Pending fault codes P1171 and P1174, tuning advice needed please

[Bone9]

Sorry only just seen this for some reason I didn't get a notification.

Got it ok I'll have a play. Thank you.

Ah I see, I thought auto calibration would shape the map. It does drive really nice on lpg now with no codes so it can't be far off I think.

[Bone9]

Hi guys. I'm back with a couple of questions.

About 10 months ago I removed the complete back end to get access to paint my chassis and replace the rusty old back end of my lpg system.

Over my Christmas break I reinstalled all new components. This included upgrading my gas line and t sizes to 8mm from ¼".

Everything switches over and works but I'm getting the lean codes again. So I plugged it in to have a look and noticed my gas pressure is now 1.9bar! Before it was set at 1.6. The reducer hasn't been touched at all and the adjustment screw is blocked by the coolant bottle so it hasn't accidentally been moved etc.

Could the increased bore of the gas line increase the gas pressure?

The car drives 100% perfect on petrol. The fuel trims on petrol are excellent too.

Cheers.

Toby.

[LPGC]

No it won't be due to wider gas lines.
You did check the pressure reading when it was running on gas? No good checking the pressure when it's running on petrol or the engine isn't running.
Does it stay at 1.9 bar when the engine is fully warmed up?
Is there liquid gas in the tank or is it running on fumes?
Have you connected the reducer's vacuum port to manifold vacuum?

I re-read some of page 1 of this thread, it seems your error codes started soon after the installer serviced / re-calibrated it?

[Bone9]

Hi Simon. Thanks for the reply.

I checked the gas pressure while running on gas and fully warmed up. It settles to 1.7 while driving and drops to 1.6 at wot.

Both tanks are full.

The reducer installation hasn't been touched so is all connected correctly.

It's so long ago now I'm not sure when they started. Imo the map looks wrong. I think I'd rather reset the map to defaults and start a fresh from there. I need to work out why the pressure is different before I can progress.

Thanks

[Brian_H]

You will find as the vapouriser ages, that the springs weaken a bit, so the pressure can creep a bit. Also as the diapgram gets older, it gets stiffer as well. The adjustment screw can be used to compensate for that though. It seems your fairly close to where it should be now, though how does it look at idle while on gas and warmed up?

[Bone9]

Ah I see. That makes sense. The vaporiser is pretty old now and covered over 90k miles. How long do these things last?

When warm at idle its 1.9bar.

I'm going to do a smoke test on the vac lines in the morning.

[Brian_H]

Seems to depend on your luck - keeping a good flow though it and keeping it warm will make it last longer. I've had one last around 60k miles before the diaphrams got very stiff, I've got another of the same model unit which is approaching 100k and has only had to be adjusted once.

Does sound like you need to knock the pressure at idle down though, the vacuum line will operate on the diaphram to compensate for driving condtions so it may change under driving condtions. Its more critical to get it right at idle as you have much less window for error there (1.9bar may mean you can't get the fueling time down short enough to get the right amount of fuel in, depening on the injectors)

A sign of failing vapouriser can be that it leaks gas up the vacuum line to the manifold, so if your pulling pipes off have a look for that (no guarantee but if your already there its worth a look just in case)

[LPGC]

Just had a quick skim through early posts on this thread, I suggested 1.7 bar in August 2021 so if it's close to that now it won't be far off the mark. As Bri says though, 1.9bar at idle may be a tad too high.

I asked about being warmed up because a lot of reducers output a bit more pressure when cold than when warmed up, asked about if there was any liquid gas in the tank because a lot of reducers output a bit higher pressure at idle when they're running on fumes. I forget what reducer you have, maybe it's a palladio or CPR?

[Bone9]

Seems to depend on your luck - keeping a good flow though it and keeping it warm will make it last longer. I've had one last around 60k miles before the diaphrams got very stiff, I've got another of the same model unit which is approaching 100k and has only had to be adjusted once.

Does sound like you need to knock the pressure at idle down though, the vacuum line will operate on the diaphram to compensate for driving condtions so it may change under driving condtions. Its more critical to get it right at idle as you have much less window for error there (1.9bar may mean you can't get the fueling time down short enough to get the right amount of fuel in, depening on the injectors)

A sign of failing vapouriser can be that it leaks gas up the vacuum line to the manifold, so if your pulling pipes off have a look for that (no guarantee but if your already there its worth a look just in case)

Cheers Brian. Makes sense. So I might be good if I adjust the pressure and remap then. As mines just been sitting there not in use for a while maybe that stiffened it up a bit? I do have a repair kit here for it. Are they worth using or is it best to just replace if it comes to that?

I'll check for leaks tomorrow. Thanks.

