M20B27 on LPG - Developments

[daimlerman]

Hmmm, triple SU conversion on an M20, go-on, I dare someone lol !

Back carb would foul the servo....but seriously,even motronic would be better!

[StuBeeDoo]

I've been pondering on the W.A.R chip during the night and I think I may have made a mistake in programming it.

I'll try it again at lunchtime, if I've got time.

[StuBeeDoo]

I HAVE NOW RE-LOADED THE BASE FILES THAT CAME WITH THE W.A.R CHIP AND THE ENGINE RUNS PROPERLY.
I WILL UPDATE FURTHER, BUT POSSIBLY NOT UNTIL TOMORROW.

THE ISSUES WERE CAUSED BY ME AND ARE NO REFLECTION ON MILLER AS A COMPANY OR THE QUALITY OF THEIR PRODUCT.

I APOLOGISE IF MY EXPERIENCE THIS PAST WEEKEND HAS BEEN CONSTRUED AS NEGATIVE FEEDBACK.

[Ant]

^^^ not all is lost then !


still cant see how you are supposed to be able to change the AFM for MAF without access to the AFM lookup tables though .........

[StuBeeDoo]

I think I know the answer to ^^^^^ that now Ant. I'll confirm here when I know for certain.

.... And even if I can't get my MAF conversion up and running, all is certainly not lost now. There's even a possibility that I may be able to forgive the slow uploading issue, if I can get the chip to do what I want.

[e301988325i]

Well that's one hell of a U-turn L O L!

[StuBeeDoo]

Well that's one hell of a U-turn L O L!

Isn't it just???

Just goes to show what happens when you (sorry, I) do things wrong.

[StuBeeDoo]

OK, I promised an update on the W.A.R chip.

Since I re-loaded my base map, the engine now starts, runs and drives smoothly on petrol. As per the edits I've made to my posts, the problems I had on the dyno on Saturday were of my own making. There is a possibility that I either didn't read the manual properly, or I didn't digest the information properly. The mixture isn't quite right, but it's sure as hell better than it was.
I'm waiting on an answer from Miller re the MAF maps. At the moment, I don't think that I would ever be able to get any MAF running correctly using the base map I have on the chip currently. What I need to know from them is;-
a) if the maps they have for their MAF will also run the M50 MAF
b) if yes to a), do I need a different base map to run a MAF (I strongly suspect the answer is "yes") and;-
c) if yes to b) have I bolloxed my MAF plan by loading the AFM base onto all 4 maps or can my AFM base map be over-written?
Depending on the answer(s) I get I'll either re-fit the MAF, only better, or refine the mixer-in-front-of-AFM mod, below.

However, I think I may now have the answer to fitting one of the mixers I have in stock inside the air filter casing I originally adapted for the MAF. That is going to be the project for two weeks time.

I couldn't wait, so I've done this this evening. I've now got a 36mm venturi ahead of the AFM. So at least now I should have proper ignition advance at part-throttle.
I checked the Leo before running the car on the road, and the stepper is going up to about 150 as soon as the TPS comes off idle, then drops back. Does that not indicate that the venturi's too big???

I need to take the laptop to work with me tomorrow to check the Leo on the road.

[GeoffBob]

I'm waiting on an answer from Miller re the MAF maps. At the moment, I don't think that I would ever be able to get any MAF running correctly using the base map I have on the chip currently. What I need to know from them is;-
a) if the maps they have for their MAF will also run the M50 MAF
b) if yes to a), do I need a different base map to run a MAF (I strongly suspect the answer is "yes") and;-
c) if yes to b) have I bolloxed my MAF plan by loading the AFM base onto all 4 maps or can my AFM base map be over-written?
Depending on the answer(s) I get I'll either re-fit the MAF, only better, or refine the mixer-in-front-of-AFM mod, below.

Stu, I have been following this quietly in the background noting your progress. Great that you had a breakthrough. It takes a strong man to admit his mistakes.

Would I by right in assuming that your Miller does indeed have a base map for the MAF, but that it is "locked away" where you can't get to it? As per Ant's earlier comment, I cannot see how a MAF can be setup without a base map, and I can only suspect that Miller woudn't want their customers changing the base map so that you are tied to their MAF. I am not sure that they want anyone using their chip/harware with someone elses MAF.

Am I making any sense? Ant, what's your opinion?

