What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Does your fuel flow fluctuate?

pa38112

Well Known Member
I am getting ready to run down an issue with my fuel flow reading jumping around, but want to verify that this is not a normal condition first.
I have a carbureted O-360-A1A. Fuel flow is measured with a red cube and an MGL FF-1 fuel computer.
At any given power or lean setting fuel flow reading fluctuates every second or so by as much as 1.2gph. I have always assumed this was because the carburetor float seats and unseats causing the flow to the carburetor bowl to fluctuate, but I have never heard anyone else mention difficulty in doing a GAMI test or setting a specific fuel flow on their carbureted engine. Am I the only one with this problem?

The first thing I want to do is verify the sensor stability. I plan on doing this by disconnecting the fuel line and pressurizing the wing tank to 25" water so that there is a very consistent flow of fuel through the red cube and into a fuel can. With a steady regulated 25" pressure the entire fuel tank should flow at a very steady rate.

If the sensor is stable then I will move to replacing the float, needle valve, and seat. That is a little expensive, so please let me know if this condition is an issue or is normal...
 
yep, it does

I've got exactly the same engine/carb setup that you've got. I've also got a floscan transducer feeding a GRT EIS and it's given fluctuating fuel flows for 14 years, pretty much exactly as you describe. The totalizer is always very close so I haven't worried about it.

If I remember correctly, floscan actually recommended a damper in the shape of a T upstream from the transducer to help with such fluctuations. Matronics still shows a variation of that for sale.
http://www.matronics.com/fuelchec/damper.html
 
If the sensor is stable then I will move to replacing the float, needle valve, and seat. That is a little expensive, so please let me know if this condition is an issue or is normal...

No harm in checking that the sensor is stable, but tearing into an otherwise healthy running carburetor based on the red cube is pretty extreme.

My red cube has been doing the same thing as yours for the entire 4 (almost) years and 325 hours that my plane's been flying. A total of three different carburetors have been on my plane through an engine swap to the O-340 stroker and ECI sending new carbs in an attempt to even out the mixture distribution. The red cube fluctuations persist. It's not really an issue operationally, and fuel totals are always exactly where I expect them to be at the pump.

There are historical reports of transponders affecting red cube fuel flow readings, especially if the transponder antenna is near the firewall as it is in many installations. Some folks have wrapped their red cubes in foil or tried other methods of electromagnetic shielding, likely with mixed results. Personally, I've never attempted to shield mine.

Anyway, the moral is, tearing into a working carb adds risk, and there may be many other environmental factors at play.
 
Last edited:
Same here. Also believe it's related to the carb. Very rhythmic oscillation when plotted at a given power setting. Carb and fuel pump both new/overhauled, and performing perfectly.

Chris
 
Last edited:
Where is the red cube located in the system? Upstream or downstream of the engine driven pump?
 
Where is the red cube located in the system? Upstream or downstream of the engine driven pump?

My cube is upstream of the engine driven pump. It is mounted at my left foot. There is about 5' of line and a glascolator between the red cube and the fuel pump.
As other have stated, the totalizer is always very accurate, but it is very annoying to not be able to set a specific fuel flow, or measure the fuel flow for each cylinder to peak.
 
Last edited:
The carb float turns fuel flow on and off. As the float bowl fuel level goes down, the float opens a valve and refills the float bowl. This may be what you are seeing
 
this doesn't solve anything but it is interesting........i have a red cube on my bing carb on a jab 3300 and it is rock steady.
 
When I had a carb on my 320, the flow reading was quite steady from the red cube.

At cruise power settings, your float is not really going up and down, starting and stopping fuel flow. It usually stabilizes at a point that produces the correct flow through the needle/seat to match the flow through the main jet.

Larry
 
varying fuel flow

If the red cube is upstream of the pump it would be normal to have fluctuating fuel flow. The engine driven fuel pump has an intermittent draw, in that the cam operated lever is pushed against the diaphragm spring to recharge the pumping chamber, and then the spring slowly pushes the fuel out of the pumping chamber towards the carburetor. During the fuel pump discharge phase no fuel passes through the red cube. To create a stable fuel flow signal, place the red cube between the fuel pump and carburetor.
 
