The place for technical discussions about GA and flying.
Forum rules: Technical discussions about GA only, please.
User avatar
By PaulSS
#1676802
When I first started test flying my friend’s new RV8 all seemed well with the fuel flow on the Dynon Skyview. The numbers were sensible and just what I would expect from an IO-360. The fuel transducer is the classic Red Cube and the default pulses settings returned normal readings.

A very brief hiatus in testing while I went to work for a couple of weeks. During that break my friend changed the fuel filter for a brand new factory unit. He mistakenly believed the whole filter unit had to be changed, as opposed to the screen cleaned, so we know the thing he fitted is extremely clean. The fuel lines were purged afterwards. He also tried to chase down a problem with the oil temperature sender (which was sending us readings that were too low). Nothing was changed with the fuel system but the oil temp sender (a new one and the existing one) were both connected directly to earth to prove we didn't have a ground problem with them. While the battery was disconnected he installed another ground strap from the airframe to the battery tray, which is connected via 6 bolts to the airframe to act as the ground return, with a rather thick bit of cable twixt the battery tray and the battery negative.

I resumed the testing program (25 hours in Oz) and noticed that leaning did not give me the same fuel flows as before. They were higher at all RPM settings , even when the same parameters were set as far as altitude, RPM and temperature lean of peak. Fuel dips showed the high readings were only reading high and that the engine was still burning fuel at the previous, original, rates. In other words Skyview was over-stating the fuel flow.

Now the strange bit :shock: : when the booster pump is turned ON the fuel flow goes DOWN and, I think, to the actual rate at which the engine is burning the fuel. Fuel pressure increases slightly, as you would expect.

We thought that some sort of air lock may have formed, causing the flow meter to read too high and that the increased pressure from the booster pump was enough to ‘squeeze’ the bubbles into solution and return the transducer to normal. But the pipes had been bled and my friend disconnected the Red Cube and made sure that there were no blockages, binding of the impeller or any other reason for it to not work physically. Still the same symptoms continued.

We considered electrical load causing some spurious readings on the Skyview, so I left the pump off and simultaneously operated the electric flaps and the electric pitch trim. Maybe not the same current draw as the fuel pump but it made absolutely no difference to the readings; booster pump on or off.

When the pump is turned on or off the change to the Skyview readings is pretty much instant. The fuel flow doesn’t ‘jump’ to the new reading but it does start to change straight away and takes only 2 or 3 seconds for it to stabilise at the new numbers. The fuel pressure change is pretty much instantaneous.

We’ve discussed the possibility of a ground loop being introduced with the extra wire from the airframe to the battery tray but really don’t see how this is a possibility, especially it being so close to the battery and so far away from the engine (the battery is behind the rear seat).

Obvioulsy, if the fuel flow dropped when the pump was turned off we’d be less confused but for it to drop to sensible readings when the pump is turned on just doesn’t make much sense to my mind. Just as an example, yesterday at 2300RPM, 5000’ and lean of peak by 6 degrees, I was getting a fuel pump on reading of 29 Ltrs/Hr and then 38 Ltrs/Hr with the pump off.

SO, kind people, do you have any ideas of the cause or have you ever heard of such a phenomena? I will be asking the same of the Dynon Forum once they’ve registered me.

I look forward to the instant answer to all our woes……so that we can concentrate on getting capacitance plate fuel levels to sensible readings :D
User avatar
By Trent772
#1676990
I would suggest you put this on here.

http://www.vansairforce.com/community/index.php

First thoughts are -

Did you change the K factor after replacing the filter ?

Is the pump/filter low on the left side, then feeding forward through the red cube ? If so, the flow under pressure may be restricted, whereas the flow under suction (when the pump is off) is un restricted. The fuel pressure will raise, but the fuel flow isn't.

Are you on a standard fuel injection - Bendix or Silverhawk - i.e no return flow to tank ?

VAF should sort you out quickest but feel free to PM for a chat.
User avatar
By Rob P
#1676994
Just to confirm that our Dynon shows a 3 - 6 lph rise with the pump on, returning to the lower figure when it is switched off. Just as you would expect.

