M20B25 Motronic 1.3 MAP Sensor Conversion

[DmcL]

peiced this together lastnight to experiment with. if it works ok ill go back and wire it up nice and tidy and loom it all up, etc. using a GM 1 bar "039" MAP sensor and standard M20 IAT sensor.



ill be fitting this today and checking voltage outputs from the MAP sensor at varying loads then ill check output of the stock AFM in the same situation to get an idea of how much tuning will be required to hopefully get this up and working. if it does work then it will open the door for upgrading to a 2 or even 3 bar GM sensor for boosted M20's.

[eta]

I want one of those, its a shame you're in N.I. It's a long way from Suffolk.

[e301988325i]

Subscribed. . .

[chu346]

Don't see the point myself?

How much would all this cost?

[DmcL]

for me.. less than £20 lol

if it works well it would be cheaper to build and sell than a MAF conversion and free up space around the intake side of the engine bay for big feeds or extra large filters, etc.

i already have everything to remap the ECU so i get to play with these kinds of things. i had it starting up fine and idling "ok" but revs would hunt at light throttle if i held the pedal in the same position up to about 35-40% of the pedal travel. it free revs ok but sounds a bit off. havent tried driving it yet, want to get it idling and revving in neutral fine first. i think i know why the revs were hunting, probably my spaghetti ball of vac hoses or possibly the vac fitting being closer to the number 1 and 6 inlet runners. think ive figured out a solution tho if that is the case and ill try it out tomorrow to see.. for now heres a little vid showing the engine running with a naked TB elbow that i took earlier.

[youtube][/youtube]

[e301988325i]

Keep up the work, I could do with the space mainly, have you just modified the AFM loom? like you would for the MAF conversion?

[DmcL]

basically yea.. the MAF i used needed a 12v supply so i had to modify a spare ECU to supply 12v to the AFM plug. i could have run a 12v feed from the fuel pump relay or something but it looks cleaner modding a spare ECU.

the MAP sensor works on 5v so its just a matter of wiring up the MAP and IAT sensors and tuning.

going to play with the MAP conversion a bit more today and see if i cant get to the stage of actually being able to drive the car with it.

[DmcL]

spent a rather long time basically totally rebuilding my 173 ECU bin definition file yesterday morning... and afternoon... and night... and into this morning. sleep? bah who needs it.. haha

should make fettling with this MAP sensor a bit more interesting.. hopefully ill have a good update later on and maybe even a test drive if im really lucky.

[Tommek]

i`m subscribed too .... keep us updated

[eta]

Keep us posted.

[DmcL]

still stuck trying to get it to rev and hold a steady engine speed.. might be a bit more tricky than i thought after getting it up and idling so quickly the other day. just have to keep playing with it.

[MillRat]

Wouldn't building an "off the self" MAP conversion be difficult because it requires every induction to sized exactly the same? With a MAF, you control the diameter of the MAF, therefore you can always calculate the flowrate, however if you change the size of the plenum, or intake length, surely the pressure at the MAP sensor changes.... or am I confused?

I am still very interested in your progress. I really do admire you guys who tackle these complicated problems head on. It really does show great resolve and dedication (particularly once it gets sorted).

[DmcL]

dont know tbh.. i think a MAP sensor would probably have issues like that with breathing mods but i figure if i can get it tuned to run solidly if i change anything intake wise i should hopefully just be able to bump whatever map or maps ive had to modify to use the MAP sensor.

i still havent even tried using the toyota/denso sensor i got. might give that a go and see if theres any difference. i know with no vacuum applied the GM sensor outputs about 4.7v whereas the toyota/denso sensor outputs 3.55v.

[e301988325i]

My old renault 19 16v used a map sensor, a fully variable TPS, air intake temp, water temp and lambda, no MAF or AFM. The MAP looked very similar to the one your using.

Looking at the above I'm not sure if a MAP sensor (relatively with only a 3 position tps) on it's own is enough to map from?

