D-Jetronic ECU Analysis Web Page

[pbanders]

I just completed a first pass at a web page on the analysis of the
D-Jetronic ECU:

http://members.rennlist.com/pbanders/ecu.htm

The page covers the following topics:

Introduction
ECU Operation Overview
ECU Photographs
ECU Schematics
ECU Block Diagram
ECU Circuit Block Analysis
ECU Waveforms
Common Questions and Answers

It still lacks descriptions of the operation of the engine speed sensor,
speed correction, and pulse width multipler circuits. I hope to add
descriptions of these circuits in the near future.

Any corrections, omissions, or clarifications are welcomed.

Brad Anders

[rjm]

Great webpage!
I'm in the middle of trouble shooting my ECU. I think I've got a speed compensation problem. I've check my injection pulse time with respect to different manifold pressures while keeping the RPM constant and the pulse width appears to be linear. But, when I check the injection pulse time with a constant manifold pressure and varying the RPM, it's not so linear. I'm looking forward to your update regarding the speed compensation circuitry.

P.S. If you have any preliminary information on the speed compensation circuitry, I'd appreciate seeing it.


------------------

[pbanders]

Originally posted by rjm:
Great webpage!
I'm in the middle of trouble shooting my ECU. I think I've got a speed compensation problem. I've check my injection pulse time with respect to different manifold pressures while keeping the RPM constant and the pulse width appears to be linear. But, when I check the injection pulse time with a constant manifold pressure and varying the RPM, it's not so linear. I'm looking forward to your update regarding the speed compensation circuitry.

P.S. If you have any preliminary information on the speed compensation circuitry, I'd appreciate seeing it.

I suspect that your SC circuit is working properly. It is a non-linear effect, but second-order. I would expect fairly small changes in the injection pulse width to compensate for filling efficiency as a function of speed. I'll update the document when I have the circuit analyzed.

Thanks for the positive comments!

Brad Anders

[bowlsby]

Nice work again Paul. You are assembling a great wealth of data on the D-Jet system

Curious about how this information might be used:

1. Evaluate an ECU for functionality - publish a docuemnt with input/output data.
2. Make a test bench for testing ECU's (need a special harness for a test bench?)
3. Add an idle mixture control knob to early ECU's without the knob?
4. Add part load/full load enrichment adjustability for turbos/bigger displacement upgrades?

What might you have in mind?

Jeff

[pbanders]

Originally posted by bowlsby:
Nice work again Paul. You are assembling a great wealth of data on the D-Jet system

Curious about how this information might be used:

1. Evaluate an ECU for functionality - publish a docuemnt with input/output data.
2. Make a test bench for testing ECU's (need a special harness for a test bench?)
3. Add an idle mixture control knob to early ECU's without the knob?
4. Add part load/full load enrichment adjustability for turbos/bigger displacement upgrades?

What might you have in mind?

Jeff

Thanks, Jeff. I mainly did it initially because I wanted to know what is going on in that skinny lunchbox. But obviously, there are things that you can do once you understand the operation of the ECU. To answer your questions:


1. Evaluate an ECU for functionality - publish a docuemnt with input/output data.

Actually, that's what Frank Kerfoot did with the ECU schematics. I'd like to do it with some additional data and waveform displays. I've got most of the equipment, but it might be a while before I get going on it.

2. Make a test bench for testing ECU's (need a special harness for a test bench?)

I'm working on this, albeit slowly. Jim Thoursen has already built a beast like this. I would have been a bit closer, except some ding-dong on Ebay decided to jack the price up to $40+ on the used injectors I was going to snag!

3. Add an idle mixture control knob to early ECU's without the knob?

It is probably fairly easy to do this. I haven't seen one of these ECU's yet to see exactly how.

4. Add part load/full load enrichment adjustability for turbos/bigger displacement upgrades?

Turbos are limited with the stock D-Jet system to about 5 to 7 psig of boost, due to the limitations of the MPS in handling positive manifold pressure. Within limits, retuning D-Jet for bigger displacement should be fairly simple, by readjusting the MPS, and if necessary, by boosting base fuel pressure up to as much as 35 psig. I don't think it's necessary to modify the ECU to handle larger displacement.

A couple of people have asked me about adding closed-loop control with an O2 sensor. I think it's do-able by adding circuitry to the the TS2 connection, but I don't see that it's at all necessary. Mixture control with D-Jet is more than adequate for good performance across all operating loads. The only reason you might want to have closed loop is if you decided for some reason to fit a 3-way cat converter to your car - which is unlikely.

A more interesting mod would be to adapt an idle stabilizer to the 914. I haven't looked into it but a lot of owners could benefit. Even when D-Jet is properly tuned idle stability is marginal.

Brad Anders

[bowlsby]

Hmmm now thats interesting. My car seems to be in pretty good tune, and usually has a rock steady or near rock steady idle *most* of the time.

Its odd because the car can have a perfectly flat idle rpm under one set of driving conditions, time of day, day of week, its mood du jour or???, and then the idle can start to vary by as much as about 100 rpm (which I understand is within factory tolerance from something official I reecently came across), for unknown reasons.

I always have attributed that variance to a vacuum leak, a miniscule leak perhaps, but a leak none the less. I have recently replaced all vaccuum lines and used sealant at the connections for an airtight fit, but no hose clamps like many do. Then connections seem very tight. I wish I could pressure test the sustem though.

You are suggesting that the ECU or sensors or some other part of the FI system other than the manifold vacuum level may be a culprit for a variable idle, whereas I would think that that part of the system would not be subject to variation. I'll be anxious to follow your findings about this idle variation issue.

[pbanders]

Originally posted by bowlsby:
Hmmm now thats interesting. My car seems to be in pretty good tune, and usually has a rock steady or near rock steady idle *most* of the time.

Its odd because the car can have a perfectly flat idle rpm under one set of driving conditions, time of day, day of week, its mood du jour or???, and then the idle can start to vary by as much as about 100 rpm (which I understand is within factory tolerance from something official I reecently came across), for unknown reasons.

I always have attributed that variance to a vacuum leak, a miniscule leak perhaps, but a leak none the less. I have recently replaced all vaccuum lines and used sealant at the connections for an airtight fit, but no hose clamps like many do. Then connections seem very tight. I wish I could pressure test the sustem though.

You are suggesting that the ECU or sensors or some other part of the FI system other than the manifold vacuum level may be a culprit for a variable idle, whereas I would think that that part of the system would not be subject to variation. I'll be anxious to follow your findings about this idle variation issue.

Jeff,

Mine's the same way. Once warm, the idle is steady, but there are several transitions it takes on the way through warm-up. First, it's low, then high as the AAR takes over, then low when the AAR closes, then stable for a while at 1K, then up to about 1.2K for a bit, then down to a steady 1K after about 10 minutes of operation. Full 100% warm-up happens REALLY fast here in AZ when the daytime temp is 106 F like today!

I know my initial warm-up is low for two reasons - my TS2 is a bit low on initial resistance, and I'm using straight-weight 40W oil which doesn't really get flowing until the motor is near full warm (I'm going to a multi-weight at my next change in a few hundred miles). But the up/down stuff through warm-up is probably due to some strangeness in the ECU. A speed-controled air bleed valve with idle sensing (like those used on Motronic) would be a very nice way of cleaning up the variation.

Brad Anders