[e301988325i]

Stu, I've found tuning a stationary car on lpg is a waste of time and doesn't give the same results as when driving.

[Gunni]

I'm waiting on an answer from Miller re the MAF maps. At the moment, I don't think that I would ever be able to get any MAF running correctly using the base map I have on the chip currently. What I need to know from them is;-
a) if the maps they have for their MAF will also run the M50 MAF
b) if yes to a), do I need a different base map to run a MAF (I strongly suspect the answer is "yes") and;-
c) if yes to b) have I bolloxed my MAF plan by loading the AFM base onto all 4 maps or can my AFM base map be over-written?
Depending on the answer(s) I get I'll either re-fit the MAF, only better, or refine the mixer-in-front-of-AFM mod, below.

Stu, I have been following this quietly in the background noting your progress. Great that you had a breakthrough. It takes a strong man to admit his mistakes.

Would I by right in assuming that your Miller does indeed have a base map for the MAF, but that it is "locked away" where you can't get to it? As per Ant's earlier comment, I cannot see how a MAF can be setup without a base map, and I can only suspect that Miller woudn't want their customers changing the base map so that you are tied to their MAF. I am not sure that they want anyone using their chip/harware with someone elses MAF.



Am I making any sense? Ant, what's your opinion?

As long as he´s got access to the fuel maps the car can be tuned. Despite the MAF curve beeing wrong.

[StuBeeDoo]

Thanks for the continued input lads.

Firstly, as Gunni has pointed out, it should be possible to set the chip up to run a MAF with the AFM base map installed. However, as I found, this will take a lot of dyno time.


Geoff's assumption is correct. There are sections of the mapping that can't be accessed. Miller have specific base maps for both AFM and MAF however the MAF base map is for their MAF and........

You can try our MAF software. However, every different model of MAF is calibrated differently meaning the 0 - 5V signal is scaled to XX kg/hr worth of air differently. Our MAF "transfer function" as it is called is different than an M50. That isn't to say you could try the our software and see if you can massage it out with the fuel and timing maps. We can't offer support for that though if it gets really tough.

I now have a MAF base map which I'll try when I get a chance to refit the MAF.

[Gunni]

Doesn´t have to be dyno time, but yes tuning is required.

[e301988325i]

It's good that miller aren't restricting anything, that would have really annoyed me! I think it would be a good idea to make up a lead to go between the AFM loom plug and the MAF conversion so you can easily swap between the two.

[StuBeeDoo]

It's good that miller aren't restricting anything, that would have really annoyed me! I think it would be a good idea to make up a lead to go between the AFM loom plug and the MAF conversion so you can easily swap between the two.

That was what I had planned but the soldering fell apart (see halfway down page 7).
Now I've got a replacement plug on the loom and matching sockets on the AFM and MAF. Once I've re-done the MAF conversion (better next time!) swapping between the two will simply be a case of unplugging from the loom, loosening the (already loose in case of backfire) hose clip on the air filter end of the inlet elbow, undoing the 2x 10mm nuts that hold the air filter to its bracket and undoing the hose clip on the mixer side of the stepper and pulling the vap. hose off, then lifting the air filter out and swapping the lid/mixer assemblies with their respective AFM of MAF. Refitting will be the same in reverse and remembering to switch the chip to the correct setting.
It will probably take less time to do than it has for me to type it here. When I get 'round to re-doing the MAF malarky (probably next weekend, because we're away this one) I'll time it.
I really want to get it to run on the M50 MAF if I can, but it may take a few goes which is why I need to be able to chop-and-change at will as easily as possible.

Without going to the expense that Chris has with his Innovate thingy, is there any way of getting a reading from the O2 sensor (other than through the Leo)? Irrespective of whether I finally end-up using the AFM or MAF, what I want to do is get the fuelling as near as possible to right before I go back to Motoscope again and being able to use the O2 would be a big help.

[Ant]

Fair play to Brody there.

Agree with Gunni, not having access to that map is not the end of the world, but it makes the whole thing much more time consuming unless you can datalog the AFM vs MAF output as 0-5V , then work out a rough slope vs slope calc out for a rough tweak of the maps to start off with .

If you can tweak the map that the other maps use to calculate fuel and timing, to correct for the metering, then the OE maps are your basemaps as such.

as long as there are 2 of you in a safe T+T'd car, road mapping is a great way to spend less £££ on dyno come the final tune, wideband 02 is a MUST though especially as stoich for LPG vs Pet are not the same.