[DmcL]

i know alot of renault use the GM sensor which is what im playing with now. havent tried the toyota/denso sensor yet.

i think the problem with MAP and motronic is that there is no variable TPS signal so manual tuning will be needed for basically any changes to engine breathing as better breathing = less vacuum which will alter the output of the MAP sensor and cause the ECU to read different points on the fuel maps.

i think its quite possible to get working but would be very tuning intensive even with small mods like switching from say an airbox to a cone filter even. probably why there are no MAP conversions out there like the Miller MAF and other MAF conversions as when airflow increases with a MAF or stock AFM the signal output also increases, looks like it might basically be the other way about for a MAP sensor, breathing increases and sensor output decreases..

[mighty_mini]

one issue i had to tackle when running ms2extra on my k-series with ITB's was that the map reading would pulse and not give a steady reading.

What i had to to was run a vaccume pipe from each port and plumb them into a collector, this would smooth out the very sensitive reading from the manifold and did the world of good. another idea would be to run a little restricter in the pipe which runs to the map sensor (like some cheap vaccume gauges do) which will smooth out the readings.




funnily enough i found this thread because i was just looking at doing the same myself today. Because if i can get it running on a map sensor i can then use ITB's and standard ECU rather than using my twin 40's

[mighty_mini]

PS - the map sensor, providing that its in a suitable position (I.E. inside the plenum after the throttle body) should be unnafected by such modifications. its an absolute reading of air entering the engine.

The denso map sensor is 0-5v and would be ideal. ill be looking at making a box up with pentomiters to trim down the resistance reading but not read too much into how the ecu uses the information given.

[DmcL]

yea i was thinking the readings were fluctuating, u can hear it in the vid when i hold light constant throttle.

i did try a little restrictor type thing but it didnt do much. what i think is happening is that because the MAP sensor is connected to the vacuum fitting above the rocker cover its closest to cylinders 1 and 6, im assuming the sensor isnt so quick that it will pick up the pressure pulse from both 1 and 6 so i think its picking up the first pulse from either 1 or 6 then essentially missing the second big pulse from the other runner thats closest so its getting lower readings from 2, 3, 4 and 5 then a spike on 1 or 6 if that makes sense..

i might try connecting the sensor to one of the brake booster fittings on the back of the TB as ive got 2 because the car was originally an auto.

i havent tried the denso sensor yet.. both are 5v sensors but at static pressure with no vacuum applied the GM sensor outputs 4.7v whereas the denso sensor outputs 3.55v, im going to try the denso sensor when i get a chance before i try plumbing the vacuum line for the sensor in somewhere else.

btw heres the MAF i was using vs a stock MAF of the same type/part number. the velocity stack is out of a VR6's airbox and fits perfectly on the MAF.

[mighty_mini]

I pulled apart my old rover k-series motorcraft ecu today and extracted the map sensor. This was the only one i had to hand and knowing its for naturally aspirated it should only output for 1bar negative output. As i started from scratch i had to work out the pinout of the barn door first and with that sorted i extended the wires for the IAT sensor then spliced in the map sensor.

I fitted this just behind the throttle body but ideally it should have been further back into the plenum, but for the test this would do.

After the initial startup it was a bit lumpy although it seemed to smooth out once it was off choke. Revs fairly free but the only issue is that from idle to WOT it bog's down a touch. I think a variable resistor to trim the voltages might be my next experiment but i didnt have my multimeter on me to compare the output voltages depending on engine load. Also i need to make sure that ignition timings are the same when running with a map sensor.

Here is a video of the engine running and revving just with the map sensor fitted.


[youtube][/youtube]


I also risked it and took it for a drive. up until 3k it responds the same as it did with the old barn door but after that its a little hesistant. Unfortunatly I dont have my wideband anymore so i cant log the afr's.