[Bone9]

Just had a quick skim through early posts on this thread, I suggested 1.7 bar in August 2021 so if it's close to that now it won't be far off the mark. As Bri says though, 1.9bar at idle may be a tad too high.

I asked about being warmed up because a lot of reducers output a bit more pressure when cold than when warmed up, asked about if there was any liquid gas in the tank because a lot of reducers output a bit higher pressure at idle when they're running on fumes. I forget what reducer you have, maybe it's a palladio or CPR?

Gotcha. So of there are no leaks tomorrow should I reduce it to 1.7bar @ idle?

Yes it's a palladio.

[Brian_H]

Cheers Brian. Makes sense. So I might be good if I adjust the pressure and remap then. As mines just been sitting there not in use for a while maybe that stiffened it up a bit? I do have a repair kit here for it. Are they worth using or is it best to just replace if it comes to that?
I'll check for leaks tomorrow. Thanks.

Depends how easily it comes apart - some will come apart fairly easily, others will have every fixing seized solid if your unlucky.

Also depends how much a replacement would set you back - I rebuilt one of my Tomosetto AT07 units without too much problem other than the temp sensor snapping off in the casting, Those are known for blowing seals when they get too cold and aren't a particually expensive vapouriser anyway.

Might be overthinking it though - see if its faulty first. The one I had which the diaphrams went stiff on wouldn't control pressure properly anymore. And literally you could hold the diaphram and it was like holding a piece of cardboard, it would hold a level line if you held it in the air rather than flexing. It had done around 4 years at that point and somewhere around 60k ish (would actually have been more, though can't tell how much more as that car had a faulty speed sensor so didn't always record mileage!)

[Bone9]

Yeah good point. For the hassle it might be worth just replacing it. Obviously only if needed.

A new one is £121.

[Bone9]

Right guys. Tested all the vac line etc for any leaks with a smoke machine. They are all perfect So I thought i'd lower the pressure. Had it ticking over for ages on gas and plugged it in. The pressure is back to 1.6bar! Exactly were it was before the backend got removed. Happy days. So I left the pressure alone and drove home. The trims looked a bit better and I made a +7% adjustment on the 12.5 & 15 lines during the drive home. This stopped the lean codes coming on so far.

Also fitted a new set of plugs.

So the main objective now is to fine tune the map and getting it looking correct. I don't like how lean the first line is and the first 2 boxes on the 18 row. The 10 row looks like it needs to go up just by looking at the map(not checked while driving yet).

Simon are my map references ok? RPM seams to be every 1000 instead of every 500rpm as per the reset default map. The T.inj are different too.

[LPGC]

Simon are my map references ok? RPM seams to be every 1000 instead of every 500rpm as per the reset default map. The T.inj are different too.

My definition of 'map references' would normally be the same as 'map reference points' which would be the rpm scale 1000,2000,3000,4000,5000,6000 and the pinj scale 2.5, 3.5, 4, 4.5 etc. Those reference points can be changed by clicking 'modify map refs'. Changing the reference points might allow for a bit more polish with the final results but changing them isn't really necessary. Anyway I don't think you'll have meant this, you'll have meant the figures in the boxes?

It can be difficult to draw conclusions from here because I can't see how the engine responds (fuel trims etc) so I don't know if the figures you've entered would be the figures I'd end up with if I calibrated your install with the parts (and pressure / settings) you have fitted. What I can say is that if your map is correct for the parts you have fitted and pressures etc, then it points to nozzles being a bit on the small side - but we know nozzles are likely to be a bit on the small side given the type of injectors you have fitted. So summing this bit up, with the parts you've got if the map you've entered gives good fuel trims across engine loads and rpms it will be correct (at least for closed loop operation).

The top row will probably never be in effect on your Discovery install (the ball never goes on the top row)? In which case it wouldn't really matter what figures were in it. But that's not completely true because the ball (pinj) will go slightly above the middle of the 3.5 row in certain conditions (over-run etc) and then the multiplier figure that the ECU uses is interpolated from the figure on the 3.5 row and the figure on the 2.5 row. I'd expect to set the 2.5 row by trying to drive in over-run conditions with the ball part way between the 3.5row and 2.5 row and adjust the 2.5 row until over-run fuelling seemed correct. But there's another couple of aspects that can be at play... The ECU may impose a minimum gas injector pulse time for the type of injectors that are set in settings, if that is the case then entering very low numbers in the 2.5 row wouldn't lean the fuelling on LPG because the minimum pulse length would overrule the figure you entered anyway. The other aspect is that if the engine features extra injection and extra injection filtering hasn't been setup properly in the LPG ECU then during extra injection the ball might flick into the top row, in which case the numbers entered in top row boxes can affect fuelling during extra injection... But you shouldn't really set (or need to set) figures in the top row to affect how it responds during extra injection because you should instead use extra injection filtering. In the case of your late Disco extra injection shouldn't need addressing but you may still need to adjust the 2.5 row due to interpolation.

