AB 1756-HYD02 Question

We have a new hydraulic system that is regenerative.. I have never been exposed to a system quite like this.. We are using a 1756-HYD02 card, Balluff PWM Xducers and Bosch Rexroth Directional Prop Valves and I am seeing a strange scenario that I can't explain... When we went to startup the system, I do what I normally do with the 1756-HYD02 card settings.. I turn the P band down and then steadily increase it to get movement.. On this setup, as soon as we enabled the axis card, we started oscillating.. we measured the voltage and we were seeing .2 etc.. volts going back and forth.. So, I started using the MDO command and set the speed to .5VDC to valve.. The speed was really fast but the cylinder moved with no oscillation... Based on seeing that speed, I can see where the voltage I seen on the valve would cause oscillation...

Is there a way to get rid of that oscillation or voltage while it is just enabled? I guess on other hyd systems I have done, I never seen this because maybe that small of a voltage was nulled out... Should the card output any voltage just setting enabled or should it just stay at 0? Also, do I have any options if .5V is too fast for the application? I did see where we have directional scaling but according to the hyd guy, the ration was 1.14 and we had it at 1.. Not sure about the friction deadband settings for oscillation..

Thanks,

plcengineer said:

We have a new hydraulic system that is regenerative..

Click to expand...

Regen is usually not good for servo control. The natural frequency drops by about 30% when using regen.

On this setup, as soon as we enabled the axis card, we started oscillating..

Click to expand...

You must of stopped oscillating to write this.

we measured the voltage and we were seeing .2 etc.. volts going back and forth.

Click to expand...

Did you have the integrator on? Does the valve have a dead band?

. So, I started using the MDO command and set the speed to .5VDC to valve.. The speed was really fast but the cylinder moved with no oscillation... Based on seeing that speed, I can see where the voltage I seen on the valve would cause oscillation...

Click to expand...

How fast is "really" fast? You can make trends.

Is there a way to get rid of that oscillation or voltage while it is just enabled?

Click to expand...

Yes but I need to know more.

I guess on other hyd systems I have done, I never seen this because maybe that small of a voltage was nulled out... Should the card output any voltage just setting enabled or should it just stay at 0?

Click to expand...

If the error is zero the output shout be 0.

Also, do I have any options if .5V is too fast for the application?

Click to expand...

Get the hydraulic designer to remove the regen but that will only help when extending. The valve is probably too big.

I did see where we have directional scaling but according to the hyd guy, the ration was 1.14 and we had it at 1.. Not sure about the friction deadband settings for oscillation..

Click to expand...

How did the hydraulic guy determine it was 1.14 to one?

I really need to know about the cylinder parameters, the cylinder orientation, valve size and the load.

Remove any hose between the valve and cylinder. Hose reduces the natural frequency even more.

Make a trend that records the target and actual position and the control output.

Peter, Thank you for the response.. I will gather some information.. I do have a question that i am wandering about.. As soon as we enable the card, we aren't giving any move commmands or trying to jog, we see the oscillation. So, let's just say I wanted to move the axis with a MAJ Jog Function, is the loop still closed for position.. I was thinking that the MAJ function wouldn't be closed loop but the MAM position commands would be.. Like we discussed, as soon as the axis was enabled, we saw oscillation.. I agree about the regenerative because from what I can see, it doesn't seem to work well and it seems it changes the dynamics of the system.

When the system was enabled, we measured voltage of -.2 to +.2, -.35, +.1 etc... This was measured during oscillation with no moves MAM or MAJ.. Just setting enabled.. So, I am guessing that maybe we were seeing error and it was trying to correct.. Could it be that a cylinder is moving slightly and we are trying to correct the error with the voltage signals?

I think that the cylinders are a 2" Bore and they are mounted in a horizontal plane.. I will get the exact sizes..

I think I had integrator hold enabled.. .5 to 1 P band and nothing else.. According to Rexroth, they were saying this valve didn't need to be setup.. Every system that I have done in past had to be nulled and the amp adjusted to get correct pressures etc for me to control.. Not sure about this valve..

It is 4WRAEB6E24-1X/G24N9DK26/MR Valve..

Hyd Power Pack Info..