Rob P
User avatar
By PaulSS
#1677133
@Rob P Yes, that is what I would expect and we'd be quite happy with that but what we have is the opposite i.e. the fuel flow goes UP when the booster pump is turned off.

@Trent772 We didn't change the K factor on the red cube but that was because nothing else changed. The entire fuel filter was changed and replaced with the same model as before, so we didn't really see any reason to change the Red Cube settings.

It is a standard set-up. There is no fuel return to the tank; the 'extra' fuel after the booster pump returns to the booster pump inlet. All standard as far as I know.

The pump is low on the left side and does then go forward through the Red Cube before it reaches the mechanical pump. We will definitely look at the way the pressure changes as it goes through the cube BUT it s strange how it was all good to start with and now things are different. He'll hate my suggestion of putting the original filter back in because it'll be upside-down again in the footwell :shock: Mr Vans is obviously a genius but some of his ideas for the fuel system are not appreciated :D :D

I'll try and post on the Dynon Forum when I can but I know my friend is already registered with the Vans Forum, so I will suggest he posts there to avoid repetition.

Thanks for both of your inputs and we welcome any ideas or thoughts anyone else might have.
User avatar
By PaulSS
#1683525
Just to close this one off (in case someone ever searches for the same symptoms), it turns out it was nothing really to do with electricky things but all down to a knackered mechanical fuel pump diaphragm, which was allowing air to enter the system. We had always suspected air bubbles etc but it was just trying to find out where they were getting in the system.

With the electric pump on the mechanical fuel pump has a cup of tea and doesn't do anything, therefore the diaphragm is not in use (basically speaking). No air bubbles and all is well. Off goes the electric pump = mechanical one puts down the cup of tea, starts doing its job and introduces air bubbles (not its job), confuses the fuel flow meter and up go the readings.

This was a brand new engine but it had sat for a long time before being used, so the diaphragm material probably perished (the Australian sun/heat on tin sheds/hangars can cause things to get a bit toasty and a bit dried out). Maybe the fuel pump is not filled with preserving fluid when the engine is manufactured.

Anyway, all is now well and the fuel flows, fuel used & fuel remaining coincide with tank dips and written calculations. Now if only we could get the stupid tank readings to work properly all would be well :D
User avatar
By Trent772
#1683537
Great follow up - thanks.

We had a similar issue with the APU on one of our 330's. Trim line that feeds it was admitting air when the APU was drawing fuel up the pipe creating an air bubble up the line and auto shutdown. Fine when the tank pumps were on forcing fuel up the trim line. Great in Goa at 30c just as the pax are boarding and it plunges into darkness.

Tank readings on 7/8/9 etc are interesting.

Unless you use capacitive senders, the float hits the top skin. You will read the lower 2/3 of the tank nice and accurately, but the Dynon will stop at around 50litres even when the tank is full.
User avatar
By PaulSS
#1683580
The poxy tank contents are a pain in the bum. There are capacitive plates fitted and we've calibrated them a couple of times from tanks empty, while adding 10 litres at a time (in the flying attitude). I've even downloaded the voltages off the Dynon files (to something like 6 decimal places), plotted them and extracted 'best fit' voltages against the 10 litre data points and added those to the system instead of just the basic readings every 10 litres.

When it's full the gauges are correct and they seem pretty good down to about 50 litres each side. At this point they seem to 'hang up' (telling you that there's more in the tanks than there actually is) but they don't 'hang up' symmetrically. After further use the contents reduce and get closer to actual fuel state but always over-read.

It's not a huge problem as the fuel flow (now it's fixed) does work well to tell us what has been used and what it calculates is left in the tanks. It worked well before the diaphragm problem and it seems to be doing a good job now. However, having fitted the capacitive plates and done all that work it's not too much to ask them to do their job.

I think at the first annual we're going to fit resistive floats and go with those instead. I don't mind if they only start reading below 50 litres each side and are accurate below that but the current ones work well at 80 litres a side, when there's no real need to look too much and then tell lies when there's less fuel :roll:

I'm glad my Eurofox will just have sight tubes (plus fuel flow/used/remaining on the EFIS) :D