Control.com - Nerds in Control

In desperate search of information, concerning the fuel pump flow control (servo-valve and RVDT feedback) for the specific model MS7001B GE Turbine, which uses the Denison Hydraulics "fuel pump"...it is really an axial piston hydraulic pump. GE PART # 185-A1728-P001...DENISON CODE AND MODEL #'S M15-20925 AND PXX1223-609-R1P-X844.

Can anyone tell me anything about the speedtronic control board that converts the .7 to 3.5 V RMS RVDT output (8.4 VAC / 3khz excitatio) to whatever signal generated by the RVDT on the pump swashplate shaft(which corresponds to fuel-rate-of-flow), which is required to be used by the control (I assume a DC voltage to be used as in input to a summing junction)?

The RVDT's are a REALLY long lead-time, and I am looking at different sensors...and need to know what the RVDT AC signal is converted to, and why, so I can look at RVDT output converters to DC.

Thanks

Brad Zuercher,

If you could describe the problem you are having it might be easier to provide some information. We also don't know what version of Speedtronic panel you have, so that is important to know, as well.

From my experience, there is no liquid fuel control "valve" on units equipped with the Dennison axial piston pump. The variable swashplate, which is controlled by the electro-hydraulic servo-valve, controls the stroke of the axial pistons, and because it's a positive displacement pump the stroke of the piston defines the flow-rate.

Also from experience, the RVDT only provides a position feedback, where 0% is zero flow ("minimum" piston stroke), and 100% position is maximum flow ("maximum" piston stroke).

You can find a lot of information about LVDTs and RVDTs on the World Wide Web. There are many videos and write-ups; I have seen some good write-ups on LVDT manufacturers' websites. GE primarily used Kavlico LVDTs, and have also used Schaevitz (sp?), and there is another site, Macro Sensors, which used to have some generic LVDT info. Some youTube videos are very good, with wave-forms that move as the LVDT core moves.

Most of the LVDTs (and RVDTs) are excited with approximately 7 VAC RMS at approx. 3 KHz, and, the RVDT/LVDT core is to be adjusted such that at "zero" stroke the output is 0.700 VAC RMS (+/-0.020 VAC RMS).

On most (but not all) LVDTs and RVDTs provided with GE-design heavy duty gas turbines are supposed to be chosen such that when the zero stroke voltage is set to 0.700 VAC RMS the LVDT output voltage at the maximum expected stroke of the device the LVDT/RVDT is attached to will not exceed 3.500 VAC RMS. This does NOT mean that for every application of LVDT/RVDT on a GE-design heavy duty gas turbine when the zero stroke voltage is set to 0.700 VAC RMS that the 100% stroke voltage will be 3.500 VAC RMS. Sometimes, it will be 2.659 VAC RMS; sometimes it will be 3.245 VAC RMS; sometimes it will be 2.317 VAC RMS.

All that 0.7-3.5 means is that the output will be linear with respect to stroke (position) when it's between 0.7 and 3.5 VAC RMS, and the maximum voltage range could be from 0.700 VAC RMS to 3.500 VAC RMS. But it doesn't mean that in every case the voltage range must be between 0.700 VAC RMS and 3.500 VAC RMS. In fact, if you try to adjust the LVDT or RVDT such that its output is from 0.700 VAC RMS at "zero" and 3.500 VAC RMS at maximum you will find that you will drive yourself very crazy very slowly. The output of any LVDT or RVDT is volts/displacement, and if the displacement (position or stroke) of the device the LVDT is attached to is less than the allowable linear stroke of the LVDT then it will never get to maximum output!

On most liquid fuel systems on GE-design heavy duty gas turbines, the liquid fuel flow-rate is measured by passive speed pick-ups mounted on the liquid fuel flow divider--and that's the PRIMARY liquid fuel flow feedback. Sometimes it's scaled in #/sec (pounds/sec); sometimes it's scaled in percent of expected maximum. GE uses the RVTD (or LVDT for LFBVs) as a stabilizing element, or inner/outer loop if you will, to stabilize flow control. (GE has even eliminated LVDTs on many of their recent LFBVs because they've found they aren't required for stable operation--but they also changed the servos when they did that, and you can't just eliminate the RVDT because the control system expects it).

