M20B27 on LPG - Developments

[StuBeeDoo]

That's really interesting, can you adjust the idle timing as well to be optimised for LPG then?

Yes, and I have. BUT... I've found that in doing so I can't get the idle speed below about 1300 unless I lower the stepper default on the Leo and doing that I've got a jerk on transition from idle to part-throttle, so I'm going to knock the ignition back a bit and set the Leo as it was before.

With your AFM stuck open on LPG, I take it you only have two ignition points to map for each RPM point, part throttle and WOT?

I could be learning something here. Does the AFM (or MAF) control the ignition, then? I always assumed it had something to do with the fuelling side. TBH, I haven't a clue about AFM/MAF because there's no reference to it anywhere in the software.

Do you know if the software output the AFM/MAP voltage when live mapping?

On idle ignition, the map has 8 rpm points; on P/T it has 16 rpm points and 14 load points; and WOT has a further 16 rpm points. Fuelling maps are the same.

[Brianmoooore]

I could be learning something here. Does the AFM (or MAF) conrtol the ignition, then? I always assumed it had something to do with the fuelling side. TBH, I haven't a clue about AFM/MAF because there's no reference to it anywhere in the software.[/b]

Many years ago I considered fitting a flap opener, and I read somewhere at the time, that the flap position does have some (small?) influence on ignition timing.

[Ant]

AFM flap is the load sensor on these, for both Fuelling and timing

no load sensor, no 3D mapping only 2D + the idle and WOT maps.

to get the best outa this the load info NEEDS to be there and in the correct sense.

without the AFM you may as well fit a Jetronic dizzy back in the block and go proper old school

HTH

[StuBeeDoo]

HTH

It does, Ant - thanks.

That being the case, I think I will convert to MAF.

[StuBeeDoo]

Right! I wanted this to be a "warts 'n' all" thread so here's where the first f***-up gets reported. Seems like I've still got a load more research to do on this. I thought I had it all sussed. I hadn't known that the AFM has an effect on the ignition map, in fact I'd read somewhere that it doesn't. OK, so I can map the ignition for idle, max load p/t and WOT, but having thrown the money at the W.A.R chip I don't really want any more compromises.

It's decision time again:
Do I try to get the LPG to run with the AFM active? Where's the preferred position for a "catflap" in the standard M20 inlet tract? TBH, I don't like the idea of any more restrictions in the inlet.
Maybe I should have bought the MAF to go with the W.A.R chip. Perhaps I should bin the AFM in preference for a MAF. If I do, will I still need a catflap? Are there any MAFs other than the Miller one that can easily be adapted to fit on the E30 air filter casing? A cone filter is a definite no-no.

[Brianmoooore]

Catflaps normally fit directly onto the outlet of the AFM, and are machined on their outlet side to take the rubber air hose. You can also leave the clips at each end of the air hose slightly loose, and can add mushroom valves to the flat area on top of the plenum.

[StuBeeDoo]

I've been out this morning and had a play. The engine won't run with the AFM active, even after clearing the Leo and resetting it. TBH, I had an idea that would be the case.

Catflaps normally fit directly onto the outlet of the AFM

That's where my mixer goes, Brian. IIRC, that's why the engine won't run on LPG with the AFM active - something to do with the AFM affecting the "draw" through the venturi. I'm sure I read somewhere that your mixer is upstream of the AFM, and that's where my flap opener goes.
So now I'm fairly certain that if I want 3D ignition mapping on LPG, I'll have to go for a MAF.

[Brianmoooore]

Mixer has to be atmospheric side of a flap AFM, because the closed flap greatly reduces the pressure at this point at low revs/idle, resulting in the mixture being far too rich. A reference pipe from the big air hose to the air side of the reducer diaphragm (as used with turbo set ups) might work.
The mixers, on the cars I have with single point LPG, are mounted inside the top of the air filter box, replacing the stock inlet trumpet.

[StuBeeDoo]

The mixers, on the cars I have with single point LPG, are mounted inside the top of the air filter box, replacing the stock inlet trumpet.

Perhaps this is something I ought to be looking at. I don't suppose you have a link to a source of supply Brian?

