Sequencer Beta into BMW M90 ( M30 )

[TheMonkey]

i'm currently beta testing the Sequencer on my '61 Pontiac. it's running, and sequencing. problem is, that i'm experiencing some whacky noise issues that i'm normally very good at tracking down, but this is proving to be tough to manage everything at once. so... i'd like to introduce the sequencer in stages, which i cannot do on my Pontiac because i don't have an existing unit on it (well, i could use the MS2, but it shares too much with seq). also, waiting on some suspension work in Pontiac to get driving results.

sooo.... i happened across a 1980 635csi. this is Euro spec with M90, which is a lot like M30, but bit different head and CR. i was just going to get in and go, but it quickly blew a head gasket which i suspect was from clogged/lean injectors from sitting just too long. i start putting in a wideband, ordered a B35 crank sensor which should bolt on 60-2 in OEM fashion, 6 pin TPS.... i started my conversion to MS, and didn't even realize i was doing it.... like a zombie.... can't i leave things alone?

anyhow, i told Bruce and Al that this might get some good info and traction on the sequencer beta testing which they agreed. my plan is to introduce the beta sequencer on this car in stages, and hopefully be running fuel very very soon.

- fuel wasted only in bare bones fashion with separate pigtails for injectors
- ignition in wasted with MS still overlayed over existing wiring (sequencer requires COP; i'll use more D585)
- introduce cam sync
- introduce IAC, fan control, etc.
- remove existing wiring & l-jet, clean up new wiring loom

i should have motor back together next week after parts arrive, with 60-2 wheel, and TPS.

the initial questions i may have will likely be BMW specific, which is why i'm posting the intro here, and sequencer specific Qs will be found in the uS beta section.

- initially, if i overlay just fuel control over the existing l-jet, can i use/share existing coolant temp sensor with the Ljet ECU, and which one of the 3 sensors is that?

- is there a particular ground spot, or has any of the block grounds worked okay for you guys?

- any noise issues with the M30 to look out for? use a filter to MS power? stock ignition wires seem to have pretty low resistance.

- i am going to use stock low-z injectors which measure at 2.4 ohms, which Bruce is advising that the drivers are built to handle it, but I'll wire the driver grounds very sufficient. at high duty cycle, there is potentially a lot of current.

- any other gotchas for BMW stuff? when i get to a point of stripping the original wiring and ECU, i thought I read something about the fuel pump being activated by +, rather than ground signal (but I can't find that info now?)?

thx for any help on getting up to speed.

Scott.

[FixItAgainTony]

Hi Scott,
I don't know much about the BMW specific, but I did convert over from L-Jet, and they are all pretty much the same.
I would be careful on sharing the sensors. The MS units operates on a 5V voltage divider. I am not sure about the L-Jet. Disconnecting the CLT sensor and measuring the voltage on the CLT wire from the ECU should give you the Bias voltage. As long as it is 5V or less, it should be fine. Anything much above 5V on the MS-2 controller pin will hurt it.

The L-Jet harness is usually separate from the rest of the car and grounds to the block (okay, intake manifold area).
Peter Florance did a nice job documenting this, you can find a copy of it here:
http://www.alfagtv6.com/MegaSquirt/ljet-to-ms.pdf

The spreadsheet I used for connector matching is here, but I am not sure that much of it applies for what you are doing:
http://www.msruns.com/viewtopic.php?f=98&t=15983 (I should update this, now running EDIS & MAF).
I built my MS to fit in the L-Jet ECU box. If I had to do it again, I would get an old L-Jet connector (desolder from a the board) and build a pigtail to adapt it to the DB-37 connector.

The harness was not so great on my car - the Fly-back 12V to a separate location modification is needed:
Item # 16 in the following link: http://www.megamanual.com/ms2/V3trouble.htm
If I did not need to retain L-Jet capability, I'd toss the harness and start from scratch.

You are correct about the high drive required to turn on the fuel pump if you use the existing relay set up. It can be implemented with an extra resistor and transistor. Let me know if you need to know how to set that up.

Hope this helps,
- Charles.

[Peter Florance]

Hi Scott
I guess you found the PDF I was going to send you.

You shouldn't have to share any sensors. Buy an old ecu and make a conversion plug using my pinout. There's one error where I used a pin 3-6 as a ground. That was for V2.2 board and won't work on V3 board.

I don't convert the fuel pump drive anymore because I have to test so many MS boxes, I needed to make mine standard. So I mounted a TIP125 on the adapter so any standard MS box set up for VR and single coil will run in my car. Collector to combo relay. emittor to +12V and base, through 1k to pin 37 of ms box. Be sure to insulate the tab unless you are lucky enough to find fully plastic TIP125's.

The BMW harness is excellent and their grounding is very good.

Let me know what else you need and I'm sure Charles will send it to you.

[cmonref]

Scott wrote: "I thought I read something about the fuel pump being activated by +, rather than ground signal "

and Charles responded: "You are correct about the high drive required to turn on the fuel pump if you use the existing relay set up. It can be implemented with an extra resistor and transistor. Let me know if you need to know how to set that up."