All good and positive though, the wall has been climbed

[GeoffBob]

Doesn´t have to be dyno time, but yes tuning is required.

It does if you want it done properly.

Agree with Gunni, not having access to that map is not the end of the world, but it makes the whole thing much more time consuming unless you can datalog the AFM vs MAF output as 0-5V , then work out a rough slope vs slope calc out for a rough tweak of the maps to start off with .

I would have thought that it is a question of resolution. You would ideally like the output of your "load" sensor (whether it be AFM, MAF, MAP or TPS) to be scaled so that the maximum load value (corresponding to maximum flow, pressure, WOT etc) is located near the top of the fuel & ignition maps, with minimum at the bottom.

Yes, I agree that you can retune so that the fuelling and ignition values are correct at that given load-RPM point (corresponding to whatever output is coming from the "load" sensor), but if you swap an AFM out for a MAF (as an example) and the two happen to have widely differing scales you'll a) have a whole load of retuning of both the fuel and ignition maps to do, and b) you'll have questionable resolution. This is particularly prevalent when the replacement AFM goes full-scale (typically 5V) at a much higher value of air flow (intended for a bigger capacity engine say) than say the original AFM does. This problem is, of course, not immune to at least partial solution by means of an op-amp in order to scale the analogue signal (the output) of the AFM. At least then a minimum of retuning would be required and the resolution of the ECU would not be compromised.

Of course, it's all the better if you have access to the base MAP, which it appears that Stu indeed has.

[GeoffBob]

Stu, since you have stated that you do indeed have access to the Miller base map, I'd recommend retuning the base MAP over anything else. If you had an engine that once ran fine on petrol with an AFM, then there is no reason why you should not be able to rescale the output of any MAF (provided it is capable of measuring your required air-flow) to produce a numeric value (which is input to the fuel and ignition maps) that corresponds to the same value that you would have received from the AFM at that same air flow.

If I may be so bold as to earnestly suggest the following:
1) Find out at what value of mass air flow (kilogram of air per second) the Miller MAF reaches full scale.
2) Similarly, find out at what value of mass air flow the M50 MAF reaches full scale.
3) Divide the 2nd number by the 1st number to determine your ”ascaling factor”a, call it k.

Now, go through your Miller base map (the one they sent you for your engine with their MAF) and scale the values by the factor k. In other words, take each value that you see before you, multiply it on a calculator by the value k, and re-enter the value (nearest integer) in the table. This process won’t iron out any non-linearity between the M50 and Miller transfer functions, but you’ll find it one hell of a lot easier to iron out those differences later (hopefully on a dyno), than you will trying to reprogram the base-map from scratch! In other words, at least you’ll hit the ground running, as opposed to not even being able to start the car due to a non-existent idle (for example)

HTH
Geoff

[e301988325i]

Without going to the expense that Chris has with his Innovate thingy, is there any way of getting a reading from the O2 sensor (other than through the Leo)? Irrespective of whether I finally end-up using the AFM or MAF, what I want to do is get the fuelling as near as possible to right before I go back to Motoscope again and being able to use the O2 would be a big help.[/b]

There are gauges available IIRC, but the 1V output is not very usefull for lighting bulbs etc, a multimeter would probably be too slow as well I think.

Geoff/Gunni,

There's a website E30SA (South Africa) that details using the M52 siemens 80mm MAF in place of the AFM, which apparantly has very similar oputputs to the M30 AFM. As far as I know the M50 MAF that Stu is using has the same hot wire insert in the MAF body that the M52 has and I guess should provide comparable outputs. It's also worthyof note, that the MAF Stu is using is from a similar HP power vehicle and hence the values at minimum and maximum Kg/Hr shouldn't be too far out.

Really glad to have both of your input on this!

[HairyScreech]

just given my mate paul a shout about what wide band lambda he has just fitted to his v6, as iirc it came in at ~£160 with bosch sensor, loom, controller and gauge out put, iirc it was compatible with the megasquirt wide band input.
think it came from either oz or south africa.

will let you know what it was when he replys as i remember thinking its a good deal cheaper than the lc1.

