V6 Bongo conversion...

[colin33]

Hi Simon, for starters I've just taken this video of the map screen so you can see how the dot moves about ! This is on petrol but does the same on gas.

http://vid171.photobucket.com/albums/u3 ... /Bongo.mp4

let me know what you think...

[colin33]

Still not managed to stop the bouncing ball!! Tried 'extra sensitivity' on all settings but no difference. Tried 'Weakening on Mazda' at 5,10 &20 % but it didn't like that either. Tried 'full group' but it just conked out. Any ideas anyone?

[LPGC]

Inconclusive from that video, partly because you don't put your foot down until the engine is above 3000rpm. 3000Rpm at higher loads is where Mazda's often switch from sequential to full group operation... Better to concentrate on what is happening below 3000rpm for now. I noted below 3000rpm with partial throttle there were some 4+ms pinj readings - see if they go higher if you stamp on the throttle at low rpm.

Simon

[colin33]

Here's a couple more short clips, but not under load. Note one of them only has one bank showing, this was before I changed settings to 2 banks. Will another road test video stamping on the throttle from tickover help?

http://vid171.photobucket.com/albums/u3 ... bongo2.mp4

http://vid171.photobucket.com/albums/u3 ... /Bongo.mp4

[LPGC]

Either you're not stamping on the throttle at idle or have connected the vac pipe to map sensor to the wrong place on the manifold. If you're stamping on the throttle, connecting the vac pipe up properly won't help with the pinj readings though but still needs to be done.

Simon

[colin33]

I'm pretty sure it's connected the right way. Drilled a new connection for the vac pipe rather than tapping into existing. Here's pictures...

[LPGC]

Hmm... Are you stamping on the throttle from idle then?

[colin33]

Here's another video stamping on the throttle from idle. If you go through the video in slow-mo you can see it does some weird things. Seems to jump about at certain revs, particularly 3500.
https://1drv.ms/v/s!AvB_hRSB0AWxhQsg2qt7jveC8u31

[LPGC]

Looks to me like it's reading incorrectly between idle and 3500rpm.
Does it do the same if you run it on petrol?
In the injector check screen are figures for all injectors around the same or do some seem to drop out?
Does it do the same if set as one bank?
Believe I already suggested adjusting extra injection filtering settings? [I know it's not signalling extra injection but worth a try]

Simon

[colin33]

Here's the same on petrol, sorry it's a bit shakey. Seems to take a dive at 3500rpm. I haven't yet looked at the injector check screen to see what it's doing. I assume it did the same on one bank, but will change it back and check. And yes, I did try adding the 'extra injection' but appeared to make no difference.
I spoke to a guy who has converted a few Bongos, he said that as far as he remembered they run on sequential, not full group. He just suggested changing my plugs and leads even if it runs fine on petrol. Do you reckon dodgy leads/plugs could cause these symptons? Not convinced myself but happy to take advice
https://1drv.ms/v/s!AvB_hRSB0AWxhQru4HcBaXC3KO6r

[LPGC]

Again your video shows mostly fairly normal readings because it shows mostly full load above 3500rpm - It would be more useful if it showed more action with rpms between idle and around 3500rpm, full throttle and partial throttle situations.

I meant adjusting the extra injection filtering settings, as opposed to adding extra petrol injection, try various settings in extra injection filtering. This shouldn't work but it might.

Problems with plugs/leads etc can occasionally lead to similar sorts of mis-readings but usually when the problem is associated with a misfire. A failed/failing crank or cam sensor could also be at the root of such readings.

Without re-reading the thread I don't remember if you connected the rpm wire. If you did then ensure the rpm wire box is ticked, if you didn't then try connecting it and ticking the box. If you get no joy with the rpm wire connected and rpm wire setting ticked, try disconnecting the rpm wire (and unticking the box). In all cases try swapping rpm detection settings between normal and weak, between single coil / double coil / rpm sensor / rpm sensor 2 / etc. To do the swaps it will probably first be necessary to (at least temporarily) have the rpm wire box ticked. Again this shouldn't work (those settings are greyed out when rpm wire box is unticked), it is totally un-intuitive and doesn't make sense but I have known this to cure rpm related issues on newer AEB systems on 6 cylinder engines... A long shot because I haven't known it to cure Pinj reading related issues directly, your rpm readings seem to be OK but I have known rpm and Pinj reading issues to be related so there is an off-chance, though usually in the latter situation would expect the on screen extra injection filtering and other indicators (beside the icon of the petrol switch) to be flashing willy nilly.