If i figure anything else out on my travels i will let you know. unfortunatly bmw's and motronics are new to me. so i dont know how to mod the ecu's as ive only ever worked with Apexi power FC's and megasquirt

[Tommek]

What car does the maf come from? bmw m44 engine ?

[HairyScreech]

i remember something about the tps from auto cars being constantly variable.

having looked at a few ms conversions and m50 management conversions that seems to be the suggestion for a fully variable tps.

with a variable tps you should be able to map alpha-n which should allow for a certain amount of engine breathing variation. as less air entering the engine = less vacuum = less fuel.

[e301988325i]

i remember something about the tps from auto cars being constantly variable.

Correct, you can use the switchable automatic output to replace the AFM flap output and effectively go alpha-n on stock motronic ECU.

[DmcL]

what just by wiring the TPS output to the AFM connection? sounds like it might be easier than trying to run a MAP sensor..

[GeoffBob]

There is no reason why you couldn't map to either a MAP senor (speed-density) or the TPS (Alpha-N). Alpha-N is less popular than either the mass-air-flow or speed-density methods due to the fact that the position of a TPS is by no means a direct indication of how much air is being inducted into the engine. If your barometric pressure changes you effectively need a remap. Alpha-N with barometric compensation (meaning a pressure sensor open to the atmosphere is added to the ECU) can be made to work extremely well, demonstrating brilliant throttle response (there are numerous modern Toyota's on the road that work this way). Take note, however, that both Alpha-N and Speed-Density tuning methods are specific to your cars intake system. Change a component (air-filter, elbow, pipe etc) and you need a re-map.

Which ever you choose, it would probably be easier for you to tune your ECU if you still had your AFR gauge! Without it you have no feedback mechanism to tell you to raise or lower a point on the fuel map.

Dmcl, check out the book Engine Management Advanced Tuning by Greg banish here. This book deals specifically with engine tuning and goes into great detail describing all the different sensors and how they work. At the very least, download chapters 4, 5 & 6 (really not that expensive).



And sorry folks for banging on so often about books. I honestly don’t work for any book publisher whatsoever, but I do buy a lot of these books myself. You learn a hell of a lot more from one book than you can ever trawl through at length over the net and all the forums.

[mighty_mini]

the only issue with using a stock ecu and tps as input would be that it will surely overfuel because it assumes that the maximum volume of air entering the engine at all times. as WOT would be 5v. thats regardless if you can map the motronics yourself thought. as long as you have a afr gauge

I just tried a GM sensor today,at atmospheric pressure i got 2.5v. at idle im getting 1.2v which is near enough to what the barn door sends the ecu. Revving was fine, but holding revs you can tell that it hunts and chokes.

Im putting this little project on hold until i get into mapping motronics, for the mean time i will finish off my DCOE manifold.

[GeoffBob]

[DmcL]

@ mighty_mini, thats what i ran up against and its also in that vid i posted. at constant throttle it hunts, on tunerpro if u hit trace the AFM 1x8 map (also in the vid) the voltage output from the MAP sensor is fluctuating which is probably the cause of the hunting. i tried a little valve in the vac line to the sensor but it didnt really help, by no means impossible but definately not as simple as tuning for a MAF.

@geofbob, cool ill look up that book and see what the damage is for it. as for alpha-n with pressure correction.. couldnt a MAP sensor just be installed but not connected to the engine so when pressure changes the MAP output changes. little head scratching would figure out how to wire the TPS and MAP up for a setup like that. motronic 1.3 (i think some older versions also) in the E30 actually have a barometric pressure compensation map but the cars dont have any baro correction systems fitted.

[GeoffBob]

the voltage output from the MAP sensor is fluctuating which is probably the cause of the hunting.

