error" with the door closed.
However the flavour of errors has altered with a more robust PSU. I regard this as a positive indication.

I have a Triac 200 Installation Manual (Heidenhain 360 - which our is not) as my guide. Not easy.
On the PC/BL30 rack top there is a 50 way screw terminal to ribbon adapter.
Question:Does anyone have any drawings which might include this? Would save me some time and effort.

Thanks
Peter Watt
Clinical Engineering
University of Liverpool

[Denford Admin]

Have added some more info to the top of this section (and moved our Triac200 installation manual here as well)
viewtopic.php?f=47&t=3125

Also some info on the BL30 drives here:
viewtopic.php?f=15&t=3126

[peter watt]

To Denford admin

Thank you for these.
Among some of the sheets I can recognise definite correspondency to our installation here.
Will keep these on file.

Since last post the have shortened some overly long ribbon cables stuffed inside rack PC and Triac 200 is now running reliably. No following errors.
After addition of extra capacity to ground on the guard door signal line, no anomalous guard door errors either.

Next: Adapt the new PC PSU to plug into the backplane by modifying the old EriPower 60W slide in PSU as a plug and then check voltage rail stability.

Thanks
Peter Watt
Clinical Engineering
University of Liverpool

[peter watt]

And another. . . .
It would appear that each remedially intended action can elicit some new reaction from this Triac 200.
Without disturbing rack PC or wiring (to the best of my knowledge) and following the same bootup sequence, Fanucms.exe now reliably hangs as follows:

MSDOS (6.2) and Fanucms.exe (run) . . . screen shows

DENFORD FANUC Servo MILLING CNC programming (v1.25)15 Nov..
IBM RS232 DRIVER (RS232) 25 March 1992
COM1 Installed
IBM EGA DRIVER (CRT) 25 Feb 1991
IBM DESKTOP PROGRAMMER (FANUC V1) DRIVER (KBD) 12 October199..
Initialising MINTSTE.

The control system has reported an error:

Comms timeout...

Suggestions sought?
Connector Com1 is wired to the side of the Triac200 (25 way D type M) and goes nowhere
Com2 cable runs toward the control pod keyboard.
LPT1 is connected to a dongle.

Thanks
Peter Watt
Clinical Engineering
University of Liverpool

PS:
It is appreciated that all systems have lifetimes and that this PC should be have already been upgraded.
It is desired to do this when funds are available.

[peter watt]

Apologies if this is getting tedious...

Have removed PC front panels (with stuffed ribbons behind) and comms timeout error disappears.
Run FANUCMS.exe successfully.

HOMEing axes (no faults) produces: X = 000.038 Y = 169.995 and Z = 299.038. I am informed these are normal values.
When HOMEing axes (with a fault) X will not go home but produces an unresettable "following error" and the X value now displays 300.000.
In this condition the lead screw can be hand turned and with each RESET press, the X values change from 300.000 up or down.
So a spurious offset is being generated.

When HOMEing axes has completed OK and the AUTO / JOG screens are used a "following error" can be produced whenever any axis is jogged and an X value of 300.000 is then displayed.

Currently now looking at hardware integrity of BL30 X axis driver board. Should I be looking at X axis encoder output?
On BL30 I see motor feedback lines running up to PL9 (20 way). Are these testpoint lines?
Comment sought. Maybe a BL30 schematic?
Thanks for any pointer.

Regards
Peter Watt
Univ of Liverpool