[e301988325i]

Right! I wanted this to be a "warts 'n' all" thread so here's where the first f***-up gets reported. Seems like I've still got a load more research to do on this. I thought I had it all sussed. I hadn't known that the AFM has an effect on the ignition map, in fact I'd read somewhere that it doesn't. OK, so I can map the ignition for idle, max load p/t and WOT, but having thrown the money at the W.A.R chip I don't really want any more compromises.

It's decision time again:
Do I try to get the LPG to run with the AFM active? Where's the preferred position for a "catflap" in the standard M20 inlet tract? TBH, I don't like the idea of any more restrictions in the inlet.
Maybe I should have bought the MAF to go with the W.A.R chip. Perhaps I should bin the AFM in preference for a MAF. If I do, will I still need a catflap? Are there any MAFs other than the Miller one that can easily be adapted to fit on the E30 air filter casing? A cone filter is a definite no-no.

Cat flap cost me 6bhp, and all the installers i spoke to told me they rarely save the AFM anyway.

The only practical MAF with standard airbox is miller IMHO. Get to keep the trumpet inside the airbox too.

I had to give up on stock air filter with MAF AND BLOS, my new air filter will sit underneath the MAf with a large 80mm u-bend. I'm hoping to encase this to some extent inside the old air box, albeit cut largely to pieces. utilising the original cold air feed.

Tinley tech had mixers as described for me, very limited on designs though, I'd go miller MAF with a round od mixer. PS) ebay is very cheap for silicone hose adaptors which are normally availble in black, although I've had to go with red to keep all the same colour.

[e301988325i]

So now I'm fairly certain that if I want 3D ignition mapping on LPG, I'll have to go for a MAF.

That's not strictly true, you could use a MAP sensor (not been done successfully AFAIK).

Or use the switchable Auto TPS which has a potentiometer built in for the autobox ECU, it still retains the idle-part throttle-wot functionality of the original TPS. You wire the output of the TPS into the AFM loom and keep the IATS Intake Air Temp Sensor as normal. This way you could retain an empty AFM (jammed open and standard airbox.

edit - and another thought, you could use an adaptor to mount a cone filter to the front of the AFM, but instead use it to connect a round hose to the stock airbox. I have IIRC one if you require.

[StuBeeDoo]

Alex & Brian, thank you for your input. Some food for thought there.

you could use an adaptor to mount a cone filter to the front of the AFM, but instead use it to connect a round hose to the stock airbox.

That sounds like a possible plan Alex. Nice one!

[StuBeeDoo]

Or use the switchable Auto TPS which has a potentiometer built in for the autobox ECU, it still retains the idle-part throttle-wot functionality of the original TPS. You wire the output of the TPS into the AFM loom and keep the IATS Intake Air Temp Sensor as normal. This way you could reatin an empty AFM (jammed open and standard airbox.

Am I reading this right Alex?

You're saying that the variable TPS could be used to provide the load signal to the ECU and there would be no need for either an AFM or MAF, other than for the IATS function??

[e301988325i]

Or use the switchable Auto TPS which has a potentiometer built in for the autobox ECU, it still retains the idle-part throttle-wot functionality of the original TPS. You wire the output of the TPS into the AFM loom and keep the IATS Intake Air Temp Sensor as normal. This way you could reatin an empty AFM (jammed open and standard airbox.

Am I reading this right Alex?

You're saying that the variable TPS could be used to provide the load signal to the ECU and there would be no need for either an AFM or MAF, other than for the IATS function??

Yes works well, known as alphaN control, although it entails a start from scratch remap. IIRC ANT may do an off the shelf chip for the conversion.

[StuBeeDoo]

Yes works well, known as alphaN control

Ah! Right. I've seen AlphaN mentioned, but always on M3 threads so I assumed it was something specific to M3s.
My Leo has capability to run with a variable TPS so I wonder it if would make any difference to the LPG side of things?

it entails a start from scratch remap

Hopefully that won't be a problem for the W.A.R chip.

I'm away to investigate further.

[e301988325i]

see here for the info required to add variable tps. I a few spare of the electronic component required to drop the voltage, although the cost is loose change anyway

http://www.e30zone.net/modules.php?name ... 0&start=25

[StuBeeDoo]

Yes, I remember reading that ^^^ thread at some point Alex.

