Hi all,

I have a dual fuel kf verada and she like to backfire every now and then. Last time it backfired It blew a big chunk of the afm into the air box. I run on lpg 98% of the time and with the afm as is the Petrol now obviously runs like shit. I disconnected the afm to see if it would effect starting on lpg. To my surprise it runs great on the afm-less tune. Starts and runs fine other than a CEL. Which only illuminates when switched to petrol.

I know I need to get the gas looked at, no need to lecture me..... Lol

But Is there any critical issues with afm-less tune I should know about until I get a replacement?

Cheers
Dayne.

The only critical thing I can think of would be to make sure that no seriously hard bits got sucked back into the intake after the event, and that there aren't any such remnants still in the intake tract waiting for the right disturbance to get delivered to a cylinder...

There is a simple alpha-N based "limp" fuel map in the ECU for when the MAF is misbehaving - easily examined simply by disconnecting the MAF as you did. As you found this doesn't usually run very well as-is (on petrol). However your gas system isn't taking any notice of the injector signalling for fuelling (just making sure the ECU thinks it still has injectors connected) so I'm not surprised it runs much better on LPG. In limp mode the ECU might also use the fixed base ignition advance (either 5° or 10° for the 3.5, which you could verify with a timing light) so performance under load probably won't be that great.

I'm a little surprised that the CEL only shows when switched to petrol though? I would have expected the ECU to be complaining about the lack of a usable MAF signal regardless...

I was surprised about CEL showing only on petrol. I switched it back and forth a few times to be sure. Only on when on petrol. I will check timing after work today to check for advance. I'm going to run the half tank of fuel I have to check performance of petrol without maf, I will also check cold start this morning on my way to work.

I got some chirpys turning right from a stop which surprised me. Running on petrol still has a lot more power than running gas especially under 3k rpm. Is there a way that the ecu would still be adjusting advance as it didn't seem to have a great deal of loss without maf? What sensors are normally used to assess load?

Cheers

Hmmm.... makes me wonder about whether the difference between LPG and petrol has to do with oxygen sensor output - if the LPG mixtures are pretty close to stoichiometric, the oxygen sensor readings would give reasonably sane short term fuel trims and the ECU decides all is good (or good enough). In comparison, the petrol fueling on the limp home map probably isn't as good so the short term fuel trims might might be outside acceptable limits, leading to the CEL...

If this is correct it would suggest the E/F series ECUs don't really insist on the MAF as such, just a workable closed loop operation.

I'm pretty sure with maf unplugged it uses throttle position for load. It is a very basic map, Merlin had a go at adjusting it on my car and it actually is much more drivable then it was.

Hints : http://www.go-lpg.co.uk/Hints.html

A couple of surprises logging the ECU of Dayno's car with the MAF disconnected:
- the LPG system seems to be supplying its own IAT signal to the ECU when LPG is selected, which is why the CEL disappears when switching to LPG;
- this mixer system, which seems to be only a few years old, uses closed loop mixture adjustment based on the stock oxygen sensor;
- the ECU is varying the ignition timing, although the actual timing is quite retarded (to my inexperienced eye, it seemed like it could be by at least 10° based on the idle readings).

It wasn't a surprise but the limited logging when running on petrol suggests the limp mode fuelling is extremely rich: the oxygen sensor was logging very nearly 1V when normally it shouldn't get to much more than about 0.9V.