Actually, the pressure is changing because the engine is hunting. Lets say you have two points on a map at adjacent RPM locations. Lets say the one at lower rpm causes the engine rpm (at idle throttle) to rise slightly (moving over to the adjacent point on the fuel map). This causes the engine to pull slighlty more vacuum (lower absolute pressure) since the rpm have gone up slightly while the throttle postion remains static (at idle position). Now, lets say that at this new (adjacent) point on the fuel map the mixture is such as to cause the engine rpm to fall. Once again, at this lower rpm the manifold pressure will rise (less vacuum). Hence, as your engine RPM hunt up and down, so will your absolute manifold pressure (vacuum). If you were to look at the output signal from your MAP sensor on an oscilloscope you would see it wobbling up and down as the ECU hunts for the one point (between the two) that brings stability to your RPM. Unfortunately, if neither point on the map is set correctly, your engine rpm will be unstable and your ECU will continue to hunt back and forth between the two points on the fuel map.

Take note, however, that if the "wobbliness" of the signal from your map sensor is synchronised to your engine RPM, then it is in fact picking up the pulsations in the manifold pressure as the engine inducts air into each cylinder. Aftermarket ECU's (designed for working with MAP sensors) have a built in filter to electronically filter out these "pulsations" (so that it see's the average pressure, not the modulated pressure). Since your stock ECU was built for use with an AFM (and thus probably doesn't have this filter) then you may want to insert a 4k7ohm resistor and a 100nF cap (to form a simple RC filter) between the output of your MAP sensor (the "signal" line) and the input to your ECU.

Put the resistor in series with the signal line. IE, cut the line and put the 4k7ohm resistor between the two points you just cut. Then put the 100nF capacitor between the input to the ECU and the 0V (ground/chassis) . You could install your RC filter at the output of your MAP module, or at the input to the ECU. Either should be fine.

Take note, however, that your hunting is more than likely due to badly tuned points on your fuel map.

[GeoffBob]

as for alpha-n with pressure correction.. couldnt a MAP sensor just be installed but not connected to the engine so when pressure changes the MAP output changes.

This is exactly how they work. A MAP sensor is little else other than an absolute pressure sensor. In other words, it produces a signal that is a function of the "absolute" pressure that is applied to it. Leave a 1bar MAP sensor "open" to the atmosphere and you effectively have a barometer. In fact, I regularly use my MAP sensor on my manifold as a barometer. I simply read the manifold pressure on the laptop with the ECU on, but the engine off. A MAP sensor left open to the atmosphere can be used to "compensate" an Alpha-N tuned engine so that the engine doesn't go out of tune every time a cold-front blows in or every time you commute from you sea-side chateau to your mountain villa .

[DmcL]

Take note, however, that if the "wobbliness" of the signal from your map sensor is synchronised to your engine RPM, then it is in fact picking up the pulsations in the manifold pressure as the engine inducts air into each cylinder. Aftermarket ECU's (designed for working with MAP sensors) have a built in filter to electronically filter out these "pulsations" (so that it see's the average pressure, not the modulated pressure). Since your stock ECU was built for use with an AFM (and thus probably doesn't have this filter) then you may want to insert a 4k7ohm resistor and a 100nF cap (to form a simple RC filter) between the output of your MAP sensor (the "signal" line) and the input to your ECU.

Put the resistor in series with the signal line. IE, cut the line and put the 4k7ohm resistor between the two points you just cut. Then put the 100nF capacitor between the input to the ECU and the 0V (ground/chassis) . You could install your RC filter at the output of your MAP module, or at the input to the ECU. Either should be fine.

it was fluctuating with the hunting RPM's did it slightly at idle as u can hear in the vid but moreso at constany light throttle, seemed to speed up with the revs to a point then just sounded "off" after that point. im pretty sure thats the problem, i was reading up on MAP sensors from bosch and checking out datasheets, etc. like all the sheets i looked at stated a RC filter may be needed.

i must check out some of the random electronic junk ive neglected to throw out thinking one day i may be able to rob some parts from the stuff.. could well find the capacitor and resistor.

i wonder how hard it would be use an autos variable TPS and wire in a MAP sensor along with it for barometric correction. cant quite think of how that would work regarding wiring them up together so the MAP sensor corrects the TPS output.