If the red cube is upstream of the pump it would be normal to have fluctuating fuel flow. The engine driven fuel pump has an intermittent draw, in that the cam operated lever is pushed against the diaphragm spring to recharge the pumping chamber, and then the spring slowly pushes the fuel out of the pumping chamber towards the carburetor. During the fuel pump discharge phase no fuel passes through the red cube. To create a stable fuel flow signal, place the red cube between the fuel pump and carburetor.

Just had the same thought. I wonder why it seems to occur more to those of us running carbs - the fuel pumps are essentially the same besides the spring I thought.

Chris
 
Another one with this issue on my 0-320. Over the years I have tried just about everything. The best solution was installing the FC-10 Fuel Computer as it has a selectable electronic dampening function which most fuel computers do not have. This makes the flow reading more stable (not perfect) but slower to react.

For my aircraft the actual cause of the fluctatuating reading are changing G forces caused by turbulence. In smooth air the readings are always stable.
YMMV

Fin
9A
Australia.
 
red cube

On my O-320 the red cube is between the pumps and the carb. If the flow drops below about 6 gpm the numbers will walk up and down about a gal. If I run it so the flow is above 7 it gets a lot steadier. I think it is the float bowl filling also.
 
Reviving this thread as my newly installed Red Cube exhibits more or less the same symptoms as the OP pa38112, using the same engine configuration. The fluctuations are over the range of about 1 USG, don't seem regular, and do not really jump around, but rather are constantly changing.
  • Was anybody with the same symptoms able to solve the problem, how?
  • pa38112, if you read this were you able to do something about it, or are you just living with it?
  • Has anyone installed the Matronics dampener mentioned in post #3, results?

Thanks for any response. FWIW, I have no leak in neither air, nor fuel nor on the carburettor shafts (o'hauled recently). Following an avionics upgrade I had to replace my non-compatible and ageing Flowscan 264 with the red cube.
The red cube was again installed free hanging downstream of the engine driven pump.

rc.jpg
 
Reviving this thread as my newly installed Red Cube exhibits more or less the same symptoms as the OP pa38112, using the same engine configuration. The fluctuations are over the range of about 1 USG, don't seem regular, and do not really jump around, but rather are constantly changing.
  • Was anybody with the same symptoms able to solve the problem, how?
  • pa38112, if you read this were you able to do something about it, or are you just living with it?
  • Has anyone installed the Matronics dampener mentioned in post #3, results?

Thanks for any response. FWIW, I have no leak in neither air, nor fuel nor on the carburettor shafts (o'hauled recently). Following an avionics upgrade I had to replace my non-compatible and ageing Flowscan 264 with the red cube.
The red cube was again installed free hanging downstream of the engine driven pump.

View attachment 55128
hard to tell from teh pic, but looks like you have the red cube BEFORE the mech fuel pump. If this is the case, you can expect oscillations, at least with the boost pump off. If you move it downstream from the pump, it will smooth out. The mechanical pump is a diaphragm style that pumps only once per engine revolution. Pump stroke is only about 30* of the 360* of rotation. It pumps, stops, then pumps again. These plays havoc with the red cube paddle wheel, which reads in both directions-fwd and backward. This only happens on the suction side of the pump. There is a little bit of pulse shock when the pump stops and can make the wheel go backwards just a bit. There is always some pressure on the output side that dampens and prevents this. When the timing is just right, this pulse causes the cube to double counts the vane - once as it passes fwd and another when it bounces back slightly from the pulse. It only happens occasionally when the pump stops with the vane right at the sensor, which is why the indication bounces around. This is a cyclic counter, so we cant use software to smooth out variations like we do on pressure sensors. Your system reads one gallon/hr when the number of pulses received in an hour matches the K factor number or .1 gallons when it receives K faxtor/10 pulses, or 10 GPH at K factor*10, etc. So, 1* K factor pulses in an hour is 1GPH. 1*K factor/3600 pulses in a second is also 1 GPH. So you can see how double counting a vane every once in a while throws off the indication a good bit.

