MyE28.com

DAVE,
On my, good to hear from you. Ill be sending you a PM on the chip you sent me, a long story.....

Anyway. I read your posts, and thanks for the info. I was skeptical of your finds, so I went and got the schematics for the M1.3 DME from the the Russian Diagcom.com.ru site. And I could not beleive my eys. The M1.3 DME actually appears(Im ~85% sure) to run the o2 sensor output, amplifies it and sends it to the ADC imput to the CPU!!!! Holly fucking shit. I would have not believed it if I hadn't seen it.

I guess BMW has such good software algarythems and accuracy, they actually use the NBO2 sensor voltage in the non linear areas of lean and rich. Wow. That is really hard to do, but they do. Its still unclear if they dictate fueling from it or just use it to do failure analysis to decide on go/no go limp modes.

So, your right, Im sure the input impedance of the amplifier circuit is very high. So the 6.2Mohm makes sense. So does the Dinan O2 clamp box, as you need the varying voltage to fool the computer.. Oh my, what a deal.

Thanks for the info, its been very helpful, even though I didn't believe it at first.

RussC

Russ,

Very interesting that you were able to get the schematic.
So, now please explain what the stuff you said about the O2 signal means......in more simple terms. Thanks.

And, trust me the DME trims fuel in WOT. That trim also affects the mixture for closed loop, at least until it trims back. The trick is to get it to trim the same amounts in WOT and closed loop and the 6.2 Mohm seems to do that.

The DME will adapt far enough to run just fine with the 42 pound injectors and no signal modification. I ran it that way for a time while awaiting delivery of a new SMT-7. Same as a 6 but with a faster chip and better software.

Beyond that if it gets angry, can't get the fuel within limits, it yanks timing like crazy and the car goes limp and overheats. Probably not for a N/A but for a turbo set up the exhaust gas temp goes through the roof.

Cheers,
Dave

Dave,
Sorry for the genuine electronics gibberish.

Ok, first lets look at the O2 signal characteristic curve:


When I talk about the non-linear areas, I mean the point where the the curve indicates rich and lean. So 0.7-0.9V and 0.15-0.02V. As you can see by those curves, very small changes in voltage indicate very large changes in fuel delivery(O2 content as compared to atomosphere). In this chart, there are 3 curves at the rich end of the ouput, those are there to show the difference in ouput of the O2 vs sensor temp. See the next propblem, the O2 content hasn't changed, but the voltage as depending on temp, its self-evident this is bad unless you control the O2 temp.

This is why you can't really use a NBO2 to measure the actual fuel mixture, only thats its rich(or lean). I can't believe that BMW is acutally making fueling desicions on the curve like that, so I believe its the really on limp mode issues since it thinks something is wrong.

On the electical stuff, I wont go into too much detail, as it would take very long to explain input/output impeadance calculations, impeadance matching, ADC theory, amplification and more. But, I will just say that the ADC(analog to digital conversion) is very suprising to me that BMW did this. The E28 059 ECU just uses a sensing circuit to see if the O2 is runing rich or lean. If its lean, the ciruit assert true and interrupts the CPU. The same on the rich side. If the voltage is in-between, theres no action. I just pressumed that the M1.3 did the same thing, as I though there was now way to accurately measure the O2 output, but I was wrong.

This is really great info for anyone using the 179 ECU, I believe I understand now why Andrews E34 535i dyno was so messed up runing 5-6psi.

Ya, Ill use the 42lb injs, that seems to be the right way to go.

Thanks,
RussC

Russ,

Thanks for the explanation. I should have been more clear, cause with all of the fussing I have enjoyed in making my car run correctly, I had learned that stuff. What I did not know was what a ADC was. Now I do thanks.

And this stuff is exactly why Andrew eventually went to the TECII or III.

Regardless I am running a fun fine smooth 15 psi now on the 42's with the stock fuel pressure regulator. In this weather I can get away with that on 91 but it is clearly happier on 94.