Trouble with me is, if something isn't plug'n'play, I lose interest very quickly.
...... And soldering? Forget it - my hands aren't steady enough for it. Brian would probably have heart failure if he saw the electrics on my Leo.

[e301988325i]

It's an option to get someone else to do it of course.

I could of course do this with mine, as it's already fitted, but the piggy back is already programmed for a 325i with the maf it came with, so it would greatly increase te overall cost having the whole map redone.

[StuBeeDoo]

The more I look at this the more I'm convincing myself that IF I'm going to do it it will have to be the Miller MAF. There's no way I can do the Alpha-N mod and I doubt there's anyone within 200 miles of me who is competent enough to tackle it.
I know I'm stubborn, but a cone filter is NOT an option.

[e301988325i]

[gareth]

Can't you use the AFM hose adaptor from a M30 AFM for that?

I've got a spare round here somewhere i think? No spare M20 AFM's to compare with though...

If someone can measure up the M20 AFM i'll see if this can fit it

[StuBeeDoo]

Right! After my rant on my MAF thread last night, I've been reflecting.
The 2.0 ran fine on LPG, it was only when I fitted the first 2.7 that it all started to go wrong.
It's known that before you can get an engine running right on LPG it has to be right on petrol. My 2.7s never have been 100%, in spite of trying 4 different generic chips (from 3 different suppliers) and a standard 2.5 (173 ECU) chip. OK, I knew the 2.5 chip wouldn't be right, but what it did do was confirm that all the 4 2.7 chips were either causing over-fuelling, or pinking (or in one case, both!!). All of the generic 2.7 chips gave similar cold-starting issues (ran like a bag of shit!), but the starting on the 2.5 chip was bang-on. One chip over-fuelled so badly that the only way I could get the car through the MOT emissions check was to submit it on LPG.
What I had wanted to do was to bin the AFM for something better before going on the rollers in 4 weeks time, hence my impatience/frustration, etc.
So, taking one step back, the first thing I'm going to do is forget for the time being that I want rid of the AFM. Then I'll get the W.A.R chip fully mapped on 99RON petrol. Hopefully Richard will have time in the same session to map the idle, full-load p/t and WOT ignition to suit LPG. Then, I can take my time looking into either going MAF or Alpha-N by the easiest DIY route.

[Ant]

Stuart, tweaking the Miller for any AFM/MAF is difficult as they dont allow you access to that section of the maps iirc

the key maps are the 1x8 measured AFM lookup map and the 1x32 calculated % open (AFM flap) in the M1.3 code, these plus the IAT/CLT and TPS data + the RPM data and rate of change data( inc/dec or static throttle ) and the fuel maps themselves are where you may get issues.

increase the PWM on petrol to get more fuel, if emulating on LPG the PWM on the Gas inj will also increase, so would required closed looping for triming puproses or mapping to suit.

LPG/Petrol can be a pita to get both right especially on LPG that uses the pet inj PWM for its trigger , commonly referred to as injection emulator iirc.

I have skimmed a bit there and simplified but I hope ya get the gist.

you will need to map the petrol, then LPG and then possibly revist both to get some of the transient "right" then move onto the timing, beware though, chgange the metering device and ALL that mapping will need tweaking again.

HTH

[e301988325i]

Stuart, tweaking the Miller for any AFM/MAF is difficult as they dont allow you access to that section of the maps iirc

increase the PWM on petrol to get more fuel, if emulating on LPG the PWM on the Gas inj will also increase, so would required closed looping for triming puproses or mapping to suit.

Ant we're both running single point LPG that turns off the petrol injectors and ignores them.

beware though, chgange the metering device and ALL that mapping will need tweaking again.

HTH

ie, put it all at once and get it mapped for that.

The link i posted on your MAF thread is stated to have very similar outputs to an M30 AFM, which is 50% bigger than an M20 AFM, the M30 AFM has an extra 50% of output over the M20 AFM giving very similar voltage outputs to each other at low flow rates.

IIRC it's the MAF that has come with my s/hand piggy back and MAF kit.