[GeoffBob]

just given my mate paul a shout about what wide band lambda he has just fitted to his v6, as iirc it came in at ~£160 with bosch sensor, loom, controller and gauge out put, iirc it was compatible with the megasquirt wide band input.
think it came from either oz or south africa.

If it's the "Wide Band Controller" made by Perfect Power in SA then I wouldn't touch it with a bargepole. I had one on my car and eventually had to pull it out. The microprocessor internal to the unit kept freezing (a PIC I suspect) and had to be re-booted (by turning the engine off) each time. It was a total PITA. They happily swapped it for me, but the replacement did exactly the same. If it's an Autronic product from OZ then I have only heard good things about them.

I am currently using a wide band AFR unit from Stack that I am very happy with. Came with a Bosch LSU4.2 wide band oxygen sensor and a very nice 52mm guage with some basic data-logging features built in. Was over £200 though iirc.

[HairyScreech]

nope, but i will remember that if anyone ever mentions perfect power.

apparently its a "tech edge" and there Australian, it was bought to go on his autronic so there may be some relation there, he says he got it through there uk reseller on ebay.

http://www.techedge.com.au/vehicle/wbo2/

http://wbo2.com/

cant find the reseller so not sure what they cost in blighty now.

[CHR1S1990]

My LC-1 kit (controller, wires, gauge etc) was £100 off US ebay, its now £100 plus postage but still damn cheap

[StuBeeDoo]

OK, so things have moved on a bit since I last posted here. When I first fitted the MAF I kept the 39mm venturi I’d used with the original set-up and mounted it between the MAF and TB. With the new set-up it made the stepper open and then drop back, making me think it’s now too big. The difference in performance was unbelievable but the LPG consumption was abysmal, to say the least.

So, I took the whole lot off and refitted the AFM but, this time, with a 36mm venturi’d mixer mounted before the AFM in the air filter and no flap opener. I drove the car 800 miles or so with this set up. It ran reasonably well, but acted as if this venturi was a just bit too big as well ”“ coming off idle the stepper shot up about 100 steps and took a while to drop back to a point where it settled. It didn’t do this when the same mixer was mounted between the AFM and the throttle body, with the flap opener fitted.
The fuel consumption was roughly what I’d come to expect from the 36mm venturi on a run, but slightly worse commuting.

After this, I decided to have another go at the MAF conversion as I now had a couple of MAF base tunes for the W.A.R chip. This time I fitted a 36mm venturi in the mixer. Again, the stepper shot up beyond normal running point, but not quite as far as before. Again, the performance was phenomenal but the LPG consumption was poor ”“ although neither to the extent that they were with the 39mm venturi.

The latest on this is that I have limited the amount the stepper rises by activating the ”afull throttle”a function. I don’t actually know if this is supposed to work, but it seems to. I’ve set it to about 10 steps above the point the stepper hunts around when cruising. It seems to have dulled the performance a bit, but hopefully it will improve the consumption fractionally.

I’ll give the current set-up a week or so and then I’ll maybe try the 34mm venturi I have here ”“ the smallest I have.

So, some conclusions. For some reason;-
1) On an intake with an AFM, an LPG mixer mounted before the AFM needs to have a smaller venturi than one mounted after it.
2) Comparing a MAF-type intake to an AFM-type one (with a flap opener), with the mixer mounted at the same point in the intake the MAF system requires a smaller venturi than the AFM system.
I can only guess that the reason for this is something to do with the air velocity through the AFM’s 52x52mm throat. The MAF I’m using has over 15% more CSA than the AFM and with the mixer in front of the AFM it’s not actually in the inlet tract at all.

So now, more than ever, I feel that my next move will have to be a BLOS, once I get some positive feedback from Alex and/or Chris.



”¦”¦ And now to the petrol side of things:-

I’m slowly getting somewhere with getting the fuelling sorted. Hopefully, I’ll be able to speed-up the process now that I’ve bought an Innovate LC-1. I could have done without the expense but it cost me less than another 2 hours dyno time would, so with a bit of luck all I’ll need is enough petrol in the tank to get the fuelling mapped out on the road and then some more to do the ignition. I just need to read-up on how to use the Innovate when it gets here and find some info on how to get the timing right. All that will leave will be a power run.
I’m hoping to be able to get the Leo to run off the Innovate’s wideband sensor so that I don’t have to keep swapping sensors.

[Ant]

LC-1 has WB02 and simulated NB02 Stuart, no worries for the Leo there.