Before I added the resistor, on long pulls up the back straight at Mosport which is a long uphill entered from corner 5 mid way through third and is WOT into about 220 kph whatever that is in MPH. I would be fine nice and rich (I am reading with a wide band) and the half way through 4th I would start to go lean as the DME pulled fuel.

Now rich all the way..... lovely,
Cheers,
Dave

Dave,
That is great, it appears Ill have to run some sort of O2 circuit like you. But I won't need it that long, Ill be going MS in the summer.

Are you still runing the AFM or a MAF?

RussC

I did the AFM to MAF switch with the SMT modifying the signal before I went turbo.

You will find the simple relay and resistor circuit pretty easy and it will work well.

Have fun,
Dave

OldgreenE34 wrote:I did the AFM to MAF switch with the SMT modifying the signal before I went turbo.

You will find the simple relay and resistor circuit pretty easy and it will work well.

Have fun,
Dave

Will do.

Another question, did you run the AFM with 42lb injs? If so, the 179 was able to adapt to that w/o AFM tension adjustments? Or what needed to be done to make 42lb's work with the AFM, if any?

Thanks,
RussC

Hi Russ,

No, never did use the 42’s with the AFM. I used the MAF/SMT with the stock injectors, N/A and then with 30’s and 42’s boosted. Regardless idle and cruise is always the same.

The numbers in the tuning map to adjust the analogue output from the MAF up or down by 39mV per unit. However, when I ran for a time with a MAF I did so withOUT the SMT. So, just feeding the MAF signal directly into the DME and the car, after adapting, ran just fine in cruise and at idle. What I have seen in my tuning, before I figured out what was going in terms of adapting, are map numbers at idle ranging from –8 up to 4. The DME has made the thing run over that huge range.

The problems appear in boost. You have to ensure the DME has adapted very little and that any adapting has been done in the richen-it-up direction. I am now tuned with a –6 at idle and –3 or –4 at cruise dependant upon RPM. As an aside you will also find that the DME acts as if it is in WOT at engine speeds above say 5,000 RPM regardless of load. You will want that resistor modified O2 signal up there as well as at WOT.

As a point of reference my tune map at WOT is about –23 with cruise at –3. This puts me at about 4.66v at 15PSI from 6000 to the rev limiter that Mark set for me at 6500. If I go above that voltage it goes bad, even if I don’t go high enough to hit the limiter, fuel cut off or whatever it is.

Causing the DME to adapt rich with a low number in the SMT tune map ensures you get enough fuel when you need it. Because you cannot feed the DME any more than 4.7 or a bit lower voltage.

Let’s say cruise tune is at a 5 (adding 5x 39mV to the MAF signal) and the DME has pulled fuel to get 14.7 at cruise. Then you go to 4.7V at WOT and you get less than if the DME has adapted to a tune of –3 (reducing the signal to the DME by 3 x 39mV) as I now have. Then the next trick is keep the DME from re-adapting while at WOT! Hence, the resistor.

If you are going to keep the AFM in the interim I would expect you will need to increase the tension of that spring quite a bit. If you can get it so that it does not peg itself against the stop until 6,000 + RPM then you will be able to run with the stock fuel pressure. For idle and cruise you could run with a MAF or stock AFM. But that isn’t what you plan to do right I suggest you monitor the AFM voltage while tuning. With the SMT I get a tracing of the voltage on my laptop.

Or, I guess if you stay at lower boost levels you will be OK. If you are like me, and I assume you have added a thicker head gasket you will not be able to get yourself from turning it up! I am running a stock bottom end with 336,000 km, ARP studs and a stock head gasket with a milled head about 0.030” so slightly higher than stock compression at 15 PSI.

I managed to stay in front of a new M5 on a local highway for quick toot a month or so back. Now that car was just out of the showroom and I understand they are stuck in the 400 hp mode until broken in. So, I didn’t have a run with a 500 hp car. And, more recently on the way to the ski hills a 645 took me on and I flat dusted him, slowed to a more socially acceptable speed – he came along side and gave me a big thumbs up.