The numbers in the boxes for 1000rpm 15ms and 18ms will only really be applied when you put your foot down at very low engine rpm, and then only momentarily because it's an automatic gearbox and stamping on the throttle will soon see rpm climb above 1000rpm. An installer might lean those boxes as a workaround if there's a misfire or hesitation on sudden booting it from idle. Misfire/hesitation could be due to too rich mixture or weak ignition system that doesn't like running correct tip-in mixture on LPG at low rpm. On an older AEB install the need to lean those boxes could point to the real issue being too much tip-in enrichment being set in other LPG settings but that wouldn't apply to the more recent Emer system. Another possibility concerns reducer pressure... Pressure can respond differently during sudden tip-in, e.g. there might be a sudden drop in pressure before it rises to more normal readings again, or if pressure would normally fall at higher rpms when you've got your foot down it could be that pressure is higher during sudden tip-in from idle rpm than in most conditions when you've got your foot down. The ECU compensates for pressure but the boxes we're talking about will only ever see momentary use during quickly changing dynamics such as pressure and rpm, so there can be a situation when figures in those boxes are changed to give correct momentary fuelling during the quickly changing dynamics. Aspects that can cause momentary incorrect readings include vacuum pipe lengths to the reducer and pressure sensor.

But you've been adjusting figures with percentage changes... Try changing the figures in the 1000rpm 15ms and 18ms to closer to be more like the 1500rpm 15ms and 18ms figures and see if there's any changes in how crisply the engine responds to sudden tip-in. Also try changing the 2.5ms row to have figures closer to the 3.5ms row and see how that effects things.

[Bone9]

Hi Simon that is exactly what I was talking about. I was wondering if I'd benefit on having an adjustment box every 500rpm instead of every 1000rpm. Like you say a finer adjustment. I did look but unsure how to increase the number of boxes for the rpm reference points. My saved map only has 6 instead of 12. Please see the attached screenshots.

My map is a current work in progress so I expect the shape will change a bit. So far I've made adjustments only based on the short term fuel trims. I will have another play tomorrow as it's my day off. I will get a helper soon to drive so I can compare the T.inj petrol info and hopefully fine tune it some more.

I'm unsure If the top line comes into play atm but will check and if it does should I match the t.inj or does this line need to be set with the lambda voltage? I vaguely remember Mick saying this should be lean so I don't waste fuel etc.

Right I understand the possible reasons for the 1000rpm 15 and 18 boxes. My ignition system is now top notch so will have a play and test with adding more fuel in those boxes and also the first line.

So the 18 line is open loop and has to be checked by lambda voltage. Correct? What voltage should I see at WOT?

Thanks for all the great info Simon, it's been a massive help.

[LPGC]

I should advise that if you change the reference points the numbers in the cells don't move with the reference points... So to keep the same calibration settings (and not effectively change the shape of the map) you'd take a screenshot of your map, change the reference points and then re-enter cell values in new positions that match what they were with the old reference points.

For this engine I'd set reference points for RPM to be something like 500, 900, 1500, 2500, 4000, 6000. Load points aren't too bad but I'd probably change them to 2, 2.5, 3, 3.5, 4, 4.5, 5, 6, 8, 10, 14, 18.
If it were a different engine I might enter different reference points. It will probably never usually use the the 2ms row, might not even use the 2.5ms row, but there are reasons (linked to some of what I wrote in the post above) why I'd still use these load references.

You can't really tune just using short term trims, you also need to use long term trims. If this were a new install (so it had only been running on petrol) you might get a very short time during which you could tune the LPG system using only short term trims but your long term trims will already have been steered by running on LPG with not quite right calibration. You should calibrate by switching between fuels while keeping a constant throttle and seeing how that effects trims.

All points in the map can be open loop during certain conditions but for intents and purposes yes the 18ms row will nearly always be open loop, and some of the 14ms cells will also likely be open loop. You want to see rich mixture during high load open loop conditions but not quite as rich as running on petrol. The thing is that narrow band lambda probes can't measure mixtures that are slightly richer than chemically correct and the mixture needs to be richer than they can accurately measure when running on petrol. For running on LPG lean it to the point where the lambda sensor starts to read just shy of it's maximum rich mixture reading and then richen it up by a few percent. With the AEB based system keep your foot down for a while because the LPG vapour and reducer temp readings will be cooler with prolonged booting it and the AEB systems tend to over-compensate for temperature (you don't want it to ever go lean when booting it even during long periods of booting it, don't want over temperature compensation to lean it off while booting it for a prolonged time).

