After a fair bit of tramping up and down my favourite dual carriageway hill in the '02, at 500 rpm increments I've got the LPG mapping as close as I feel I can. Stable trims, good lambdas etc.
Normal road driving its trims are stable and don't wander off after periods of LPG running.
Which is great, except when running on LPG on foot to the floor acceleration it's way down on power to the extent that switching back to petrol during the exercise is like hitting turbo boost.
It's obviously (well, to me anyway) running weak on full throttle.
If I increase the LPG maps by any significant % to richen it up, it throws out the trims for "normal" driving.
Any ideas anyone?
There is a 4wd rolling road dyno in Plymouth, but it seems like a very expensive exercise, not to mention the mirth it'll cause among the Scooby/ Evo drivers who seem to be resident in the building!

On my tartarini, there's an option for 'enrichment during acceleration'. Don't need to alter the map. You not have that?

On a Zavoli it's under a tab called Modify Carb which should deal with it but no doubt LPGC will pop up and confirm how it should be used. Have you confirmed it's going lean from the lambda readings on full bore?

Not yet. I'll have to do a recorded instrumented run with the Nano. My driver for the calibration runs, while I'm monitoring the LPG on the laptop, has a reluctance to do full throttle power runs from TC lock up to infinity, so I'm going to have to drive myself :)

Orangebean wrote:

There is a 4wd rolling road dyno in Plymouth, but it seems like a very expensive exercise, not to mention the mirth it'll cause among the Scooby/ Evo drivers who seem to be resident in the building!

Sounds like we need to invade one day, just to see the look on there faces,, 😳

Summer Camp and Dyno Day Chris? That'd be different :)

Enrichment in acceleration works in different ways depending on age of ECU.

It makes sense to explain how the newer ECUs work first, then it's easier to see why the enrichment in acceleration facility is there at all and why it is more important on older ECUs... Most of you will have the older type ECUs.

On newer ECU's the slider doesn't allow for enrichment at all, full right on the slider gives just the mixture that the map/pressure compensation/temp compensation provides, left of full right leans the mixture during acceleration. On the newer ECUs the only real correct position is full right, anything left of full right points to calibration/setup issues. That's because the newer systems are truly sequential - they take Pinj readings from every petrol injector and (supposedly) each cylinder's gas injector pulse length (Gnj) is calculated individually based on each individual cylinder Pinj's, furthermore the gas pulse should start at almost the same moment that the petrol injector pulse starts (where almost means at the same time or within about 2ms based on other settings such as 'extra injection filtering' where extra injector filtering in theory allows very short Pinj's to be ignored altogether).

The older type AEB based ECU's are not really truly sequential, they take the Pinj reading only from the front cylinder on each bank (blue wires), then each cylinder on the same bank gets the same Ginj as the front cylinder Ginj. Now suppose that between the front cylinder injection pulse and another cylinder's injection pulse on the same bank the driver stamps his foot down... At high rpm the manifold pressure won't have time to increase much before the other cylinders on the same bank inject fuel, at low rpm the manifold pressure will have more time to increase between injection pulses on that bank.. In cases of both high rpm and low rpm the mixture for the rear 3 cylinders (particularly the last cylinder in the firing order) will be leaner than the front cylinder mixture because it didn't read the rear 3 petrol injectors pulse times it just assumed they would be the same as the front cylinder's pulse time. To address this the system has the enrichment in acceleration slider.. The default setting of the slider (usually somewhere near the middle) gives a bit of enrichment to the rear 3 cylinders when the ECU notices manifold pressure and/or Pinj on the front cylinder rising quickly... The enrichment is a fudge, but no more of a fudge than the acceleration enrichment that a petrol ECU provides. Since the newer type ECUs read individual petrol injector pulses to calculate individual gas injector pulses the newer ECUs don't need to add in the fudge themselves because they are reading Pinjs that have already had any necessary acceleration enrichment fudge added by the petrol ECU.

The gas ECU applying an acceleration fudge (instead of the petrol ECU's fudge) isn't really a drawback. The only real drawback of the old type ECU over the new type is that it doesn't necessarily pulse gas injectors at the same time as petrol injectors are pulsed - the new type uses start of petrol injector pulse as trigger for start of gas injector pulse whereas new old type does this for the front cylinder but then spaces the rear 3 cylinder gas injector pulses according to only rpm. Not a problem on older engines but can be a problem on newer engines with features such as VVT/Atkinson cycle.