I am not familiar with the BMW fuel pump setup, but I am wondering if it is the same as the Porsche 911 fuel pump setup in 1980. During start, the fuel pump is powered by the starter circuit. After start, it is powered by switched power thru the NC position of the relay. There is a single switch in the engine compartment that is normally open, but if closed, provides a ground to the relay that will open the NC position and shut off the fuel pump. So, ground that wire, shut down fuel pump. Can't directly use the MS signal for THAT ...

Charles: Does the "extra resistor and transistor" work with the MS fuel pump signal to reverse the sense of that signal? The desire would be to use the transistor's output to act as an open switch to the OEM wire when the MS says "fuel pump ON" via its ground, then show a closed switch to the OEM wire when MS says "fuel pump OFF" via an open circuit from the MS.

There is probably a more elegant way to describe this, but I hope you get the point.

Brian

[FixItAgainTony]

cmonref,
I think the '80 Porsche used L-Jet, what you describe is close to the L-Jet operation. You are correct on the transistor modification, it is just to change the polarity of the signal to provide enough current to energize the relay. On LJet, the relay coil always has a ground, it is power to that is switched to operate the relay from one of two sources in a wired 'OR' [diode] set up. Either starter motor running OR AFM flap open energies the relay. This is to insure that the fuel pump is turned off if air stops going into the engine - accident, etc. Look at the explanation of the left side of the wiring diagram that Peter wrote up. MS only supplies ground paths. To provide the voltage / current required to close the L-Jet relay a "high side driver" is required. Something like what is described here does the trick:
http://www.eckhard-gosch.de/download/digio.pdf see figure 6a, part b to the right.

Peter, thanks again for taking the time to document all of the L-Jet stuff. A lot of people have found that useful.

Scott - the "conn" tab has wiring differences between what Peter used (V2.2) and what I used (V3) are there along with some additional comments at the bottom on the different L-Jet signals as obtained by looking at the circuit board & harness from the L-Jet used on a 4 cylinder Fiat.

- Charles.

[cmonref]

Charles,

The Porsche 911 used K-Jet (CIS) thru 1983 then switched to Motronic. The Porsche fuel pump relay works "backwards" from the L-jet pictured in Peter's wiring diagram -- the relay is "at rest" and HAS power running thru the NC terminal to the fuel pump (which has a full-time ground). If the engine stops (ignition still ON), the airflow switch closes, providing a ground to the relay. The relay activates, opening the NC terminals and stopping the pump. Different from the L-jet, so the "high side driver" is not the solution to my problem.

I can "fix" the problem with a relay, but a little circuit would be so elegant -- and with fewer moving parts"!

Brian

[FixItAgainTony]

Brian - sorry about the mix up on the injection system. A friend had a similar vintage Porsche (different model which I think used the L-jet). A tree ran in front of it. The tree kept running, the car didn't.

Any chance the normally open contact can be used for the fuel pump? If available, then one side of the relay coil could be connected to 12v and the standard MS could ground the other end of the coil - standard MS operation and relatively simple to implement. I have had luck in the past using a small flat bladed screw driver to remove the wire & connector in a relay receptacle - just push down on the retaining tab and pull.

- Charles.

[TheMonkey]

Thanks you guys for the information. I've read over it a couple times, very helpful.

I suppose based on this, even for the first phase of fuel only, rather than piggybacking wiring to injectors I should do a full swap of the existing L-jet and wire the sequencer directly to harness. (Charles - I want to retain L-Jet capability also). I'm trying to find a donor L-jet case with recepticle harness now. For some reason I was thinking that ignition was intigrated into it, but ignition is entirely within the small unit located near the coil?

The only thing about full removal of existing L-jet for 'fuel only' phase, is that it will require a couple extra steps - like addressing the fuel pump relay. But... it eliminates any messing with some other wiring which I'm really warming up to. I'll have to do the fuel mod outside the MS unit since sequencer may be swapped back into my other project.

Charles- I don't think your flyback mod applies to sequencer since it does not have PWM control. Bruce says that low-z injectors will get triggered like high-z and that it will handle the current with sufficient grounds.

If anyone has a donor L-jet, let me know. Thx.

[Peter Florance]

This is an updated dwg that includes the transistor and TPS gnd pin update

[TheMonkey]

a potential interesting solution to the fuel pump relay signal:

donor L-jet arrived today from Peter. As he had initially indicated, the pins on this unit have a mildly different layout. Interestingly, the only pins that are not in the unit that otherwise would be used are: 20 (fuel pump on), 28 (combo relay ground), & 29 (switched power to ECU from relay).

since i have several other wires that will bypass the harness anyway, this may be an opportunity to retain the traditional grounding signal from MS. why can't i provide switched power to MS directly from the ignition switch under dash, jump 86b (fuel pump on + signal) also to switched ignition power, and connect the 'combo relay ground' 85 at relay (pin 28 on l-jet) with the fuel pump relay signal from MS? essentially the idea is to connect positive relay to an ignition switched source, and ground the relay using normal MS fuel pump signal. this wasn't possible before because the power to MS came from relay source which would have been self dependent, but if it comes from switch, then shazam?

