916 Fuel Flow

Thanks Jeff.

What fuel flow do you use? The only solution on a seri9i is to connect your Garmin on the Bus CAN and you get the exact consomption if you mesure physically something different you must find the calibration menu of the Garmin.

Rgds

Kristof

This is a longstanding, known issue with Rotax iS engines. I spent a good bit of time going back and forth between the Garmin and Rotax booths at Oshkosh, trying to get a resolution for this issue. That was two years ago and nothing has been resolved.

Rotax acknowledges the problem, but when I asked them to fix it I was told “it is not possible.” I think Rockwell makes the ECU and Rotax doesn’t want to pay to have it corrected. I told them someone is going to run out of gas because of this, but the best they would offer is to add a note in the OM. Not good enough, IMHO. 

I asked Garmin to allow us to make adjustments to the FF on the G3X (since K-factor doesn’t apply on iS engines). Garmin said that Rotax requires they present the data as provided from the Rotax ECU, without adjustment. I asked, why not allow us to add an adjustment to the fuel burned totalizer? He said that might be possible.

This Garmin rep was building an RV-12, so the issue will affect him, too. He was going to the Rotax booth in person to discuss the issue, but again, that was two years ago. I’ll email him to see if there has been any progress behind the scenes. 

I love my engine, but it’s frustrating that Rotax won’t take the initiative to fix this known issue, and our only hope is a workaround from Garmin.

Most ECUs calculate fuel flow by counting injector pulses, and compensating for the pulse width and fuel pressure. These values are then totalized. In theory, each injector pulse can deliver a slightly different amount of fuel. The injector pulse width is based on a map for various engine loads, so you would think that would be accounted for in the fuel flow calculation.

However, in the Rotax system the ECU does not know the fuel pressure. It relies on the accuracy of the vacuum controlled fuel pressure regulator. If a fuel pressure sender is installed, fuel pressure is displayed to the pilot, but not shared with the ECU. The fuel pressure can change with one or both pumps running, and other factors. I’m guessing here, but I would bet that more fuel is delivered with each injector pulse when both pumps are running than with just one. And the ECU would not see this difference. On my 912iS aircraft, at times I can see a slight drop in RPM when turning off the aux pump after climb out. Fuel is power, so I assume that is the result of slightly less fuel being delivered to the cylinders. However, the change seems minor, and likely can’t account for all of the error people are experiencing.  

In summary, I believe the reason Rotax says “it’s not possible to fix” is because of limitations of the ECU. We know it can’t account for fuel pressure changes, so the calculation may be simplified and based on averages. I do see a reduction in fuel flow when I go from Power mode to Eco mode, so I assume it’s calculating for at least two different pulse widths. Still, you would think an adjustment to the firmware could improve the accuracy.  

Hi Jeff

You are correct, it is a calculated burn by the ECU and was never intended to do more than work with the internal calculations within the ECU.  There is no substitute for an actual fuel gauge to know what is in your fuel tanks.  

The mechanical fuel pressure control is for sure not monitored and as you stated any differences between them would alter the flow.  Fuel filter restrictions can also change this flow.  As you noted the biggest differences will be seen between ECO mode and max power.  

In talking to the factory people they maintain that the flow rates are accurate in all their testing.  I believe their rather expensive equipment is correct and under the test conditions and the R&D aircraft it seems to be fine.  But as you have noted there are many variables that change with the literally hundreds of different installations over the manufacturers different models.  It would seem that some adjustment on the instrument side would be a good idea to correct for those variables.  The firmware within the ECU is not something Rotax can change except for the entire Fleet.

Cheers