ASSEMBLY NOTES:
1. MAXIMUM SYSTEM PRESSURE: 3450 PSI. (238 BAR).
2. WORKING PRESSURE: 3000 PSI. (207 BAR).
3. PUMP: AA10VSO100DR/31R-PKC62N00 (destroked: 73.0 cc, 31.05 GPM
Actual Flow)
4. HYDRAULIC FLUID: ISO VG46 (Mineral Oil) @ 40 °C.
5. HYDRAULIC FLUID ISO CLEANLINESS LEVEL TARGET: ISO 17/15/12
6. APPROXIMATE WEIGHT DRY: 647.23 kg. (1426.9 lbs).
7. PAINT SPECIFICATIONS:
a) CLEAR COAT ANYTHING NOT HAVING AN EXISTING COATING.
b) ZINC COATED COMPONENTS TO BE LEFT AS IS.
8. WARNING - PER ISO4413 IF THIS SYSTEM UTILIZES AN
ACCUMULATOR, DRAIN OIL TO TANK BEFORE COMMENCING WITH
ANY MAINTENANCE.​

9. CLEANOUT COVER BOLT TO BE TORQUED TO 15 ft/lb

I will try to get more information. I can run some trends on speeds when I get onsite next week... Thanks Again...

Rant, never trust the hydraulic guys!!!

You should use the 4WRTE series from Bosch Rexroth. I have one or two.
4WRAE valves are not suitable for hydraulic servo control.
The hydraulic 'designer' is trying to save a buck at your expense by using a cheaper inferior valve. He doesn't care how long it takes you to make the system run.

[RANT]
Many thousands of HYD02s have been installed.
Why would anybody think there is a problem with the HYD02 instead of the one of a kind hydraulic 'design'?
[/RANT]

Peter, You are correct.. The electrical engineers always end up solving mechanical, electrical and hydraulic issues..hahaha..

Yes, I have installed 4 HYD02 cards and I never had any trouble like this.. Like I said earlier, I always tune very soft, then slightly increase bands to get movement.. Then, I trend speed/position command/feedback for tuning.. No issues at all.. As soon as we turned power on this thing, it started oscillating...

Can you tell me if the enabled axis still is trying to keep zero position error all the time? I am thinking that the valve may be drifting so when the axis is enabled, it is trying to correct for the drift causing oscillation.. I wasn't sure if the card (once enabled) would self correct for position error without any commands.. if the card always corrects for error once enabled, then I understand it a little better now.. I just wasn't sure why the card would be putting out voltage when it was just enabled..let me know your thoughts..

Thanks,

Just chiming in on your question about loop closure...
When you execute the MSO instruction, the loop is closed and servo action is enabled. At the same time, command position is set equal to actual position, so no movement is commanded. Once the loop is closed, the module will attempt to correct any un-commanded movement -- which may be due to vibration or drift in the valve controller or other external forces. If you have a non-zero integral term, oscillation will be exacerbated.
Also, the MAJ is closed loop - it simply changes the command position at a specified rate.

plcengineer, what are you going to do? It isn't a HYD02 problem. You need a system that is controllable and you don't have one yet.

Peter, Yes, they hydraulic designer has admitted that he didn't think his design was sound.. The main problem that I found with the valve setup was the fact that I couldn't maintain stability at low speeds.. We have to face a part and we need really slow speeds and no matter how I tuned the controller, we would oscillate at slow speeds.. or,if we could run slow speeds with no oscillation, we would get position errors because we weren't stiff enough to keep position tolerances.. We have one station that just does boring with a drill and we can run faster speeds just fine with the previous valve.. As long as we keep the speeds above 1 in/sec, I can bore with accurate speeds and positions.. So, they are proposing this rexroth valve for a new setup... Rexroth Part # 4WREE6V1-16-2X/G24K31/A1V or Rexroth Part # 4WREE6V08-2X/G24K31/A1V.. I see that these valves have spool feedback and I was wondering if this is a closed loop system within the valve.. I know I have done some valves that had spool feedback but we just used it for monitoring and not control.. Does this valve close the loop in internally? Do you think that one of these two valves will work with our setup? Thanks for all of your expertise.. I really appreciate the input..

Peter, Also, we took out the regen plates because that was a nightmare.. So, we are not running regen now.. They should have never had regen on this setup..

Hold out for the 4WRTE valve.

Also, at slow speeds your system can suffer from a slip/stick action if the natural frequency is too low and the ratio of the static friction to the dynamic friction is too high.