Insane Hydraulics
Bold and audacious blog about fluid power
I am going to share with you my "empirical takeaways" from adjusting the neutral swash-plate position of a Rexroth A4VSG500 closed-loop pump equipped with the EO2 displacement control. My experience with these pumps is limited, and, unfortunately, I didn't manage to find a "formal" zeroing technique for this control, so if you possess the know-how and find whatever I am saying ridiculously incorrect, please shoot me an email and explain what it is I am wrong about and/or what can be done better. I assure you - the small IH community will appreciate both my "shameful adjustment attempts " and your precious expertise.
As a person who's dealt with loads of different closed loop systems, and even wrote a series of articles on the very sensitive matter of correct setting of the neutral swash-plate position in common closed loop pumps, I've always considered myself to be quite proficient in this matter, that is... until I was asked to check a hydraulic system based on four (yes - freaking four!) A4VSG500 closed-loop units which exhibited the (very familiar to me) symptoms of floating zero.
In that particular case, the four pumps were driving a two-cylinder slurry pump, and the intermittent off-center condition was causing one of the cylinders to extend slowly, hit the end of travel, and then trip the high-pressure alarm set to a very conservative 300 bar. The system was able to operate normally, however, the need to reset the high-pressure alarm every time the pump stopped definitely cried for fixing.
I had some experience with 355cc A4VSG units with hydraulic proportional controls, and I must tell you right here and right now that the techniques that I describe in my "brilliant swash-plate-zeroing series" can't be applied to these pumps because they use a differential servo-cylinder that can't be by-passed to eliminate the hydraulic zeroing system and I don't even want to go into the details of the fancy mechanical feedback arrangement of these controls - I am sure that if you ever repaired an A4VSG with the HD control, you will understand. It's not complicated, but... fancy. And by the way, nothing external on this pump suggests that the servo-cylinders are unequal, which has the potential of making things "a bit more interesting" for techs who may assume that they are dealing with a common spring-centered equal-area servo-piston, and then suddenly discover that the servo-pressures are "all over the place" for some reason...
But I am getting carried away. My point is - having dealt with a large closed-loop pump that had what looked like a classic servo-cylinder design from the outside and turned out to be something entirely different made me "permanently wary" of large pumps in general and especially closed-loop pumps.
But these "runaways" were very different from the A4VSGs I knew. They were the A4VSG500-EO2 units. The EO2 is the closed-loop electronic control that sets the displacement of the pump with a 4/3 direct-operated proportional directional valve (4WRE10), driven by a VT 5035-1X amplifier card, that reads the pump's swash-plate angle with an LDVT position sensor. Sophisticated stuff! Just for the record - I never liked calling these "amplifier cards", I think that the term "control card" or something would be so much more fitting, but everyone and their grandma call these "amplifiers", so I have to stick to the convention, I guess. Check out the sheer size of these "babies" in comparison with my boot:
As I was studying what info I managed to dig out on these pumps and the EO2 controls, I found out that they do have spring-centering of the swash-plate, but it is "weak", and does not guarantee the centering of the swash-plate during high-pressure operation. Very different from the HD control, I must say. A quote from the catalog: "Spring centering of the control cylinder is standard. It is used for settings and adjustments in the unpressurized zero position, however without a defined reset during high-pressure operation. The spring centering is not a safety device." Here's the schematic of the control, and the cutaway view of the servo cylinder:
I also noticed that the electronic control was actively driving the proportional DCV even when the pump was centered. You could feel the tiny but constant DCV spool corrections when you touched it with the manual override, and also hear how the sound of the compensated pump supplying the 150 bar to the servo pressure changed when the amplifiers were activated.
This was a bit surprising because I thought that these cards would release the solenoids when the pump was in neutral, and I did check the card to make sure there was no input (there's a convenient measuring socket "W" on the front panel of the card where you can read the command value). My initial idea was that when the transducer was reading zero angle of the swash-plate, the control system was supposed to allow the mechanical centering to take over, but apparently this was not the case for these controls, which means they are actively overriding the spring centering system all the time. I suppose this has to do with the fact that the spring-centering is indeed too small and too slow for these chunky boys.
The system had separate pumps for servo-pressure supply, which allowed me to operate the displacement controls (and play with them) without engaging the two 250 kW drive motors, which was very convenient.
As you can see from the cutaway view of the servo-cylinder, the servo-piston can be left loose between the centering screws, and I did check if any of the screws were loose, and it was all OK. This was to be expected - the pumps had less than 3000 hours of operation, almost brand new by A4VSG standards - too soon for wear issues to appear.
The manual for the amplifier cards gave me the information that you could set the transducer zero point by adjusting the "Zx" potentiometer. So that was the first thing that I tried, and it worked like a charm. For two pumps. But it didn't work that great for the other two...
Let me say this again - I managed to perfectly center the first two pumps by turning the Zx potentiometer on the VT card (less than two full turns), and I didn't manage to do the same for the other two.
I did confirm that the mechanical zero was play-less and true - something that is easy to do with these pumps due to the fact that there is a convenient swash-plate angle indicator on the side of the pump which you can use for your reference. I even developed a "secret technique" for this, but this is a matter for a separate post, and if I ever come back for this system, I'll make sure to document it and publish a detailed article. But I digress, again...
So, two of the pumps responded great, one responded so-so, and on the last one I did a full 10 turns on the Zx potentiometer without seeing any result - the pump was still off-center. I checked the transducer wiring and it was OK, I even tried replacing the VT card with another one and got the same result. Very strange! My last resort was - since I didn't manage to set the zero with the pot, I could still see how the control responded to the change of the position of the LVDT feedback coil. Here's the coil with the protective cover removed (it's actually a combination of three coils):
As you can see - the three-coil assembly is jammed between two nuts, and you can adjust its position by at least a couple of mm in both directions, and this did the trick!
I have zero idea why the control system did not respond to the zero adjustment on the VT card and responded so well to the feedback coil position change (all while maintaining the precision of the displacement control). All I know is the pump is centered and has worked problem-free for several days now.
In any case - my takeaways from this "adjustment case" (for now) are:
There's still a question in the air - why did these pumps go off-center? I have a theory for that. As it turns out - the power pack was designed so that the pumps are the perfect stepping platforms to get inside the power pack for routine maintenance procedures, and by the looks of it the traducer assemblies are stepped on (and even stood on) all the time, with the cables occasionally tugged on "unintentionally but hard". I believe that even with the protective guards the repeated abuse moved the coil assemblies and caused all the trouble.