Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[chrisgixer]

Where can I get one, why do I need one, how will it help me, is it the best way forward, etc etc...?

I don't have any major problems with the lpg as such. Day to day its fine. Average fuel trim is +5 and +4 on each bank and I (or Kev to be more accurate) haven't touched it for months. So on average there's no issues really.

But being fussy, which we have every right to be;

1) it hunts slightly at tick-over. No big issue, it takes a good few minutes of  stationary tick over to build up a slight hunt on mine. Some of the v engines with lpg have it worse than others on the newer stag kit. Old kit, I believe, isn't a problem(?)

2) avoiding lag on auto gear changes at wot and general fuelling at max injector duration.
On my first car the Lag was pretty horrific as TB will testify when he drove it at Newent. It was tuned out to an extent, but it would still rear its head from time to time.
 2nd and my current car has the newer stag plus ecu and slightly better injectors, which was much better re lag. And was almost 100% lag free. Excellent. ... but the tune tailed off for sone reason, giving heavy lag, and "we" (Kev and I) couldn't quite get it back. During mapping on the road we also noticed....

3) a problem with the petrol map which gave a slight lag on liftoff from wot then back on again at high rpm (in 3rd iirc). Also, lately, with trims at the previously mentioned average of +5 and +4 on cold mornings just above 0c I get lag on gear changes ON PETROL, before it changes to lpg. Engine temp gauge shows the needle just lifting off the bottom.



So, If wanting to sort these minor gremlins, what to do.

First thought for me, is, given the lpg is a piggy back system running off the petrol map, ecu and fuel trims etc (or whatever) its important that there are no gremlins in the petrol map. Fair enough most people won't notice the on off lag on petrol at wot flat out in third. We're getting on for serious speeds there, on private land , and most people won't encounter it at all.
 But, add in the lag on petrol occasionally, that makes me think the standard petrol map could do with a tweak. ...Or is there a fault? Has the lpg dragged the fuel trims off a bit to cause the lag. What's pulled the trims out if so? And so on.

Aaaaanyway.
My particular car has upgraded Hanna injectors rated to 300bhp, they are Prins copies and do all their injector timing on duration, there are no nozzles to drill . While the part throttle settings are crisp and pull strongly, there is no lag at all, and the mid range is noticeably strong.... Wot is too rich, as far as I can gather the petrol map is calling for maximum duration on correctly sized petrol injectors, and the over size lpg injectors do the same and are then running it too rich. Ok for short bursts, but leave it at Wot for too long and it will eventually miss fire due to excess fuel and the eml comes on. All the miss fire codes will be present as will fuel trim codes iirc stop, re start, and all is well until prolonged wot again.
 So I've built in a fault there myself but I can drive round that and it's not where the car spends most if its life. However there are some throttle settings that approach wot, where the car feels sluggish by comparison. Back off the throttle and engine response improves. Again, it's too rich, that could be mapped out.

So reading through all that its evident there's a lot of seat of the pants and guess work going on. I want to re map the petrol side of things so the lpg has a better base to work from then be able to map the lpg and know exactly what the fuelling is doing. While the lpg software is good at setting up initially its not helpful when trying to sort the finer points.

On speaking to Kev, a wide band lambda sensor is needed. Although is that going to be necessary if re mapping on petrol if the lpg can be set up on the day? I guess a rolling road will be needed and presume they will have ways of checking the fuelling accurately themselves....?

[Marks DTM Calib]

Bottom line is that the LPG ECU's are not upto the job and dont have the inputs required to do a 'proper job'.

I cant see that you can get a step change in setup with a wideband 02 given the lack of TPS input etc.

It may help tweak the setup but little more.

Now if you were setting up a full injection setup with all the usual inputs then yes, a big help.

[chrisgixer]

Bottom line is that the LPG ECU's are not upto the job and dont have the inputs required to do a 'proper job'.

I cant see that you can get a step change in setup with a wideband 02 given the lack of TPS input etc.

It may help tweak the setup but little more.

Now if you were setting up a full injection setup with all the usual inputs then yes, a big help.

That's pretty much where we're at Mark tbh. Just need that extra little bit of info.

[Marks DTM Calib]

Trouble is, how does the LPG ECU know your at WOT.....without a TPS input

So how can you address the over fueling at WOT

[Kevin Wood]

Yes, a piggy back ECU is a major bodge, really. That they work as well as they do is a surprise, in my view.

In addition, where you talk about lag issues when coming on and off the throttle, a lambda isn't going to give you a picture here because it simply doesn't respond fast enough. The petrol ECU is seeing gearbox retard and throttle position in real-time (and on a DBW it's controlling the throttle!) yet, still, it's not perfect on petrol let alone LPG.

The LPG ECU is only following injector durations that it's observing from the engine ECU, so it's at least a complete engine cycle behind what's actually happening, probably quite a bit more. Consider that it has to measure petrol injector duration, calculate from the map and live data inputs the corresponding LPG injector duration and then actually time the opening of the LPG injector.

