Fuel Flow Jabiru Engine

[JimJab]

Hi All,

Well, here is an update on my Fuel Flow Sensor's really bad pulse frequency fluctuations.

I decided to remove my Fuel Flow sensor and bench test it. The results on the bench test where the same as in the plane with the pulse count fluctuating all over the place for any given gravity feed rates between 2Lph to 35Lph.

BenchTestFuelFlowV1.jpg (187.51 KiB) Viewed 9475 times

So I cleaned out the impeller and found nothing that might have block the impeller. I check the 2mm jet was seated all the way in and put the sensor back together. Using a length of 8mm tube, I then blow a very light amount of air through the sensor and watched the impeller turn under a bench magnifying glass. At these very slow rotation rates and at faster rotation speeds, I could not see the slightest hesitation or binding in the rotation of the impeller at slower rotation speeds or as it slowed down from spinning faster.

I was hoping the problem may have been fixed, so I put the sensor back on the bench test rig. Sadly the oscilloscope continued to showed a wide range of pulse fluctuations, I mathematically calculated this could cause up to +-15% variation in fuel flow rate.

MagnetsFuelFlowImplellerV1.jpg (103.89 KiB) Viewed 9475 times

So I ordered a new fuel flow sensor from Stuart the MGL Dealer in New Zealand, I have found nothing but great service to me here Australia.

So as soon as the Flow sensor arrived, I bench test this new unit to find the pulse frequency was very stable with in 0 to 0.5Hz for the lower fuel rates and from 0.5Hz up to 1Hz for the higher fuel rates of 35lph. I also used Stevo's box, as I called it with the 12v DC input supply to 10v regulated DC output.

I certainly like the look of the oscilloscope fuel flow pulse as in the pic with both the regulated 10v DC supply and the new flow sensor.

So what's wrong with old fuel flow sensor, my guess it has a problem with the internal electronics that is causing those wild pulse fluctuations.

BenchTestFrequencyWithStevo12vBoxV1.jpg (59.59 KiB) Viewed 9475 times

I also used a spread sheet to keep a check on what the K-factor would calculate out to be for any given Lph and pulse frequency. The bench testing of this new Flow Sensor and my spread sheet calculations showed the k-factor very close to 7000 as in specs for the 2mm jet. I my case the K-Factor worked out to be 6665.

So where to next, well I am going to connect and send a pulse signal down the flow sensor wiring in the plane to the RDAC/EFIS, to emulate the pulse rates for different LPH I recorded during my bench testing for the new fuel flow sensor.

If all the pulse frequency show a correct solid LPH on the EFIS and these fuel rates don't change as the Avionics, Lighting, Radio or the electric fuel pump are turn on or off. I will then start the engine and ground run it to see if this causes any interference to the emulated pulse rates and the EFIS's indicated fuel flow rates.

Cheers
JimJab

[JimJab]

Hi All,

I have re-written this post as it was suggested it was to hard to read, I think this is a lot better.


Yesterday I decided to do the following tests.

Test 1. Without any other aircraft system turned on other than the EFIS. Use a signal generator to emulate the fuel flow pulses going through my
aircraft's fuel flow wiring to the RDAC and EFIS. This test is to test my wiring and the accuracy of the EFIS's reading of the fuel flow rates.

Test 2. Run Test 1. again, whilst turning on and off all the aircraft's electrical, lighting, avionics and pumping systems looking for any Electrical
Interference to the pulse signal and flow rate.

Test 3. Run Test 1. and 2. whilst ground running the engine to see if there is any Electrical Interference to the pulse signal and flow rate,
while the engine is running.

So to Test 1.

I connected my signal generator to the signal wire. I set my signal generator to emulate the pulses from a fuel flow sensor with the pulse voltage set to 4.5v, the RDAC supplies 5v, so I thought after a little voltage drop getting to the back of my aircraft, 4.5v would be in the ball park. The EFIS's and spreadsheet K-Factor was set to 6665 which is the K-factory worked out from my bench testing.

PulesGenFuelFlowV1.jpg (49.26 KiB) Viewed 9464 times

I emulated 25 different Flow Rates/Pulse Frequencies with only the EFIS and signal generator turned on. I checked the EFIS's LPH readings against the mathematical calculations for that Pulse Frequencies. Through out the range from 0.54 to 54 LPH all the EFIS's readings were within 1% of the calculated flow rate. So the Voyager EFIS and aircraft wiring harness both passed this emulation testing with flying colours.

FuelFlowSpreadSheetV1.jpg (46.67 KiB) Viewed 9464 times

Test 2.

Now turning on and off all the different electrical systems whilst doing Test 1. to see if the EFIS's LPH readings would vary. The EFIS's LPH remained rock solid. So none of the aircraft's electrical systems caused any interference to the fuel flow pulses and the EFIS's LPH remained rock solid and within 1% of the calculated flow rate.