On the input side, the flow is constantly starting and stopping, but on the output side it is a steady flow thanks to the back pressure and hte pumps check valve. On this side, only the pressure is oscillating and not the flow. THerefore the cube does not react to it and double count.
 
Last edited:
thanks for chiming in lr172, interesting explanations.
My red cube is installed downstream of the engine driven pump (EDP), or expressed differently, between the EDP and the carb.
 
thanks for chiming in lr172, interesting explanations.
My red cube is installed downstream of the engine driven pump (EDP), or expressed differently, between the EDP and the carb.
Interesting. HOw old is it? The paddle wheel spins on a jeweled bearing and it is possible that some debris has built up and is not spinning freely any more, causing similar issues. I have heard some have short term success by running carb cleaner through the cube to clean up varnish.
 
thanks lr172, the red cube is brand new. I made the first flight today with it, 1:20... maybe there's some air trapped in there, though it is installed according the instructions e.g. outlet pointing up and to the carb.
Guess i'll fly a few hours more and see if things improve, if not perform a flush by disconnecting the hose at the carb with the elec pump on, or/and follow your suggestions above.
 
As another data point, My fuel flow indication is also rock solid.

I have an RV-4 with O-360-A1A and carb (Marvel MA-4-5). The red cube is located after the fuel valve and before the engine driven fuel pump on the cockpit side of the firewall. The con to this location is that fuel flow indication increases when the fuel pump is on, the pro is that the red cube is in a 'gentle environment' and their are fewer hose connections on the engine side of the firewall.
 
Thanks YellerDaisy
Probably a simpler system than mine...

One thing I did not mention is that I have a fuel return line Teed after the EDP. Installed is an Andair dual fuel selector valve, so the excess fuel pumped by the EDP, or the EDP + the elec pump is returned to the respective fuel tank. All of this was installed to favor Mogas and fight vapor lock.
OTOH it should not influence the FF reading since my red cube is between the EDP and the carb...
 
I see some fluctuations on my O-360 in my 6A. I chalk it up to the carb bowl getting full and the float stopping fuel flow into the bowl. I could be completely wrong and it is just something strange in my setup.
 
A question for people cleverer than me:

Fuel is an incompressible fluid, for all intents and purposes. Not a lot of elasticity in the fuel lines either.

If there's no air in the system, why would the location of the Red Cube make any difference?

Rate of flow at the suction end is always going to be the same as rate of flow at the discharge end, and you should be able to measure it anywhere at all along the fuel line.

(older sensors which are inferring flow by measuring pressure deltas would likely be different; But the red cube isn't one of those, it's measuring the rotational velocity of a turbine wheel inside the cube, so it should be a direct measurement of flow rate)

- mark
 
FWIW.... O360-A4M in a Piper Archer III, Red cube is installed between the electric boost pump and mechanical pump. Using the EI FP 5L fuel flow instrument.

Fuel flow readings are " dead stable ". When refueling the plane I am typically within .3 gallons reported use vs the fuel ticket.

Boost pump is used only for TO and landing but has minimal impact on accuracy.
 
Reviving this thread as my newly installed Red Cube exhibits more or less the same symptoms as the OP pa38112, using the same engine configuration. The fluctuations are over the range of about 1 USG, don't seem regular, and do not really jump around, but rather are constantly changing.
  • Was anybody with the same symptoms able to solve the problem, how?
  • pa38112, if you read this were you able to do something about it, or are you just living with it?
  • Has anyone installed the Matronics dampener mentioned in post #3, results?

Thanks for any response. FWIW, I have no leak in neither air, nor fuel nor on the carburettor shafts (o'hauled recently). Following an avionics upgrade I had to replace my non-compatible and ageing Flowscan 264 with the red cube.
The red cube was again installed free hanging downstream of the engine driven
I recall having similar issue. Turns out I had mounted the transponder antenna right behind the firewall which in turn was in close proximity to the red cube. Turned off the garmin GTX45R and flow stopped fluctuating. Relocated the antenna behind the aft baggage bulkhead and all is fine.
 