In addition, LPG injectors are a lot slower than petrol injectors. The volume of fuel they must pass dictates a much larger valve and ditto the moving mass behind it. At 6000 RPM we have 20ms to open the injector, flow enough fuel, and close the injector again for each engine cycle. With an injector that takes 4 or 5 ms to open and close, it's never going to give precise control of fuelling. All we can hope is that, when it gives up and reverts to 100% open, the fuel mixture is "about right".

A wideband lambda would certainly help see what's causing the misfire at wide open throttle but I don't think there's a solution to much of the problem. The petrol ECU is a black box which you can't get inside and map, and the LPG ECU is simply a sow's ear adjustable via a really blunt instrument to be acceptable for "normal" driving.

Frankly, I think switching back to petrol over 5,000 RPM is probably the best and cheapest solution.

[cem_devecioglu]

Yes, a piggy back ECU is a major bodge, really. That they work as well as they do is a surprise, in my view.

In addition, where you talk about lag issues when coming on and off the throttle, a lambda isn't going to give you a picture here because it simply doesn't respond fast enough. The petrol ECU is seeing gearbox retard and throttle position in real-time (and on a DBW it's controlling the throttle!) yet, still, it's not perfect on petrol let alone LPG.

The LPG ECU is only following injector durations that it's observing from the engine ECU, so it's at least a complete engine cycle behind what's actually happening, probably quite a bit more. Consider that it has to measure petrol injector duration, calculate from the map and live data inputs the corresponding LPG injector duration and then actually time the opening of the LPG injector.

In addition, LPG injectors are a lot slower than petrol injectors. The volume of fuel they must pass dictates a much larger valve and ditto the moving mass behind it. At 6000 RPM we have 20ms to open the injector, flow enough fuel, and close the injector again for each engine cycle. With an injector that takes 4 or 5 ms to open and close, it's never going to give precise control of fuelling. All we can hope is that, when it gives up and reverts to 100% open, the fuel mixture is "about right".

A wideband lambda would certainly help see what's causing the misfire at wide open throttle but I don't think there's a solution to much of the problem. The petrol ECU is a black box which you can't get inside and map, and the LPG ECU is simply a sow's ear adjustable via a really blunt instrument to be acceptable for "normal" driving.

Frankly, I think switching back to petrol over 5,000 RPM is probably the best and cheapest solution.

very good write up Kevin


the only set that doesnt create problem at high revs/powers are Prins kits as far as I know..  they have the best injectors  (Keihin).. however I agree with Kevin,  lpg is for economy if you are going to drive above 5k rpm using normal fuel is a better approach than purchasing an expensive kit.. the friends omega with brc kit also switches to normal fuel at high rpms


ps: also prins supplies an expensive direct injection kit but never seen any installation of these kits on omega

[Kevin Wood]

Of course, the best solution would be to replace the petrol ECU with a true dual fuel ECU that can cope with LPG and petrol natively.

That's a different project for every engine, though, and keeping all the other systems happy via CAN bus and other proprietary interfaces such as the fuel flow signal to the MID, etc. and then mapping it properly would make it quite an undertaking just to support 1 vehicle, so it's not hard to see why LPG systems are what they are.

[chrisgixer]

So the 3.2/Dbw ecu can't be mapped?

...and mean time tweak the lower end of the map to try and stop the hunting, if I can be bothered ...and set it to switch back to petrol at 5,500 rpm

....and stop moaning about it.

[zirk]

So the 3.2/Dbw ecu can't be mapped?

...and mean time tweak the lower end of the map to try and stop the hunting, if I can be bothered ...and set it to switch back to petrol at 5,500 rpm

....and stop moaning about it.

Bit off topic, but Im actually in talks with a ECU Mapper I know at the moment, on the possibilities of 3.2 remap, with a bit more grunt at the top end on Petrol and a bit more economy lower down and mids when on LPG, via some sort of auto sensor or via the LPG switch, not sure if that would aid on the above discussion. 

[Kevin Wood]

So the 3.2/Dbw ecu can't be mapped?

Not by us. Maybe there are places that will no doubt offer you a "remap" but we're dealing with "baseball cap on backwards" types then. Are they going take time to map it fastidiously on a rolling road or just upload a map they've messed about with a bit and then send you off with an impressive figure to brag about down the pub? I seriously doubt they'll have the experience to map out glitches of the type you have noticed, even if it's within the limitations of the ECU to do so. In fact, I'd say that GM would have had the engine running in a dyno cell and on the road for months getting the map right, so if it were possible to make it work any better (on petrol)... 

...and mean time tweak the lower end of the map to try and stop the hunting, if I can be bothered ...and set it to switch back to petrol at 5,500 rpm

The thing to remember is that the engine is running closed loop at idle, so:

1) The mixture can't simply be out. Closed loop control will bring it back to where it should be or you'll get fuel trim codes. Tweaking the LPG map only makes a temporary difference until the block learn map in the petrol ECU has updated.