Cheers,
Dave

Dave,
Holy Shit, If I understand what your saying, and I want to be sure of this.

Your saying you connected the MAF directly to the ECU AFM input and it worked with not other electical interface device???? It would start, idle and cruise in that state??? If thats the case, this is huge. I would have never thought that. That means with a Apex-i SAFC-II you can use a MAF on the 179 with no other mods! Does that sound right?

Cause that would be the cheap bomb MAF upgrade for cheap!!

What MAF are you using? Can any Ford or GM unit work? Any input would be great.

On the O2 stuff, aren't you using a WBO2? Cause a LM-1 would be able to replace your relay/resistor setup, I believe. The LM-1 WBO2 has a narrow band output that can configured for any output levels in the NBO2 range. It's my contension that you could configure it to olny output to to a lower voltage, thus keeping the fuel trim correct in the WOT range. I have an LM-1, so Ill try this out when the stuff gets hooked up.

Thanks for all the info on the triming of MAF levels. I'll have my Apex-i SAFC-II hooked up, so I can do alot of the stuff your doing with the voltage trim's.

Man, this is really exicting, this is big news if what your saying works, nice work and thanks again.

RussC

Russ,

Correct, the MAF signal directly connected the DME, no signal interface at all. That is with the 42’s. It will be dam difficult to start at first, you will be rich at cruise and idle. Trim will take maybe 15 to 30 minutes driving. I am sure you know there are 2 trims, short term and long term. Short term will have you able to run in say a couple of minutes and then it will take the longer period of time to get smooth with the longer term trim.

The but is that after you are trimmed with no interface signal adjustment you will be lean in boost. That is what all of the chatter was about getting the trim ‘right’ so that you had enough fuel at the max input voltage of 4.7V, because that is an absolute limit.

Now I don’t know what an Apex-i SAFC-II will do or can do, but if you can modify the MAF signal based on engine speed and load on some sort of map you are there. You will need a chip from Todd or D’Sylva. I am running Mark D’Sylva chip which is 100% stock, except for timing and a 6500 rev limiter. I am adjusting the timing further from Mark’s chip and will be willing to give you my timing map to allow Mark to burn you a chip with my exact timing map. It is smooth and quiet, likely a bit conservative – which keeps engines together!

I have a Mustang MAF, big old plastic thing 3 ½” in and 3” out, I think. Regardless it’s huge inside relative to that little square hole in the AFM.

I am using a Bosch NB through the resistor to the DME and a Bosch WB with a Daytona TwinTec data logger and NB output simulator. I feed the NB simulated output to an LED O2 gauge, the sort intended for use with a NB and to feed the laptop for SMT tuning. Again I don’t know what an LM-1 will do. If you can get the LM-1 to send the signal you want the DME will trim as you desire. BUT, you will still need to remember that whatever trim you get at WOT will affect cruise and the same going the other way. This is what took me so freaking long to get right.

What I’m saying works, it just took me a long time to figure it out ….. so I am more than happy to help anyone ready to attempt the same thing.

Have fun,
Dave

Im off to look at my "theoretical" wiring harness. Im holding my breath its right. Report soon.

Dave, Ill reply when I get back.

Russc

OldgreenE34 wrote:Russ,

Correct, the MAF signal directly connected the DME, no signal interface at all. That is with the 42’s. It will be dam difficult to start at first, you will be rich at cruise and idle. Trim will take maybe 15 to 30 minutes driving. I am sure you know there are 2 trims, short term and long term. Short term will have you able to run in say a couple of minutes and then it will take the longer period of time to get smooth with the longer term trim.

The but is that after you are trimmed with no interface signal adjustment you will be lean in boost. That is what all of the chatter was about getting the trim ‘right’ so that you had enough fuel at the max input voltage of 4.7V, because that is an absolute limit.

Now I don’t know what an Apex-i SAFC-II will do or can do, but if you can modify the MAF signal based on engine speed and load on some sort of map you are there. You will need a chip from Todd or D’Sylva. I am running Mark D’Sylva chip which is 100% stock, except for timing and a 6500 rev limiter. I am adjusting the timing further from Mark’s chip and will be willing to give you my timing map to allow Mark to burn you a chip with my exact timing map. It is smooth and quiet, likely a bit conservative – which keeps engines together!