Test 3.

Now with the engine ground running and going through Test 2. again. I saw no change in the EFIS's LPH readings and all remained within 1% of the calculated flow rate, once again.


So I am really happy to see my aircraft systems don't seem to introduce any EMI interference when using these test pulses.

But the real test is to come when I install the fuel flow sensor back in the aircraft and the Fuel Flow Sensor is using the aircraft power supply to generate it's pulses.


Cheers
JimJab

[JimJab]

Hi All,

So I am up to my next test with some anticipations.

I installed my new fuel flow sensor in my Jabiru, in the process, I changed the positioning of the sensor further up stream from my electrical fuel pump(pulse type), so the fuel sensor is now about 0.5m from the electric pump by fuel hose and about 1.7m from the mechanical fuel pump on the engine.

But sadly with this modification and with the engine being ground run, my new fuel sensor being powered from the RADC's 5v power supply had a pretty flaky EFIS LPH readings at all different engine RPM's.

And it seem to get worst when the electric fuel pump IS turned on, the EFIS's LPH would swing dramatically which is understandable as the pump pressure builds up. Then as the LPH stabilises on the EFIS, the reading would end up being well below (about 1/2) the LPH reading when the electric fuel pump was turned off.

So it's disappointing to get this results after having such good results with all the emulation pulse testing.

So the next test while powering the Fuel Flow Sensor from the RDAC will be to insert an old style resistor wheel, to test different pull up resistor values while monitoring the pulse signal on the oscilloscope.

Then I will put in Stevo's regulated 10v power supply to power the Fuel Flow Senor and give a much stronger pulse signal to the RDAC for it to work with.

I will report after all this.

Cheers
JimJab

[Tasmag]

It sounds like the fuel pump pulses (from both mechanical and Electrical) are the issue at this point. When I was trying to get my fuel flow to read accurately I tried an extra fuel filter after the turbine as an expansion chamber, with limited success.

Another option was an anti-hammer pipe, similar to what is used in house plumbing.

I have also heard of people inserting a rubber outboard bulb to stop the pulsing.

Mike

[JimJab]

Hi Tasmag,

Now that I have moved my Flow Sensor further up stream from my Elec fuel pump, and now is about 0.3m from the header tank that is behind my rear seat and sound curtain.

I will try adding another Jabiru fuel filter just up stream from the Elec fuel pump.

Then I will ground run the engine to see if the oscilloscope pulse frequency remains the same at the RDAC as I turn the electric fuel pump on and off.

I have also heard adding the extra filter for some people worked and for others not so well. I am wondering if in some cases there is more than just the one thing to do to fix the problem.

I am hoping also Stevo's 10v DC regulator supply idea for the Flow Sensor will also help resolve the issue.

Cheers
JimJab

[JimJab]

Hi All,

If you have been reading this thread, you will know I had check every aspect of the Fuel flow Sensor, EFIS and sensor wiring and the last of these tests is to sort out the issue of the electric fuel pump when turned on, it changes the EFIS's LPH reading to go higher or lower and remain there until the pump is turned off.

This pic below shows the way my Jabiru 430 fuel flow Sensor was originally set up.

The electric fuel pump is just out of the pic on the left lower corner and on the bottom right corner is the black rubber cover of the fuel filter.

It showed from all my testing when turned on the electric fuel pump's short distance from the fuel flow sensor, could caused cavitation's and send fuel pulses back up into the Flow Sensor, this could changed the sensor's pulse rate and in turn changed the EFIS's LPH reading whilst the pump was turned on and this would make the flow rate very flakey and erroneous when the pump was on for anytime.

Fuel Flow SensorR3.jpg (189.87 KiB) Viewed 9415 times

So I change the positioning of the Fuel Flow Sensor as far up stream from the electric fuel pump as I could, and added another fuel filter as Tasmag suggested to help dampen the fuel pulses and cavitation's from the electric fuel pump.

In the pic below, on the bottom you can see a hose clamp on the electric fuel pump and on the right side you can see again the black rubber on fuel filter and down stream is the Fuel Flow sensor and then the extra fuel filter I am using to dampen the fuel pulses from the electric fuel pump.

So what is the outcome of all this, well due to the weather on the day I made these changes, I could not take the plane for a fly after we did some exhaustive ground running at different RPMs with the wheels chocked and the electric fuel pump on and off at each RPM setting. This extensive testing was as much about making sure the new fuel hose routing was not going to cause any fuel flow issues.

As to the EFIS reading of the LPH, I am pretty excited to see at any RPM, when the electric fuel pump was turned on, the EFIS's LPH would rise for a couple of seconds and then settle back to the same LPH before the pump was turned on. When the pump was turned off the EFIS's LPH would rise for a second and then settle back to the same LPH reading.