2) The ECU will be making small adjustments to the mixture constantly. On petrol, you can see the idle hunt up and down slightly as this happens. They have set the correction steps small enough that they are only just noticeable. On LPG, who's to say the exact same steps have the same effect on the actual mixture? If they are right at the lower limit of injector duration they make a much bigger impact on the mixture, so the hunting is more noticeable. This has been demonstrated where the Stag injectors have been drilled out a little too large, then engine stumbles below a certain LPG injector duration because the injector doesn't have time to open at all!

....and stop moaning about it.

 

To a certain extent, though, you do need to accept that an LPG conversion has positives and negatives. On the plus side, it's a shed load more economical to run. With that comes the down side: it's a bit of a bodge that will affect some aspects of the engine's refinement.

[Kevin Wood]

So the 3.2/Dbw ecu can't be mapped?

...and mean time tweak the lower end of the map to try and stop the hunting, if I can be bothered ...and set it to switch back to petrol at 5,500 rpm

....and stop moaning about it.

Bit off topic, but Im actually in talks with a ECU Mapper I know at the moment, on the possibilities of 3.2 remap, with a bit more grunt at the top end on Petrol and a bit more economy lower down and mids when on LPG, via some sort of auto sensor or via the LPG switch, not sure if that would aid on the above discussion. 

There is one massive improvement that can be made on LPG running. Advance the timing a good bit. Burning LPG with an ignition map designed for petrol is a big compromise.

[chrisgixer]

Interesting Zirk

..its not lost on me, that if if re mapping to aid the lpg, then it makes sense to explore any improvements on power or economy etc as well. Might give the lpg a better chance at the top end maybe

I know nothing about these things, but if the petrol mapping can er on the side of reducing the amount of time spent on max duration and more controlling injectors times, would that give it more of a chance? I wonder?
 I guess there's only so much that can be done within a single rpm at 6750rpm or whatever it is, though.

Only one way to find out.

[chrisgixer]

So the 3.2/Dbw ecu can't be mapped?

Not by us. Maybe there are places that will no doubt offer you a "remap" but we're dealing with "baseball cap on backwards" types then. Are they going take time to map it fastidiously on a rolling road or just upload a map they've messed about with a bit and then send you off with an impressive figure to brag about down the pub? I seriously doubt they'll have the experience to map out glitches of the type you have noticed, even if it's within the limitations of the ECU to do so. In fact, I'd say that GM would have had the engine running in a dyno cell and on the road for months getting the map right, so if it were possible to make it work any better (on petrol)... 

...and mean time tweak the lower end of the map to try and stop the hunting, if I can be bothered ...and set it to switch back to petrol at 5,500 rpm

The thing to remember is that the engine is running closed loop at idle, so:

1) The mixture can't simply be out. Closed loop control will bring it back to where it should be or you'll get fuel trim codes. Tweaking the LPG map only makes a temporary difference until the block learn map in the petrol ECU has updated.

2) The ECU will be making small adjustments to the mixture constantly. On petrol, you can see the idle hunt up and down slightly as this happens. They have set the correction steps small enough that they are only just noticeable. On LPG, who's to say the exact same steps have the same effect on the actual mixture? If they are right at the lower limit of injector duration they make a much bigger impact on the mixture, so the hunting is more noticeable. This has been demonstrated where the Stag injectors have been drilled out a little too large, then engine stumbles below a certain LPG injector duration because the injector doesn't have time to open at all!

....and stop moaning about it.

 

To a certain extent, though, you do need to accept that an LPG conversion has positives and negatives. On the plus side, it's a shed load more economical to run. With that comes the down side: it's a bit of a bodge that will affect some aspects of the engine's refinement.

Quite so. We have however, improved on the initial kit by some ditance IMO. A natural process has a beginning and an end. I don't think/know we've reached the end, just yet.

If we have got the best out if it without going to great expense then fair enough.

[chrisgixer]

So the 3.2/Dbw ecu can't be mapped?

...and mean time tweak the lower end of the map to try and stop the hunting, if I can be bothered ...and set it to switch back to petrol at 5,500 rpm

....and stop moaning about it.

Bit off topic, but Im actually in talks with a ECU Mapper I know at the moment, on the possibilities of 3.2 remap, with a bit more grunt at the top end on Petrol and a bit more economy lower down and mids when on LPG, via some sort of auto sensor or via the LPG switch, not sure if that would aid on the above discussion. 

There is one massive improvement that can be made on LPG running. Advance the timing a good bit. Burning LPG with an ignition map designed for petrol is a big compromise.

One of these?

http://www.ac.com.pl/en/produkty/12/timing-advance-processors

[TheBoy]

I'd say that excessive hunting at idle (like TBE) is LPG setup, and you need to look at the 3d map for high RPM, given no MAF/TPS feeds into the LPG ECU.

The high RPM lag, when coming off power and immediately back on, with no gear change inbetween will only ever be resolved with a manual box IMHO. The 3.2 TCM is significantly behind the 3.0 TCM in terms of refinement, and I suspect its more than just a TCM map, but a major firmware enhancement, to get it on a par. I'd say outside the scope of anyone affordable.


Motronic remaps are likely to be just that, remaps. Not firmware changes, so doubt the lag on retard will be able to be dialled out, and a piggyback ecu will amplify this.


Sorry, I know its not what you want to hear.