I have a Mustang MAF, big old plastic thing 3 ½” in and 3” out, I think. Regardless it’s huge inside relative to that little square hole in the AFM.

I am using a Bosch NB through the resistor to the DME and a Bosch WB with a Daytona TwinTec data logger and NB output simulator. I feed the NB simulated output to an LED O2 gauge, the sort intended for use with a NB and to feed the laptop for SMT tuning. Again I don’t know what an LM-1 will do. If you can get the LM-1 to send the signal you want the DME will trim as you desire. BUT, you will still need to remember that whatever trim you get at WOT will affect cruise and the same going the other way. This is what took me so freaking long to get right.

What I’m saying works, it just took me a long time to figure it out ….. so I am more than happy to help anyone ready to attempt the same thing.

Have fun,
Dave

Dave,
You are a true pioneer. I would have never even considered hooking a MAF directly to the ECU You are a real adventurous type.

Ya, I believe your lean at WOT due to the way MAFs measure air mass as compared to a AFM. Im looking at the MAF curves and I can see why that is. Its probably also due to the fact that MAF dont measure to very high voltages on the high side of air flow. So you may be down a little there.

Yes, the S-AFC-II is a ECU similar to the SMT stuff, but not nearly as complicated or have all the features either. It is map based, and can modify an incomming signal +/- 50% vs RPM to the ouput. It alos has a TPS input to swicth maps depending on TP. So you can have a low throttle map and high throttle map. It can also monitor knock input, inj duty cycle and RPM. It can data log, but only to screen, not to file. I believe it will do enough that I can get the MAF to work plenty fine.

Can you post a pic of your MAF. Ive looked at some of the Pro-M, GM, Granatelli and some others on Ebay. If I see what yours looks like, that would help alot. In doing some first order calculations, the MAF needs to be at least 3", and 3.5" would be better. I noticed the Bosch HFM5s go up to 82mm but only measure max 1000kg/hr @ 4.33V. Thats only ~360hp worth. So a little bigger is good.

Thanks for the O2 routing. Ill be using the LM-1 to route the NB output to the DME with like a 0.75V limit. I can adjust it if needed to. And still use the WB for gauge and datalogging. If the LM-1 active NB clamp doesn't work, Ill go to your resistor method.

Have you dyno'd the car at 15psi? That would great to see if you did.

Thanks Again,
RussC

Russ,

Not a pioneer, I had a failure of sorts and had to connect the MAF directly to the DME and when it ran for a week or so until I could affect a repair I was pleased. Necessity and all that.

OK it sounds to me that you can use the S-AFC-II to do a MAF conversion. Cool!

I managed to break our digital camera skiing with our 13 year old son. It looks like a Mustang MAF from about 1995 because it is. I think that part is pretty easy, go to the Ford shop, their parts are cheap!

Oh, ya do not wire the air temp sensor. I found that the DME is too sensitive to air temp changes with the MAF. And, that makes sense as with the MAF you are essentially alreay looking at air mass with the AFM there is a need for pretty big correction. I just bridged the input with a resistor. I do not recall what resistance, but when I plug in an OBDI, I get an air temp reading of 13 C which is about 50 F, I think. Regardless that never changes from the DME's perspective.

I like your LM-1 plan, and hope it works. Let me know.

Sorry, I have not been on the dyno since I got the thing right. However, before I did and when I was going into limp mode above mid 4000's the car made about 320 at the wheels below 4000 RPM. Given the way the car feels now I am in the high 300's at the wheels. Todd suggests 375 and I think he is cautious but who knows?

And remember my engine is pooched, 336,000 km. A fresh engine is going to make MORE.

Did I tell you my forged bottom end engine is in the machine shop and will be in the car my mid May? First track event this year is late April, second track event early June on the new engine! Mosport is the the home track.