Some engine management systems run open loop during over-run conditions, some run closed loop (obviously only while still providing some fuelling), some run open loop when you first lift off the throttle but then switch to closed loop during extended over-run. You basically want the LPG system to deliver the same amount of fuel that the petrol system would deliver in like for like conditions, including during over-run - You want fuel trims (in case of closed loop) or lambda readings (case of open loop) to be around the same on LPG as they are on petrol during overrun.

One of the reasons why there are often smaller numbers in the fuelling table / map for very low loads (for small pinjs) is that gas injectors take time to open/close.. In simple terms (and discounting lots of other factors) the peak multiplier (max cell value) normally occurs at around the point that the gas injector response becomes linear (has time to fully open before having power turned off and closes itself). If we pulse the gas injector for less than that time we will cut it's power before it is fully open but the momentum of the injector internals may see it continue to open further, so it overshoots the amount of fuel the electronics might anticipate it will flow. So for pulse times below the peak multiplier point the map figures are smaller because the injector over-shoots, and for pulse times above the peak multiplier point the map figures are smaller because the 'dead time' (or time it takes for the injector to start moving at all from closed towards open) and the partial flow time (an injector won't flow as much when it's plunger is only slightly off it's seat as it does when the plunger is fully open) become a smaller percentage of total pulse time.

If we fit small nozzles on an injector that would otherwise flow a lot of gas if it had big nozzles fitted it can have the effect of increasing the overshoot for extremely short pulse durations, because there's a bit of volume between the injector internals and the nozzle and that volume is going to go through the nozzle anyway. E.g. If we were converting a 27ltr V12 Rolls Royce Merlin engine using 1 injector per cylinder we might want 6mm nozzles, there'd be no point fitting 6mm nozzles on injectors that only had 3mm internals because they'd still only flow as much as 3mm nozzles, we'd want the injectors to have 3mm internals. But if we took those big injectors and fitted them on a normal engine and fitted 3mm nozzles on them then for very short pulse durations there'd still be a massive dose of gas getting out of the injector internals, and it would still go through the 3mm nozzles, so we might try to compensate for that by entering smaller numbers in the very low load area of the map but we would still never be able to get it very accurate and we'd probably find that even for the smallest amount of injector opening (plunger off seat) the injector would still flow far too much gas for low loads for the small engine, so we'd try to decrease the numbers in the map further and then suddenly there'd come a point where the injector didn't open at all for the pulse length and the engine would get no fuel... in that situation we'd either have too much fuel for the low load condition or no fuel at all, the big flow injectors even with small nozzles would be too inaccurate at low pulse durations on the small engine. But on the big engine that does need 6mm nozzles those big flow injectors would be just about as accurate at small loads (small loads for the big engine) as the smaller injectors on the small injector at low loads.

[Bone9]

Hi Simon. Thanks again for a detailed reply and the advice about changing the reference points and re entering them in the correct boxes etc. My rpm boxes have changed to every 500rpm now which is excellent. I'm slightly reluctant to change the load points atm as my peak is at 7 and you don't include that in your points. On over run it goes off the scale past my current 2.5 row.

I only had 3o mins spare for a tune today but have improved it further. This time I watched the petrol timings and matched them while switching between gas and petrol. Its smoothed the map out more which is good. I added to the 2.5 row to bring it closer to the 3.5 row as it was well into that region when on overrun. Also increased the 10 line which smoothed that out. When I was tuning with the trims only I was monitoring the short and long not just the short. The long never went more than 5 out.

At wot my stft lean out. It seems my lambdas aren't connected to the lpg software so I need to find another way to monitor them at the same time as watching my lpg map. Any suggestions? I do have a Bluetooth odb2 adapter that I use for torque pro.

My current map.

[LPGC]

Your peak is at 7... Have you set it up as full group injection in settings?

Is it a Thor or Gems engine? Gilbert or Bri might chip in here! I thought all 2003 petrol Discos had the sequential (not full group) fuel system, in fact a bit more advanced than the P38 system with the inclusion of EGR? I'm basing that on old memories though.

Fuel trims may well go to zero during open loop conditions, especially short term trims.

You can set up most OBD scan tools to simultaneously monitor fuel system status (open/closed loop), fuel trims and lambda voltage readings. You don't really need the status because it's obvious they've gone open loop whens stfts stay at zero while booting it etc.

[Brian_H]

I believe all disco 2 v8s are Thor types? You or Gillbertd would likely know better about that. Mine is pre face-lift 2000 x reg and that is a Thor if that's any help?

I'd assume Thor to be fully sequential?

[LPGC]

You think same as me then Bri, and Thor is sequential.

But 7ms peak pinj implies it's set up as full group in LPG software.