Other things to try would be re-routing LPG system wiring (injector break cuts, the rpm wire even if it isn't connected) / Trying an older AEB ECU in place of the new one. Not going to stick my neck out yet - It is possible that the readings seen are actual, in which case the Mazda is running one of Mazda's unusual mode changing injection methodologies and any ECU would show the same readings, or less likely that the readings are effected by EM interference (hence the re-routing suggestion). But it is also possible that the problems you're having would not have occurred if you had fitted just about any none AEB ECU (or had fitted an AEB ECU more than just a few years old) on the 6 cylinder engine.

Also worth trying your new AEB lead with another computer, using a computer with a different OS and trying different drivers for the cable. Some fairly recent versions of software (not the latest) could sometimes do funny things in certain OS's (particularly 64 bit ones), like showing fuel type as CNG when you definitely set LPG, greying out certain other important settings and displaying settings that are totally undocumented and otherwise never seen. In such cases the Pinj readings can also be displayed wrong.

Simon

[Tinleytech]

The problem is most definitely not the fact it is an AEB ECU on a 6cyl engine, not quite sure why anyone would think it was, we have supplied many hundreds of such for 6cyl (including Bongo's) without problems, suspect an underlying problem with the car or possible installation error but have not got to the bottom of it yet but expect we will in the end.
AEB are the biggest and most successful maker of LPG ECU's so rest assured, they work just fine on 6cyl engines.

[classicswede]

Does seem to be an issue with short injector durations. The injection filter might be able to sort that out.

You could also try putting the injection type as MJ sequential

[LPGC]

The problem is most definitely not the fact it is an AEB ECU on a 6cyl engine, not quite sure why anyone would think it was, we have supplied many hundreds of such for 6cyl (including Bongo's) without problems, suspect an underlying problem with the car or possible installation error but have not got to the bottom of it yet but expect we will in the end.
AEB are the biggest and most successful maker of LPG ECU's so rest assured, they work just fine on 6cyl engines.

Dai (Classicswede) and I have in the past spent hours on the phone to each other (calls both ways) trying to solve unusual reading type probs on certain vehicles converted with certain ECU's where all the problems went away after fitting a different type of ECU - most of those situations involved rpm reading and extra injection related problems rather than pinj reading problems (pinj seems to be the problem here).. Those other situations will remain mysterious regardless of the outcome here, you'll recognise both of our knowledge, experience and diagnostic abilities with AEB and a very wide range of other systems. I will have converted plenty 6 cylinder Mazdas (and other marques) using older AEB ECU's without problem but (Mazda's aside for the moment) have had unusual issues with newer AEB ECUs particularly on 6 cylinder engines. That said (and putting things in perspective) modern AEB ECU's are still my usual choice for 8 cylinder engines, I'm not anti AEB or trying to do anyone a dis-service by mentioning my experiences/findings. Above I mentioned how changing greyed out rpm settings can solve unusual problems with extra injection lights flickering, rpm dropping out and ball bouncing - probably not something most have thought to try, shouldn't work but does, but why does it work? Though I'd bet you won't agree changing greyed out settings can have an effect and I don't think we reached a concensus the last time we discussed systems on forum http://www.lpgforum.co.uk/viewtopic.php ... 3&p=108508

On the main topic I reckon Dai's suggestion of setting MJ sequential a good idea. The issue does seem to be that the ECU isn't interpreting Pinj properly (for whatever reason) - probably not due the type of ECU on the type of engine in this case (I said I wasn't sticking my neck out in this case, because it is different to issues I've seen before), could be the Mazda injection methodology point again, in which case Dai's suggestion has the potential to sort all the issues and do a good job of it too (I reckon AEB ECU's were/are particularly good at sensing injection methodology changes on Mazda's etc, i.e. the leaning on Mazda control works better than on other ECUs). Other than that, what might we suggest is wrong with the car if it pulls properly on petrol under all conditions? It wouldn't pull properly if manifold pressure reaches atmospheric but fuel dosage is only half what would provide correct mixture. If Col has followed all advice so far and Dai's MJ injection setting suggestion also doesn't work (even after trying the MJ setting in conjunction with various extra injection filtering settings) it will be a case of what to suggest next though... In which case if the car were here I'd try an old spec ECU, an easy thing to do which might help rule out or confirm ECU related issues, reckon I can think of a couple of other people who would likely try the same.