The acceleration slider won't make any difference to mixture with your foot in a constant position on the throttle, even if you're holding foot flat out, because with constant throttle manifold pressure and Pinj's stay about the same. The slider is used on older ECUs to dial out drive-ability problems when you're changing throttle position (hesitation from lean running with too lean mixture / misfires from too rich), can even be used to compensate a bit for long pipe lengths (needs to be richer during acceleration so more gas reaches the cylinder's intake port in time to make mixture correct with rising manifold pressure). Usually set while the engine is idling then stamping on the throttle (idle to flat out /idle to near flat out / idle to partial throttle) looking for crisp throttle response, and adjusted again while driving.

If you want to change top end mixture while throttle is held constant,you use the other mapping tools (map itself etc). Going too rich on LPG will result in less power than a stoch mixture on LPG because the greater amount of gas displaces what could otherwise be airflow into the engine (hence negatively effecting volumetric efficiency) to a greater extent that going too rich on petrol... You still want a rich mixture at high loads but not always as rich as would make more power on petrol. If you have a 0-1v lambda probe, where on petrol max power might arrive at richer than the narrow band lambda is capable of reading, on LPG max power might arrive at as low as 0,7v but we might still go a bit richer than 0.7v to protect the valves. Set the map first, then adjust the acceleration slider.

Wouldn't bother with a rolling road, they never give same under bonnet temps as driving on a real road. Under bonnet temps effect the LPG vapour temp (and reducer temp) readings which effect the mixture. If you've been sat idling in an LPG car for a while and under bonnet temps get high, if you suddenly set off and boot it the vapour temp reading will for a while be high, causing the mixture to be richer than it will be after a while of driving on the road when vapour temp readings have settled down to normal. Greater volume / faster throughput of gas through the reducer mean that gas temp readings fall when booting it but there is lag between booting it and the temp sensor reflecting the correct temp of gas reaching the injectors - If you have hot under bonnet temps to start with and then set off booting it for a while the mix will start rich and get leaner until temp correction has fully caught up, should be setting mixture when compensation has at least nearly caught up. A rolling road won't keep the rad/engine as cool as real road speeds, AEB systems also compensate mixture for reducer temp. Seat of the pants while reading trims/mixture is probably better than a rolling road.

In the case of Orangebean's install (where adjusting the top end mixture throws off calibration at other points) I'm not sure what ECU he's got. Some ECU's have map screens with only a few denominations for engine load, so the top end fuelling box also effects near top end calibration / some have a lot more denominations (up to about 12 as memory serves) for Pinj's between 2.5 and 18ms as standard, and furthermore the denominations can be adjustable (i.e. the 18ms column could be made a 25ms column). If you don't have enough denominations the only way to get things absolutely correct is by playing with injector spec, pressure and nozzle size... but the get around is to set top end fuelling correctly at the expense of correct trims at near top end fuelling.

Also worth bearing in mind that depending on ECU type, year of ECU and firmware the system may or may not automatically switch back to petrol temporarily (accompanied by flashing yellow light on the switch but without beeping) when it senses Ginj exceeding available Ginj window (rpm versus gas injector pulse time). If rpm's are 6000 there is only a 20ms window for Ginj, if Ginj is above 20ms the injector won't have time to close before opening again to meter fuel and the engine will run lean regardless of Ginj being 20ms or even 50ms because the gas injector has already reached the point where it is open constantly... In which case you'd need to increase pressure, and/or or fit bigger nozzles, and/or fit faster opening or bigger flow capable injectors.

Thanks for comprehensive info LPGC. I'll read through it this evening, digest and probably come back with more questions!

Very interesting....

I think I need to find a time where I have someone else who can sit with the laptop and monitor LPG maps as I'm driving. Mine feels a little bit underpowered when I put my foot down, and I'm wondering if the map is set too rich as you mention, and needs leaning off a bit in places... but it's hard to tell that when you're trying to drive on your own and monitor the laptop too!

I will have to also have a look and see where the enrichment in acceleration slider is set on mine. I have a new King 'D' series ECU, so going from your notes, it should be fully sequential...

Cheers for the info LPGC!

Marty

I found it worked best done the other way round Marty
You monitor, test pilot drives! Problem was mine bottled doing full power acceleration runs. Need to find a new test pilot.

Interestingly, I floored it today to sprint onto the motorway and almost straight away it bleeped a lot and switched itself to petrol.
Filters are all new so I'm wondering if my 2005 vintage vapouriser is due for replacement...
(and no, I wasn't low on gas!)

Yeh Marty, yours will be truly sequential so your slider should be set fully right. The only real exception for modern AEB ECU's is on a tiny minority of vehicles that do a hell of a lot of petrol enrichment during acceleration (where acceleration is defined the same as in the slider context, i.e. increasing throttle position)... usually dodgily remapped turbo stuff that shouldn't have so much acceleration enrichment if the petrol remap was good. Too lean usually decreases power more than too rich, for much decreased power from too rich you'd probably have to be so rich that if you wound the window down a bit you'd pick up on that eggy smell.