I am very happy with all this at this stage. I have not installed Stevo's 10v regulated power supply as yet, but I think I will do some flight testing first with the 5v from RDAC, before I install the 10v regulated power supply.

Cheers
JimJab

New Flow Sensor SetupV2.jpg (156.73 KiB) Viewed 9415 times

[JimJab]

I am very happy with all this at this stage. I have not installed Stevo's 10v regulated power supply as yet, but I think I will do some flight testing first with the 5v from RDAC, before I install the 10v regulated power supply.

So yesterday I took my Jabiru 430 up for a flight to check out the EFIS's LPH readings.

After take off and over the next several minutes, I took the following readings.

Full power at climb the Jabiru book value is 38-40 LPH and my EFIS was reading 39.9 LPH, at top of climb I dropped the revs to 2700rpm and the EFIS was reading 20.7, the book value is 20.0 LPH, next at 2850 rpm in cruise the EFIS reading was 24.6, with the book value at 25.6 LPH.

At each of these rpm settings I turned the fuel pump on and I let the reading settle and then turned off the pump, in all cases after a number of seconds the pump reading on or off was within a few 10th's of the litre rate before electric pump was turned on.

Boy, I thought I was pretty close in getting this issue sorted. Then all of a sudden I saw the EFIS's LPH rate @ 2850 rpm go from 25.6 and heading up to 29.5 to 35 and back to 30 LPH in a space of several seconds as it stabilised. But now the LPH reading was well over the book value that I was getting only minutes earlier.

What the heck, so I check the K-factor, it had not changed. I then set the revs every 100 rpms, from 2400 through to 3000 rpms to record the EFIS's LPH readings and to test the electric fuel pump being on and off, as you turned on the pump the LPH rate would increase for few seconds and then settled back to the same LPH rate before the pump was turned on.

So it looks like the extra fuel filter is controlling and stopping all the physical pulse waves going back up the fuel line from the electric fuel pump.

So now this problem must be an electrically issue, so what caused this to happen suddenly after 5 to 7 minutes into the flight? Could it be a case of static discharge, but the LPH reading is still very stable, even though it's wrong.

Maybe the aircraft battery reached full charge and this changed the electrical quality of the aircraft's power supply and this brings us back to Stevo's solution of needing a 10v regulated power supply to give a higher voltage and to get a cleaner pulse wave from the fuel flow sensor.

So I am thinking, I will not change anything and take the plane for another flight and see if all this happens again in the same way.

Cheers
JimJab

[JimJab]

So I am thinking, I will not change anything and take the plane for another flight and see if all this happens again in the same way.

Well here is a little update, I took the plane for another flight, to check my fuel flow rates at 100 RPM intervals from 2400rpm to full power, using the RDAC's, 5VDC with a Pull-up resistor.

So the outcome from this second test flight was, all my LPH readings were the same as the reading at the end of my last test flight. But I noticed during this second test flight, that there was some variation in the LPH value for the same RPM setting.

So my next step, to install Stevo's 10VDC regulated power supply and see if this reduces this small LPH variability for any given RPM.

Cheer
JimJab

[rainier]

Perhaps somewhat related.

a) You can power the sender from 12V direct - no issue at all this usually caters for cases where the sensor output signal is not high enough.

b) A fuel primer bulb as used often with marine outboard engines is a cheap and super effective way to add a pulse damper between a pump and a flow sender. In addition these contain simple valves as part of their pumping mechanics which make them "one way" and tend to prevent most of any fuel shunting issues.

[JimJab]

Hi All,

I made this simple test rig, so I can insert it into the Fuel Flow Sensor wiring via the cable plug near the Flow Sensor. I have not done any inflight testing with it yet, my work has got in the way once again.

It took a little bit to make, but I wanted an easily reliable way in flight to switch between the two different power sources, the RDAC 5VDC supply and the regulated 10VDC, which is being supplied from my aircraft's 12VDC system. We can also easily change the pull-up resistor to try different values.

I did not want to deal with the possibility of having bridging clips, unclipping, shorting or causing any type of intermittent connections or false reading inflight from vibration or a bad connection.

So now, I am in the process of borrowing a battery operated signal generator and oscilloscope. So we can test inflight the different pulses signals and see the difference between the LPH reading from the Fuel Flow Senor using the different voltages. I will connected the oscilloscope to the RDAC signal input to see how the different pulse signals look inflight as well.

With many aircraft types you just can't go and do this type of thing, so you should check first. But having an Experimental Register aircraft does allow you to experiment with it, with properly skill people.

P.S. the labels on the test rig may not be right, eg. "Sender" instead of "Sensor"

Cheers
JimJab.

FuelFlowVoItageTestRigV2.jpg (113.29 KiB) Viewed 9236 times