Simon

[colin33]

The problem is most definitely not the fact it is an AEB ECU on a 6cyl engine, not quite sure why anyone would think it was, we have supplied many hundreds of such for 6cyl (including Bongo's) without problems, suspect an underlying problem with the car or possible installation error but have not got to the bottom of it yet but expect we will in the end.
AEB are the biggest and most successful maker of LPG ECU's so rest assured, they work just fine on 6cyl engines.

After autocalibration a message says " autocalibration complete" then underneath it says "calibration NOT optimal"
Do you know what this actually means, and what triggers it to give the "NOT optimal" message?

[colin33]

Been away for a few weeks but now back to trying to solve this problem.
Today I tried the various settings for the RPM signal. The only one which appeared to show the actual correct rpm was as it was set (standard, rpm sensor)
I also tried the various injector type options, but none made it any better so stuck to 'OMVL Fast"
I then selected "MJ Sequential" and did another autocalibrate. Then took it for a spin, it switches to lpg ok and runs but lacks power compared to petrol. The ball is still frantically up and down on the map though.
Not sure what to try next...?

[LPGC]

Calibration not optimal most likely points to the LPG ECU guestimating that your combination of nozzle size and working pressure are too high or too low. In the display screen there should be a slide bar graphic that shows where it thinks your combination of nozzles/pressure are at regards big / small.

This won't affect the ball bouncing (and neither will type of injectors selection), because the ball bounces on petrol when none of those types of settings are relevant. Settings that could be relevant are those which control how the ECU interprets petrol injection pulse duration and perhaps rpm.. So we were concentrating on settings concerned with things like petrol injection strategy, leaning on Mazda, extra injection filtering and rpm detection. There may yet be some combination of these settings that works properly but it is looking like a long shot. Still may be worth you experimenting with combinations of those settings...

There is still a bit of a chance that although the ball bounces up and down, behind the scenes the ECU could actually be seeing and responding to petrol injector pulse duration correctly, i.e. the ECU itself sees correct readings but they are not reflected on laptop screen. I expect that's very unlikely though in this case. But worth pointing out, and also pointing out that if any major setting such as injection type (you changed to MJ sequential) is changed you need to do calibration again from scratch - and autocalibration by itself might not produce good results for all engine running conditions, you may need to adjust the map too before even getting drive-ability something like OK. You said 'runs OK bit lacks power compared to petrol', can you define/explain this a bit more? E.g. does it seem to be all over the place perhaps sometimes seeming like there's nothing there momentarily when you put your foot down (this wouldn't really be running OK), misfire (could be under certain conditions) or just lack power (probably lean) under certain conditions? How do the different injection type settings (currently MJ sequential) compare?

Presuming the idea is to persevere with this ECU, again to recap/confirm some of the less usual points already checked... in 'injectors check' screen did you monitor petrol injectors duration readings when running on LPG (to confirm none of the readings hangs or reads unusual pulse duration compared to the others - which could point to incorrect wiring of the petrol injector connections on the affected cylinder)? Did you check routing of wiring for possible EM interference (wires routed near alternator / HT components etc)? Main earth and live to ECU connected to battery? Where does the red/white ignition on wire to ECU get it's feed (just possible that under certain conditions the voltage on this wire falls so the ECU momentarily sees ignition as off leading to all sorts of problems. You could disprove this by temporarily connecting the red/white directly to battery + INSTEAD of to the engine, disconnect red/white from engine first!)? What version of software and firmware are you running? Have you tried a different laptop / version of Windows / cable?

Simon

[LPGC]

Fitted a new filler point on a Mazda Bongo V6 yesterday, if I'd had more time I might have offered the owner a check of calibration etc giving me chance to see how the install was configured. Not the same system as yours though and the owner reported no other issues. Having the vehicle here, and your engine pic, both served to remind me that I have converted and repaired Bongo V6's before and not had any problems like yours.

Simon

[colin33]

Regarding 'calibration NOT optimal' here is a pic of display. Looks correct doesn't it?

[colin33]

Today I moved all wiring away from the HT leads/coil. The closest were about 5cm away but I moved them to about 15cm distance. However this seemed to make no difference to the bouncing ball so I can safely assume interference from the HT circuit is not the issue. I will move the wires back to where they were in the interests of tidiness