I calibrate and drive at the same time but I'm well used to it and can just glance across... Still I prefer calibrating slow stuff rather than really fast stuff. If someone else drives you need to drill it into them that constant throttle position doesn't mean constant speed / a touch more or touch less throttle means maybe a 1% shift in throttle position not a 30% shift in throttle position / stamping down and then holding full throttle means just that / etc. You'll probably see what I mean if someone else drives! If the driver is uncomfortable holding full throttle etc it's probably safer if you drive. Full throttle doesn't mean full speed, you can slow to a stop and accelerate hard at full throttle without hitting unsafe speed at least for a short while and on some cars you can extend that time by riding the brakes at the same time as acclerating though not to the extent of cooking the brakes..

Sometimes on repair jobs the customer drives but mostly a re-calibration gets done quicker if I drive while the customer sits with the laptop sideways on his knee so I can glance at the screen. I often regret letting someone else drive, many struggle with the concept of constant throttle (not unsafe but makes calibration more of a pain), some seem incapable of steering smoothly round a bend while holding throttle steady.

Just ordered a nice shiny new one of these

After 12 years and 167 000 miles I figure it was about time to pension the old one off.
I'm thinking that these will come with the output pressure pre-set?

The reducer in your pic (Zavoli Zeta S or Zeta N) is a decent unit but a bad point about these models is that they do come with pressure preset and are not very adjustable away from preset pressure which is 1.4 bar on the S and 1.2 bar on the N. You'd want the 1.4 bar Zeta S for the flow requirement of a P38 but if your injector nozzles are 3mm you'd really want less than 1.4 bar pressure maybe around 1.1 bar..If you do manage to adjust this model reducer from 1.4 to 1.1 bar it's flow ability may go down dramatically (much more than the difference between 1.4 and 1.1 bar would suggest). Lack of pressure adjustment is less of an issue for an installer fitting a new install than it is when replacing an existing reducer because the installer can choose injectors and injector nozzle size to work with the engine for the given pressure. We might assume that the original installer did that, so replacing a reducer that comes with preset pressure like for like should make for good results but the thing is a lot of installers who fitted a reducer with preset pressure won't have fully addressed nozzles / pressure so better results might be had fitting a reducer that isn't so limited in pressure adjustment.Before I could advise fitting a reducer with very limited pressure adjustment I'd need to know spec of your injectors and nozzle size, or at least know how ginj compared to pinj when the old reducer was good and calibration was correct.

I'm afraid I don't have ginj and pinj for when the old reducer was fitted and working properly, 'cos it's never really worked properly for long enough since I bought the car a year ago. ECU died first and had to be replaced and since then the cars been on and off the road with various resto tasks.
It did run happily on that LPG setup for 160 000 miles though so I'm hoping that a like for like replacement will do the trick.
We'll see, when the new one (Zeta S) arrives next week. Next time I'm near the matrix injectors I'll see what size the nozzles are...

Hey, Lpgc.

Don't know if you want to branch out, but I've been a member on another forum for a few years now, and a guy is having issues with a 2001 XJR switching to petrol under heavy acceleration.

This is his take, and it sounds a familiar lament...

"Well i appear to have another issue. if i floor it when running on gas it coughs and splutters then switches back to petrol. On petrol it's fine; it pulls like a demented steam train. I'm thinking of getting the LPG side of things serviced asap. Only thing is, having owned a few LPG cars, it seems it's the perfect scapegoat. Even LPG pros like to blame it. "The fitters must have been cowboys - i can't help you..." etc. It's one of my least favourite things about owning an LPG car; no one wants to touch it! I'm hoping i have a bit more luck with this one."

The thread link is here...

http://forum.retro-rides.org/thread/200305/2001-jaguar-xjr?page=6&scrollTo=2365841

if you fancy changing his views! :-)

I'll have a look George thanks.

Well, shiny shiny arrived today- great service from Tinley, but I'm slightly puzzled. Ordered a Zavoli Zeta S, was invoiced for a Zavoli Zeta S, what I have looks similar to my Zeta S, but it's a Tomasetto (see pic). I seem to be missing a male/male connector for the Valtek 07 solenoid as well. Temp sender fitting is different to the 10mm threaded hole that I have in my current one, so my sender won't fit.
Before I make a fool of myself with Tinley, by accusing them of sending something that I didn't order, have Zavoli now become Tomasetto or something?

Give them a bell, they are very friendly and don't bite. What you have doesn't look like any of the other Tomasetto reducers shown on their site, in fact, as you say, it does look very much like a Zavoli. Maybe just a bit of badge engineering or the two companies have become one.