Fluctuating Fuel Flows With G Forces

Ever since day one I have had fuel flow readings that vary considerably. I have tried various transducer locations including between the carb and the engine driven fuel pump, between the electric pump and the gascolator and between the fuel selector and the electric fuel pump (the current location). I have replaced the engine monitor and the transducer (under the free exchange offer by JPI on VM 1000C units) with no improvement.
Lately I have discovered that the fluctuations are greatest in turbulence and mainly occur with the small G force changes. Today I went flying in smooth air and the readings were reasonably steady at 5.6 g/h. A mildly aggressive but brief pull on the stick momentarily decreased the reading to 3.5 g/h and a similar push on the stick shot the reading up to 13.8 g/h. The same maneuvers with the boost pump on gave similar results; from about 7.7 g/h down to 3.9 g/h and up to 19.8 g/h
Is this normal? Do others experience these fluctuations? Any suggestions would be appreciated. About the only thing I can think of is perhaps the carb float may be partially filled with fuel and the G forces are causing it to rise and fall, however the readings are the opposite that I would expect with changing G force. The engine always runs perfectly through the fluctuations and has never shown any signs of flooding or leaking fuel from the FAB drain hole.

Fin
9A
0-320, VM 1000C (new version)

Does it average out to the correct amount? I've seen that carb'd engines FF varies a great deal, but averages out fine. That's because fuel really flows pretty unevenly to the carb, but evenly out of it.

osxuser said:

Does it average out to the correct amount? I've seen that carb'd engines FF varies a great deal, but averages out fine. That's because fuel really flows pretty unevenly to the carb, but evenly out of it.

Click to expand...

I agree with the above statement. A "G" load departure from 1 will cause the carb float to over react and fuel will flow unevenly through the impeller. The total used should be accurate no matter.

A long time ago I had an auto version of fuel flow installed with a 0235 with float and it worked quite well, unless the light bulb in the impeller unit burned out.

Here is 4 minutes of data from a typical straight and level flight. You will see that the fuel flow varies between about 36 L/H (9.5 g/h) to about 13 L/H (3.4 g/h) for a constant MAP, RPM and mixture setting. This was in relatively smooth air, in any sort of turbulence the FF readout is all but useless. Looking at the graph you could say it sort of averages out around the flat line segments of about 22 L/H. Is this the sort of thing I should expect from a FF transducer on a carb?

Fin

http://i211.photobucket.com/albums/bb6/fpgath/FuelFlow.jpg

It would seem to me that the sensitivity of the system you are using, is not dampening out the float level variations as the float opens and closes. In another words, as the float chamber fills and empties the flow into the carb varies slightly and you are seeing that on the fuel flow gauge. I think, the system used for a carb'd engine needs to be less sensitive or have a lesser average reporting rate then an FI engine because the flow to the crab isn't as steady as it would be on an FI engine.
Good Luck,
Mahlon
?The opinions and information provided in this and all of my posts are hopefully helpful to you. Please use the information provided responsibly and at your own risk."

Sensitivity setting?

Hi Fin,

I wonder if the JPI guys can provide you with a sensitivity parameter to set for the FF. I have a Dynon EMS with a floscan FF sender and I don't see this variation(sender between pump and gascolator). I seem to recall in an old thread(or maybe on the dynon board) that they have a routine that damps out the temp spikes in FF. Not sure if JPI works the same way.

This might be no help, but thought I would mention it. I'm still jealous that your 9A is faster than mine......

I know for the JPI Flowscan 450, they have different senders for carbed and FI engines. Maybe you have a FI sender on a Carb'd engine?

Thanks all for your comments. I'll contact JPI and see what they suggest.

Fin