The installation instructions on the FT60 say to install downstream from all pumps to avoid “…vapor lock & jumpy readings.” Also says to perform flow & pressure tests found in AC 23-16.

I have IO but imagine this covers both systems.

The instructions provide a formula to calculate theoretical pressure drop. I got TS Flightlines‘ bracket mount. Tom helped me on the install. Happy with it.


Warm regards,
 
Reviving this thread as my newly installed Red Cube exhibits more or less the same symptoms as the OP pa38112, using the same engine configuration. The fluctuations are over the range of about 1 USG, don't seem regular, and do not really jump around, but rather are constantly changing.
  • Was anybody with the same symptoms able to solve the problem, how?
  • pa38112, if you read this were you able to do something about it, or are you just living with it?
  • Has anyone installed the Matronics dampener mentioned in post #3, results?

Thanks for any response. FWIW, I have no leak in neither air, nor fuel nor on the carburettor shafts (o'hauled recently). Following an avionics upgrade I had to replace my non-compatible and ageing Flowscan 264 with the red cube.
The red cube was again installed free hanging downstream of the engine driven pump.

View attachment 55128
Mine is installed in a similar fashion with a few exceptions. The inlet and outlet fittings are straight fittings to minimize turbulence entering and leaving the cube. Mine sits level. As I recall, the instructions say not to have the outlet lower than the inlet as it may trap air and cause fluctuations. It appears your inlet is above the outlet. My readings are steady.
 
My FT-60 is installed on the firewall right after the tank selector, then to the Weldon boost pump (also on the firewall, then to the engine-driven pump, then fuel servo then engine. 90° angles and everything. Choice made by the guy that my airplane's OB selected to do firewall forward 12 years ago. Works well, no fluctuations.


IMG_8373.jpeg
 
Thanks for all the replies and suggestions guys.

My xpdr antenna is well behind, as in below the fuse next to the flaps trailing edge. No other antenna in the area.
The outlet is definitely pointing up, and the cube installed according the instructions.

I will now go fly some more, and see if the totalizer works. And I'll do some troubleshooting next time the cowls are off...
 
My FT60 is between the servo and the spider on an IO540. So theoretically in the perfect position.
It jumps around all over the place +/-1lph (1/4 gph).
The totalizer however is amazingly accurate. It’s a running joke with the fuel truck. I give him a number before refuel and it’s never been more than 2L (1/2 gal) out on a typically refuel 250-280L (65-73gal). Often it’s within 1Litre.
So I’ve just decided that it’s either engine vibration or mech fuel pump pressure pulses/harmonics plus inadequate smoothing in the engine monitor algo (Garmin G3X).

On my RV7 the FT60 was in the cabin between the boost pump and the mech pump. It would change flow rate with the selection of boost pump (which was annoying) but was always smooth. Yet the totalizer was also similarly accurate. Admittedly the pump is only on for a few minutes per flight.

Not all that helpful I guess but another data point.
 
A question for people cleverer than me:

Fuel is an incompressible fluid, for all intents and purposes. Not a lot of elasticity in the fuel lines either.

If there's no air in the system, why would the location of the Red Cube make any difference?

Rate of flow at the suction end is always going to be the same as rate of flow at the discharge end, and you should be able to measure it anywhere at all along the fuel line.

(older sensors which are inferring flow by measuring pressure deltas would likely be different; But the red cube isn't one of those, it's measuring the rotational velocity of a turbine wheel inside the cube, so it should be a direct measurement of flow rate)

- mark
Fuel inside of a delivery system is absolutely compressible. How else can your pump create 25 or 45 PSI of pressure in the fuel system. We could argue about the physics and what components are either compressing or yielding to allow the pressure, but it does happen. Not a physicist, so cannot say whether the fuel is compressing or the delivery system components are yielding and acting as a diaphragm; I just know that pressure can be produced.