Have a great weekend,
Dave

Hi Russ,

My MAF looks like this one but not polished,

http://cgi.ebay.com/ebaymotors/Ford-MAF ... dZViewItem

or you get this, which I think is the same but in aluminium and add the sensor from the cheap plastic unit above,

http://cgi.ebay.com/ebaymotors/C-L-80mm ... dZViewItem

I have wondered about getting the aluminium unit and converting to blow through. Maybe some tuning issues but throttle response would be better, maybe...

Dave

Dave, could you email me a "fast and dirty" schematic of your modified O2 circuit? I'm 99.63771% sure of you've done, but would like to bump that up to 99.63928%
Thanks!
Sweeney

Boru635 at msn dot com

it can look like this:


Dave will look similar. I added the trasistor circuit depending on what kind of drver the pressure switch is. You may need it to get enough drive output to throw the relay coil. The diode prevent reverse current flow for protection.

If using no trans, the signal will be a +V to GND. If using the trans, the signal will be GND to +V. V being 12V. the relay is a 12V type.

RussC

Hi Sweeney & Russ,

Sweeney, I have sent a sketch in .xl, I hope it makes sense.

Russ, bridge the N/C contacts with the resistor, nothing is connected the the contact which is closed when the relay is energized. Simply, when the N/C contacts open the circuit passed through the resistor and when the contacts are closed the O2 signal takes the lowest resistance circuit....

Simple and simple is good.

Cheers,
Dave

Ya,
You can do it that way also. I just sketched that up since you mentioned a relay.

With what you described, you could use just a pressure switch and not use a relay at all, in absense of some other switching device for the transfer point of vacum to boost.

RussC

Hi Russ,

Absolutely a pressure switch is all that is needed. I used a relay as I have MAP driven output from the SMT and because it allowed me to, with one activation, close the WOT circuit and introduce the resistor.

Yesterday coming back from the ski hill cruising at a socially acceptable speed, the car just didn’t feel right. A look at my O2 gauge to see I am pretty big time lean. My guess is that the relay hadn’t proper reset or closed the N/C contacts after my last brief blip into boost. So the resistor was in the O2 signal at cruise and it had taken a good few seconds for the thing to adapt itself to a lean condition. (maybe a good way to improve mileage?)

Regardless, I gave it a punch to cycle the relay it bounced rich until it ‘re-adapted’

I think I will add a new relay later this week.

Teach me what the diode does across the relay inputs. Is it possible the relay is sticking because I don’t have a diode?

Cheers,
Dave

Dave, have you considered using a solid state relay?

Sweeney wrote:Dave, have you considered using a solid state relay?

I'm using a Vishay LH1502 SSR for, umm... a "thing." One of the SSRs within the 1502 family will probably work for the type of switching you need, just remember that its max switching current is 150mA and that since it's an optoelectronic (i.e. LED inside), you'll need to have an external current-limiting resistor on the input.

OldgreenE34 wrote:Hi Russ,

Absolutely a pressure switch is all that is needed. I used a relay as I have MAP driven output from the SMT and because it allowed me to, with one activation, close the WOT circuit and introduce the resistor.

Yesterday coming back from the ski hill cruising at a socially acceptable speed, the car just didn’t feel right. A look at my O2 gauge to see I am pretty big time lean. My guess is that the relay hadn’t proper reset or closed the N/C contacts after my last brief blip into boost. So the resistor was in the O2 signal at cruise and it had taken a good few seconds for the thing to adapt itself to a lean condition. (maybe a good way to improve mileage?)

Regardless, I gave it a punch to cycle the relay it bounced rich until it ‘re-adapted’

I think I will add a new relay later this week.

Teach me what the diode does across the relay inputs. Is it possible the relay is sticking because I don’t have a diode?

Cheers,
Dave

Its hard to believe that the relay contacts are failing that quickly, especially with the amout of current your switching. Unless the relay coil has failed, They can from time to time.