[e301988325i]

i remember something about the tps from auto cars being constantly variable.

Correct, you can use the switchable automatic output to replace the AFM flap output and effectively go alpha-n on stock motronic ECU.

what just by wiring the TPS output to the AFM connection? sounds like it might be easier than trying to run a MAP sensor..

Clearly the outputs are not the same, so it would need a custom map, and that seems to be about it, the output on mine is roughly 0.5V idle to 4.5V WOT ;

'ANT' has done this mod to good effect. . .

[DmcL]

yea i know i mean as for physically connecting it up. tuning is well covered.. i sell the odd chip and do custom tunes on these cars so just a matter of peicing the kit together then slap it on and get the wideband and laptop stuck on the car.

any idea of how i would go about also wiring in a MAP sensor left open to the atmosphere to use for correction of TPS output signal according to barometric pressure? i was googling it but information doesnt seem to be as easy to find as i had hoped. i know both sensors require power and ground and that leave 1 output signal wire on both sensors but what do i need to do/make in order to use the MAP sensor to correct the output of the TPS sensor? output of MAP sensor to power input of TPS sensor or what?

EDIT: actually thinking about it now that probably would work.. the TPS takes a 5v input and when i was playing with the GM and Denso MAP sensors i picked up i noticed the GM sensor outputted 4.55v open to atmosphere in the house when i was checking it actually worked with my multimeter.. anyone think otherwise?

[Gunni]

you´d need to create a circuit that will change the TPS voltage based on MAP voltage. This needs to be done electrically , maybe with a gain transistor or something. It´s no easy task and not really within the capabilities of the Motronic code either.

The is nothing wrong with a map sensor signal on a stock M20, it can be very steady, say 45-46kpa at idle as an example.

Basically with a MAP sensor or Alpha-N conversion you are just feeding a 0-5v signal into the AFM signal line to the ecu, from there you need to readjust scales and tables and table values to make the input work properly to give out the right pw and spark angles.

Its not simple as changing a MAF , there is a lot more to it in terms of adjusting within the ecu.

[GeoffBob]

It´s no easy task and not really within the capabilities of the Motronic code either.

But well within the capacilities of a good aftermarket ECU, which is what I think you should be looking towards DmcL. I know that these goodies come at a price, but they also usually come with instruction manuals and on-lone forums detailing how to set them up to work with MAF's, AFM's MAP's etc in all the various tuning modes. Megasquirt might be a good place for you to start since it comes in kit form and is great fun for those who want to experiment. Otherwise forget Alpha-N with your Motronic and stick with speed-density (MAP) until you get your tuning right. Hard as hell to tune without an 02 sensor though Bit like playing darts in the dark.

[e301988325i]

yea i know i mean as for physically connecting it up. tuning is well covered.. i sell the odd chip and do custom tunes on these cars so just a matter of peicing the kit together then slap it on and get the wideband and laptop stuck on the car.

any idea of how i would go about also wiring in a MAP sensor left open to the atmosphere to use for correction of TPS output signal according to barometric pressure? i was googling it but information doesnt seem to be as easy to find as i had hoped. i know both sensors require power and ground and that leave 1 output signal wire on both sensors but what do i need to do/make in order to use the MAP sensor to correct the output of the TPS sensor? output of MAP sensor to power input of TPS sensor or what?

EDIT: actually thinking about it now that probably would work.. the TPS takes a 5v input and when i was playing with the GM and Denso MAP sensors i picked up i noticed the GM sensor outputted 4.55v open to atmosphere in the house when i was checking it actually worked with my multimeter.. anyone think otherwise?

I've had two thoughts on this;

1) IIRC correctly E30 M3's had an altitude sensor built in, mounted on the bulkhead, they use roughly the same ECU.

2) Wouldn't the intake air temp sensor compensate quite well for this, ie high intake air temp = lower air pressure?