[Peter Florance]

a potential interesting solution to the fuel pump relay signal:

donor L-jet arrived today from Peter. As he had initially indicated, the pins on this unit have a mildly different layout. Interestingly, the only pins that are not in the unit that otherwise would be used are: 20 (fuel pump on), 28 (combo relay ground), & 29 (switched power to ECU from relay).

since i have several other wires that will bypass the harness anyway, this may be an opportunity to retain the traditional grounding signal from MS. why can't i provide switched power to MS directly from the ignition switch under dash, jump 86b (fuel pump on + signal) also to switched ignition power, and connect the 'combo relay ground' 85 at relay (pin 28 on l-jet) with the fuel pump relay signal from MS? essentially the idea is to connect positive relay to an ignition switched source, and ground the relay using normal MS fuel pump signal. this wasn't possible before because the power to MS came from relay source which would have been self dependent, but if it comes from switch, then shazam?

You could do that as you have no interest in reverting back to Ljet

When I first installed the V1.01 board in mine, I wasn't confident that my bag of parts would work and I carried an AFM and ecu in my trunk for months. I never needed them.

[TheMonkey]

...
You could do that as you have no interest in reverting back to Ljet
...

i think if i keep my changes documented, this swap back would be easy.

i made a first draft of my wiring with sequencer. this is using the 106 box without pins 20, 28, & 29, so those items get wired outside of harness, with changes implemented as described. the fuel pump combo relay changes are not shown in my sequencer diagram. Note that Red side grounds need to be very heavy since injectors are low-z without PWM.

is there an IAT sensor existing inside the stock AFM box? it looks that way on Peter's diagram. if there is, how did Peter retain the function if the AFM box was in the trunk?

link if pic does not fit on screen: http://img213.imageshack.us/img213/8814/ljetdiag10.jpg

[TheMonkey]

L-Jet surgery is finished. posted a couple pics, and I moved the updates thread to the beta section since i'll start needing some SEQ specific help: http://www.microsquirt.com/viewtopic.ph ... 39#p148239

[TheMonkey]

Wiring question:

i don't see any fuses going to the injectors?

big red wire going to terminal 88z at combo relay comes directly from starter, and through relay to terminals 88b & 88e which seem to go directly to injectors. why wouldn't factory have injectors fused?

looks like ign power going to terminal 86c does not have a fuse either.

i'm inclined to fuse terminals 88b & e according to those bank power needs.

[Peter Florance]

Wiring question:

i don't see any fuses going to the injectors?

big red wire going to terminal 88z at combo relay comes directly from starter, and through relay to terminals 88b & 88e which seem to go directly to injectors. why wouldn't factory have injectors fused?

looks like ign power going to terminal 86c does not have a fuse either.

i'm inclined to fuse terminals 88b & e according to those bank power needs.

BMW just didn't fuse those lines. Not sure why, but they never did. I have the ETM for that model if you want some scans

I may be doing this on my car as well. Perhaps that will be helpful to you?

On later models the factory used inductive pickup on one of the spark plug leads (#6 IIRC)for some sort of sensing. I wonder something like that would be useful as a cam sensor?

[TheMonkey]

Peter-

I've got the manual for the car, thanks for scan offer though.

I'll just put 15 amp fuses coming out of combo relay to each bank.

When it comes time for cam sync, I was just going to first modify the reluctor wheel inside the distributor. If all works well, I'll prob swap the timing cover for later style and make a sync of the end of the cam. Or maybe modify the gear wheel at end of cam and use the distrib hole to mount a sensor. Not sure yet. First thing will be to get fuel running wasted.

Thx, Scott.

[TheMonkey]

another query about factory setup: why do you suppose the ground for the 'aux air valve' goes from the valve, through the harness, through firewall, into pin 34, through board trace to other grounds 16,17,35, back through harness and firewall, and then ground at intake? why not just ground from the air slide valve to the intake?

anyway.... my fuel relay wiring worked out really well.... i wired the jump from 86b to ign power downstream of the sequencer power fuse so that all i have to do is pull that fuse and it's ready for factory unit again. also able to retain grounding signal to fuel pump.

maybe i'll post a diagram of how this worked out, but the key to this was that 20,28,29 all bypassed the factory harness.

[Peter Florance]

another query about factory setup: why do you suppose the ground for the 'aux air valve' goes from the valve, through the harness, through firewall, into pin 34, through board trace to other grounds 16,17,35, back through harness and firewall, and then ground at intake? why not just ground from the air slide valve to the intake?.

I wonder if they thought they might make it an idle control valve.
I was able to plug a VW idle control valve in to the aux air valve connector and wire the PWM transistor right to that pin. Like BMW meant for me to able to do so.

[TheMonkey]

...
I wonder if they thought they might make it an idle control valve.
...

you know what..... after all my IAC motor woes on my Pontiac.... i have grown VERY fond of the aux air slide valve on this BMW. it works perfect. i love it.

[FixItAgainTony]

Scott,
Is there any concern with the low impedance injectors overheating without the PWM? I was under the impression that coil heat was the main reason for PWM.
- Charles.