Have you stroked the swashplate and used a True AC RMS voltmeter to monitor the RVDT output as the swashplate moves? If so, what did you observe? I have seen pumps where the RVDT setscrew was loose and the RVDT core didn't move when the rod attached to the swashplate mechanism moved, so the output didn't change. Tightening the setscrew solved that problem very quickly. Re-calibration wasn't even necessary in those cases, because the calibraton hadn't changed.

So, tell us a little more about the problems you are having and what kind of control system you have and if you haven't taken some measurements, take some and let us know what the results are. From the description you provided, there is something odd either in your explanation or your understanding or something.

Lastly, the accuracy of the RVDT calibration on the Dennison pump swashplate isn't critical. The primary flow feedback is from the flow divider (usually), so even if the Speedtronic thinks the swashplate is at 37.6% when it's really at 28.4% as long as the actual flow-rate feedback from the liquid fuel flow divider equals the flow-rate reference then the swashplate won't be moved. The swashplate will only be moved when the actual flow-rate feedback (from the flow divider speed pick-ups--usually) differs from the flow-rate reference--and then the Speedtronic will use the servo-valve to change the swash plate position to make the actual feedback equal to the reference, regardless of the actual position or the indicated position.

The regulator is almost always a flow-rate with position feedback (for stability) loop, and not a position loop. If it's a typical GE-design heavy duty gas turbine with Speedtronic turbine control system it will be a flow-rate with position feedback loop, and all that's necessary is for the RVDT feedback to linear with respect to position and fall within the measuring range of the Speedtronic's LVDT/RVDT input circuitry (which is usually about 0.500 VAC RMS minimum).

One more thing--the Speedtronic doesn't absolutely have to have 0.700 VAC RMS at "zero" stroke or 3.500 VAC RMS at maximum stroke. As long as the feedback is linear with respect to stroke, it could be 1.2 VAC RMS at "zero" stroke and 4.2 VAC RMS at maximum stroke, or it could be 0.5 VAC RMS at "zero" stroke and 1.20 VAC RMS at maximum stroke--as long as the feedback voltage is linear with respect to position. (Some older Speedtronic panels couldn't go quite so low (0.5 VAC RMS; and some Mark V's didn't go much lower than 0.5 VAC RMS, either), but the key is still the feedback has to be linear with respect to stroke no matter what the voltage range is.

But we need more information to be of any more help.

I recall the original 7B had no speed pickups on the flow divider. (Speed pickups on the flow divider came about when they went to the gear pump/bypass valve for heavier fuels.) I also seem to recall that on the original Mark I, VCE was input to the servo card, and the output of the servo card went to the fuel pump servo assembly, passing a signal proportional the VCE and producing 0 to 100% stroke of the pump. The pump servo assembly contained the servovalve and feedback LVDT for the swashplate position. i.e., I think the loop was closed in the pump servo assembly.

As CSA said, more info would be needed.

be_anon,

Thanks for the history lesson! I have only had the good fortune to tinker with a Mark I in a "lab" setting, and have never worked on one in the field, nor replaced one with a newer Speedtronic panel.

I've also never heard of a liquid fuel flow divider without speed pick-ups. I've seen them with "starter" motors but never without speed pick-ups. I've worked mostly on "universal" liquid fuel systems, and the only 71B I ever worked on had a Mark IV. So, I'm learning a lot from your post; thanks!

I think the original poster seems to actually be some kind of pump refurbisher/repair facility, so he may not actually have a Speedtronic of any vintage.

We'll have to wait and see if he responds with any more information.

Hopefully you'll continue to post here so we can all learn more!