[CHR1S1990]

as above stu ^ works fine, however i have now binned off the stepper motor after fitting the BLOS.

I must say, the wideband goes hand in hand with the BLOS, took me litterally 5 minutes to setup on idle, then very little time on the road once all warmed up.

it currently runs between 15 and 16 AFR at idle, about 15 AFR when cruising and when i give it full beans between 14 and 15 - sometimes as low as 13. Throttle response is much better. I cant really comment much on economy just yet. Performance wise is does seem to have more poke, but it's hard to tell as I have also just fitted a BTB3 manifold and found a more suitable (larger pipe diameter) water feed for the vap so i get no fuel starvation at all in such cold weather. BTB btw sounds amazing, definite performance increase though it has moved the power up the rev range by about 500rpm maybe more - not really ideal on the b20 as it stands but will be perfect on the b27 soon to go in.

all in all, i would definitley recommend the BLOS. And i would also recommend taking off the stepper motor (leave it in the boot) as the BLOS is pretty damn accurate at getting the desired AFR and renders the stepper motor perhaps only a restriction. Also to note, it has never started better on LPG than with the BLOS - it also idles mucchhhh better, doesnt "search" as much as it did with the "large" mixer i had - so the best of both worlds in this case (small and large mixer benefits)

[damageandyy]

looks like ill be buying a BLOS then, do you think a wide band is definitely needed for setting up or could it still be done pretty well using a normal air fuel ratio meter?

[mr_dink]

Chris, that's great news on the BLOS. Are you using it with an AFM opener?

I've been watching this thread and Stu's development thread with interest as I've recently picked up an e28 525e (m20b27) which I was planning to convert to LPG and was waiting for some feedback on the BLOS.

I'm hoping to keep down the cost and complexity of the conversion so wasn't planning to make any modifications to the AFM. Neither the ebay listing or this info sheet mention that an AFM opener should be neccessary. Yes I accept that it will cause a restriction but I'm after economy more than power with this engine. Rev limit is only approx 4500 so would hopefully show up any restrictions less.

If you are using an AFM opener is there any easy way you could try it without to see if there are any noticeable running problems?

Thanks!

[e301988325i]

Chris, that's really good news. How accurately does BLOS control the AFR under load and cruising?

Stu, your findings with mixer size with AFM and MAF are exactly as I'd expect.

[CHR1S1990]

no need for an AFM opener mr_dink in my application, ive mounted it before the AFM (onto the air filter). It really is as simple as that.

Its surprisingly quite good alex, on part throttle ive managed to get it between 15 and 16 AFR, full throttle it goes somewhere between 13.8 and 14.5 AFR. You can also literally start it up and drive away on a morning - i doubt i'll be touching petrol again for a while.

[mr_dink]

no need for an AFM opener mr_dink in my application, ive mounted it before the AFM (onto the air filter). It really is as simple as that.

Its surprisingly quite good alex, on part throttle ive managed to get it between 15 and 16 AFR, full throttle it goes somewhere between 13.8 and 14.5 AFR. You can also literally start it up and drive away on a morning - i doubt i'll be touching petrol again for a while.

So is the BLOS fitted inside an airbox or are you using a cone filter?

Would appreciate some pics of your set up when you get a chance.

Thanks

[CHR1S1990]

I'm currently using a cone filter, without a heatshield (dirty i know). In the process of making something up out of carbon fibre. Heres a pic when i was assembling it the other day



At present it is fitted as above ^ pointing downwards behind the light, but it will be mounted using a bracket on the (m30) AFM, and will have a 90degree silicone hose behind the headlight

[e301988325i]

Looks good Chris, the M30 AFM should work nicely on LPG as any errors won't show up as much as it would with petrol because LPG is more tolerant.

[CHR1S1990]

Yes it runs great with the m30 AFM, a little stretch of the rubber boot with some hot water and on it goes. Only wish i had some actual numbers to back up each individual mod, people rarely do this. I havent really had the chance to test on petrol yet and set the AFM up as my fuel pump has gone (again) and the one i bought off ebay last week was a duffer. Ill probably put a fuel pump cut off switch in today to prevent it happening again (me being to lazy to put some fuel in and the pump siezing whilst it is still running on LPG).

[e301988325i]

You must have the most unrestricted M20 b20 in existance