The diode it used to disipate reverse current stored in the coil when the it is switched off. If there is no path except the output of the driving circuit, that stored current can damage those output drivers. Essentially its a voltage clamp(+V + 0.7V[or the diode drop]). You dont need them for solid state diodes.

And, yes Sweeney and Shawn, that SSR will work fine. Shawn is correct making sure the opto input nees proper currect limit resistors. The 150mA load will work easily for the O2 sensor as its very low current. I dont beleive the 20ohm resistance through the mosfet switch is an issue with the O2 circuit. That 20ohms is usally a killer in the fields Ive worked in for SSRs.

RussC

John,

Nope, didn't know there was a solid state relay, I just went to my local place where I hunt for resisitors and picked up a 12 v relay.

How do I spec. a more reliable relay?

Thanks,
Dave

Shawn,

Your suggestion is propably great, but I need simple and rugged. I have a 3 day track event in April and then I have to drive to work on the Monday.

Thanks,
Dave

OldgreenE34 wrote:Shawn,

Your suggestion is propably great, but I need simple and rugged. I have a 3 day track event in April and then I have to drive to work on the Monday.

Thanks,
Dave

Dave,

I can see you wanting "simple" in order to get it done by April, but optoelectronics are quite rugged and will have a lifespan greater than your car if you use them properly (i.e. use the proper limiting resistors and don't exceed the rated switching current).

-Shawn

OldgreenE34 wrote:John,

Nope, didn't know there was a solid state relay, I just went to my local place where I hunt for resisitors and picked up a 12 v relay.

How do I spec. a more reliable relay?

Thanks,
Dave

That may be the problem righ there, is the place your going to a surplus store(I love the one here by the way). If so it was probably a used unit. So longevity may be suspect from the get go. If it new, then you just found a bad one. Try again with new unit. At those currents, the relay should last many years.

RussC

Right Russ,

I am going to grab 2, one to install and one for the spares inventory.

Cheers,
Dave

Well,
My E24 harness showed up at Edge Motorworks. I going to look at it tomorrow. If it checks out to be the correct harness, swap happens next week...

Update and pics tomorrow.

RussC

Finally,
I have found a harness I believe I can use with little modification. First, see the E34 M30b35 harness


Here the M20 E30 harness


The E34 harness is much neater. But it appears the M20 harness will work. All the connectors are there. My harness is from rs4pro3. This harness he was using from a E30 that had a M30b35 swap'd in. I just got this harness from him. The M20 harness is in two pcs, the main portion and the inj'r harness. He installed the inj harness in a M30b35 housing so it would look factory. Nice touch. More pics




This M20 harness was in E30 cars from ~'89 to the end of the E30 run, '92. These were all M1.1 and M1.3(same as E34). Then the E36 hit and the system changed from inductive pickup on the #6 cylinder plug wire to a real cam sensor.

Ill be cleaning this puppy up tomorrow. Its my belief(finally) the car will be at Edge Monday to begin install. Its taken 7 weeks to get here..

RussC

Hi Guys,

Since I am in the process of doing an e34 M30B35 into my 87 e30 325is, I was directed to this thread by Dave, just came off the phone with him.

I see Russ making mention about Dave's brilliance and creativity. There is not a whole lot more one can do to reinforce this, Dave is well respected in the Toronto area for who he is and what he is.

Reading all the technical information exchanged between Dave and Russ (very smart and creative as well) I am impressed about the findings and solutions these guys have come up with. It makes it all the more interesting and exciting since my background involves electronics.

Russ there have been a few people who have done the m30b35 swap into their e30's using the existing m20b27 engine harness. I have researched it and though mine a bit different being an 87, will work as well. I was a bit mad at one of my friends who worked on stripping a 90 e30 with me becaused he damaged the harness in the process.

If it helps, here is a bit of write up on another guy who did the m30 swap into an e30
I know you are doing the swap into an e28, but there might be links from his site which may be useful as well.

http://www.jaredsquires.com/ndp/335i/m30/page13.html

Thanks to all of you guys for sharing information.

BTW, I will be turbocharging my m30 when the swap is done so may come back for more info

Cheers,

Raph.