[StuBeeDoo]

Stuart, tweaking the Miller for any AFM/MAF is difficult as they dont allow you access to that section of the maps iirc.

So are you saying, then, that their MAF probably has identical outputs to the standard AFM? When you buy the chip they don't ask which you are using, so I assumed that you just mapped the chip to suit the rest of your set-up - AFM, MAF or possibly even Alpha-N.

beware though, chgange the metering device and ALL that mapping will need tweaking again.

ie, put it all at once and get it mapped for that.

Amen to that! It looks like I'll not be doing the conversion before the summer, if at all, so I'm still going ahead with the dyno session next month. That way, if I don't get 'round to the MAF conversion the engine will still be running 100% on petrol with the AFM, which it isn't right now and hasn't been for over 2 years. If I leave things as they are now and never get 'round to the MAF conversion then the W.A.R chip will have been yet another waste of time and money.

The link i posted on your MAF thread is stated to have very similar outputs to an M30 AFM, which is 50% bigger than an M20 AFM, the M30 AFM has an extra 50% of output over the M20 AFM giving very similar voltage outputs to each other at low flow rates.

Yes, I've been looking at that Alex. If I'm going to do it, I think that's the way I'll go. I might even be able to cope with that myself. Thanks for that link, it could well be very useful to me.



My current thinking goes like this.....
There doesn't seem to be anything to be gained from using a MAF that is a lot bigger than the diameter of the TB unless the main priority is for max power on petrol, which mine isn't. Am I wrong? If so, someone please correct me.
If I'm right, then I'm looking at a 70(ish)mm unit such as the M44B19 one Ant mentioned on the MAF thread, or the M50B20 one I was originally thinking about. Also by keeping it smallish, I shouldn't have to go beyond the 325i injectors I'm using, right??

Please keep the comments coming, I'm grateful for them all, but remember that I'm an old man and my brain cells aren't as fast, or as numerous, as yours so keep it as simple as possible or allow me time to catch-up.

[Ant]

Single point kits... PHEW !

makes life a lot simpler then : D

smaller MAF will have more resolution than a bigger MAF , bonus when mapping low rpm part throttle transient areas.

[StuBeeDoo]

Shamelessly copied from my MAF query thread.......

I've taken some measurements this morning.
A 70mm MAF would fit nicely into a modified E30 air filter casing. Coupled with the info in the link Alex posted, it looks like I've got a plan.

[e301988325i]

Stu, the stock E30 TB to AFM boot can be stretched over the 80mm dia of M52 b28 MAF if you replace the AFM. No real bonus of the extra dia in powerr terms for us with the mixer restriction, just that it is stated to give similar outputs to AFM's.

Gareth the M30 adaptor won't work on the M20 AFM, but is exactly the same idea as what I suggested with the cone filter adaptor.

[e301988325i]

Stuart, tweaking the Miller for any AFM/MAF is difficult as they dont allow you access to that section of the maps iirc.

So are you saying, then, that their MAF probably has identical outputs to the standard AFM? When you buy the chip they don't ask which you are using, so I assumed that you just mapped the chip to suit the rest of your set-up - AFM, MAF or possibly even Alpha-N.

FYI the miller MAF comes with a chip to match their MAF!

[StuBeeDoo]

I was on about the W.A.R chip Alex.

[StuBeeDoo]

Single point kits... PHEW !

makes life a lot simpler then : D

I like simple!

[e301988325i]

Ok, rephrase, the miller maf comes with a map for the WAR chip to closely match your engine/ although inherently that's probably available for free.

[StuBeeDoo]

Ok, rephrase, the miller maf comes with a map for the WAR chip to closely match your engine/ although inherently that's probably available for free.

[e301988325i]

Ie there won't be so much mapping to be done buying a known quantity of the Miller MAF.

See here for pics of my intended setup, comments welcome. . .

http://www.e30zone.net/modules.php?name ... 44#2078044

[StuBeeDoo]

Ie there won't be so much mapping to be done buying a known quantity of the Miller MAF.

Valid point.

See here for pics of my intended setup, comments welcome. . .
http://www.e30zone.net/modules.php?name ... 44#2078044

Thanks for that link Alex. I've posted on there.