While it may seem that the fuel flow is the identical before and after the pump, that is only from a static overview across a reasonable time frame. Same way that it appears to you that the lights in your house are always on and receiving full power, but they are not -voltage goes from +120 all the way through zero and then to -120 and back again 60 times per second. You just don't see it as an observer, as your brain cannot process cycles that fast along with the fact that the resistive element takes a bit of time to stopping illuminating, creating a capacitive effect; But that doesn't mean it isn't happening. Suction side of the fuel pump is simiilar; constantly starting and stopping even though the ultimate flow rate over longer periods of time is identical to the output. Ever stop aggressively and then feel a slight rearward force after the stop? This is the effect that is happining to the red cubes water wheel type vane each time the pump stops. Sorry, not a physicist, so can't give the correct scientific term for this.
 
Last edited:
As another data point, My fuel flow indication is also rock solid.

I have an RV-4 with O-360-A1A and carb (Marvel MA-4-5). The red cube is located after the fuel valve and before the engine driven fuel pump on the cockpit side of the firewall. The con to this location is that fuel flow indication increases when the fuel pump is on, the pro is that the red cube is in a 'gentle environment' and their are fewer hose connections on the engine side of the firewall.
I have pretty much the exact same setup in my RV6. O-360-A1A Marvel MA-4-5 and red cube mounted between fuel selector and Facet boost pump inside cockpit. Fuel flow is very steady with the boost pump on or off, and it is accurate with the boost off, but over-reads by several GPH with the boost pump on. As long as I don't accidentally leave the boost pump on longer than necessary, the red cube is accurate to within a couple tenths of a gallon on a 30 gallon fill up.

My RV6 red cube and boost pump wiring goes to the panel in the same bundle, so I've sometimes wondered if the problem is electrical noise when the boost pump is on. Hasn't bothered me enough yet to wiggle down under the panel to separate out the wires and see if it helps.
 
My FT60 is between the servo and the spider on an IO540. So theoretically in the perfect position.
It jumps around all over the place +/-1lph (1/4 gph).
The totalizer however is amazingly accurate. It’s a running joke with the fuel truck. I give him a number before refuel and it’s never been more than 2L (1/2 gal) out on a typically refuel 250-280L (65-73gal). Often it’s within 1Litre.
So I’ve just decided that it’s either engine vibration or mech fuel pump pressure pulses/harmonics plus inadequate smoothing in the engine monitor algo (Garmin G3X).
Thanks Richard.
Though our engines differ, I have exactly the same experience. Went flying for close to 4 hrs today, and once in my Carson speed super eco cruise at 7'000 ft on 19"/1'900 was burning 19 liters, so slightly more than 5 USG. The FF seems stable for some periods, then goes down as low as 17 litres and the shortly goes to 27 before going back to about 19. No regular cycle can be observed, and the indication changes don't happen instantly, but over about 2 seconds.

As you also state, the totaliser (or fuel used on the G3X) seems very accurate at the end of the flight.

PS
Still wonder about that Matronics pulse dampener mentioned above... any good?
 
For a carb, you will get some fluctuations. I have experimented with the pulse dampener and didn't find that it did any good. The sensor should be mounted more horizontally than you currently have it and that may help.

Vic
 
For a carb, you will get some fluctuations. I have experimented with the pulse dampener and didn't find that it did any good. The sensor should be mounted more horizontally than you currently have it and that may help.
Thanks for the feedback and info Vic, appreciate.
 
FWIW...

I have the same symptoms in fluctuating flow. It's more exaggerated at lower fuel flows and pressures (boost pump off, in cruise). There is also a corresponding change in pressure with flow (inversely..flow increases, pressure drops). I can only assume what others have said is that the carb bowl is filling. At higher flows and pressures my guess a constant buoyancy of the float at certain fuel flow/pressure keeps the needle off the seat and constantly filling the bowl, so there no observed fluctuation. This same carb, before and after a recent overhaul, has the exact same fluctuation. My FT-60 is installed per instructions, after EDP and before carb.

Once calibrated, my totalizer is within 0.1 of a gallon of refueling total.
 
Back
Top