[Dan Staples]

Hi,

New to the forum so please go easy!

I have a ‘72 CSL, currently on very old, obsolete and unobtainable Bosch D-jet engine management.

Long story short but after buying the last of “NOS” parts for eye watering prices I’ve grown despondent of having to fix the EFI every year.

The engine is a standard M30 inline 6. I would like to keep the engine as stock as possible. It has a distributor and high impedance injectors.

D-jet is incredibly simple, using only TPS and Map to control fuel. CTS controls choke.

I’m hoping to hide a modern day map and throttle pot in original D-jet bodies and use the standard thermistor CTS for enrichment.

What I’m struggling to determine is what hardware I need. I’m planning on using the 0.4.3 board with mega2560. Will I need any additional hardware for the -ve coil signal to determine engine position?

could anyone tell me if adding narrow/wideband O2 sensors are worth while? If so what is best? Being 6 cylinder I image I’ll need 2 O2 sensors/controllers?

Thanks

[PSIG]

The engine is a standard M30 inline 6. I would like to keep the engine as stock as possible. It has a distributor and high impedance injectors.

Hi, Dan! This should be a fairly simple project.

D-jet is incredibly simple, using only TPS and Map to control fuel. CTS controls choke.

Speeduino can be operated similarly. Are you going to use the original ECM case and connector?

I’m hoping to hide a modern day map and throttle pot in original D-jet bodies and use the standard thermistor CTS for enrichment.

Speeduino already has a MAP sensor on-board, so only a hose connection needed. Your cold start valve is not needed and may be disconnected. Your warmup air valve will work fine as it is.

Will I need any additional hardware for the -ve coil signal to determine engine position?

This is a decision point. You have the option to build a simple circuit for coil(-) input in the proto area of the PCB, for Speeduino to run fuel-only. You also have the option to control the ignition, for better tuning and results. This is as simple as taking off the distributor coil(-) wire, and sending it to Speeduino. Speeduino then sends a firing signal to a new ignition driver module ($10) to fire the coil. Oddly a bit simpler than not controlling it, and what most users choose, but it's your choice this time and you can always change your mind later.

could anyone tell me if adding narrow/wideband O2 sensors are worth while? If so what is best? Being 6 cylinder I image I’ll need 2 O2 sensors/controllers?

Absolutely, WBO2 is much easier to tune, and tune better. You only need one, either where exhaust merges, or on one set of cylinders if they don't. You are not planning to run your engine differently from one cylinder or set to the next, and so no operational reason to watch all of them now.

[Dan Staples]

PSIG thanks for the reply!

I’ll be taking the original loom and ECU off the car. The ECU is located under the rear seat and the loom will be very easy to replicate as it uses only Bosch plugs that can still be found on cars today and brought new!

So for the MAP I can bind a vacuum tube into my new loom and plug directly into the ECU? Great, saves me having to cut up one of my MAP sensors I’ve accumulated over the years!

Ideally want to keep it as simple as possible with ignition, if it’s easier to control the coil rather than leave it standalone then that’ll probably be the route I’ll take.

My only concern really with running off coil- would be injector firing as there would only be 1 reference point. Would this mean that all 6 injectors fire at the same time? I can fit a toothed crank pulley off a later motronic m20 engine that will make little to no aesthetic differences and would this mean the injector pairs will open at 120 degrees from each other? To be honest is this just me overthinking (and tell me if I am), the car didn’t run sequential injectors anyway and later Bosch CIS systems had the injectors always open regardless of engine position.

My main reason for going down this route is I’m a do’er, I like the idea of building my own ecu, even if it is a kit of pre-allocated components that need soldering to a board. I have spoken to other companies that want over £1000 as it’s over 4cyl for an ecu that’s all singing and dancing when really I just want to replicate something that was developed in the 60s!

[jonbill]

IMO, a missing tooth trigger wheel is preferable over the 'basic distributor'.
More accurate timing of events and the ability to synchronise injection with cylinder position.

[PSIG]

So for the MAP I can bind a vacuum tube into my new loom and plug directly into the ECU? Great, saves me having to cut up one of my MAP sensors I’ve accumulated over the years!

Yes, very simple, and testing has shown no detrimental effects to an internal MAP, compared to an external MAP. And, no extra wiring or electrical connectors. I developed a semi-rigid MAP line kit, which allows connections through the case (bulkhead connector), and small flexible N11 semi-rigid line so the MAP "hose" can be bundled and hidden with the wiring in the harness. But you can route it any way you like.

Semi-rigid_MAP_line_kit.jpg (73.08 KiB) Viewed 1244 times

Ideally want to keep it as simple as possible with ignition, if it’s easier to control the coil rather than leave it standalone then that’ll probably be the route I’ll take.

This you can take in steps if you prefer. Fuel-only to replicate the original control, then once you are comfortable with fuel-tuning step-up to timing control, and so on. Taking the distributor points signal for ignition control requires no extra parts. Taking the coil(-) signal for fuel-only does. Example circuit to take coil(-) signals you can build in the proto area, a simplified version of one tested by @zylinx:

Coil-_RPM_input_opto - zylinx.png (16.76 KiB) Viewed 1244 times

My only concern really with running off coil- would be injector firing as there would only be 1 reference point. Would this mean that all 6 injectors fire at the same time? I can fit a toothed crank pulley off a later motronic m20 engine that will make little to no aesthetic differences and would this mean the injector pairs will open at 120 degrees from each other? To be honest is this just me overthinking (and tell me if I am), the car didn’t run sequential injectors anyway and later Bosch CIS systems had the injectors always open regardless of engine position.

Your choices. Fuel-only coil signals would provide 6 pulses per-cycle (3 per-rev), as would distributor signal mentioned earlier. Either can provide 2 or 3 injections per rev (120°) of injector pairs, providing averaged fuel distribution.

Accuracy is nearly irrelevant with fuel-only, and 6-signal per-cycle is good enough to work well for ignition (4-cyl's like the MX-5 only get two signals per rev!). But, yes, a toothed missing-tooth wheel would provide even better accuracy for everything. Again, your choice to start limited-but-functional, or jump straight to a more-optimal setup. Each level would be a good learning experience, so take your pick.

[Dan Staples]

I’m going to go ahead and order the board, initially going to run off coil- purely for the aesthetics of not having to fit a different crank pulley and sensor. If I decide to run the ignition off the ecu at a later date then that’s a step I’ll have to take.

Finally, what would you recommend as an enclosure? There are plenty out there but what do you guys use as a break out to connect the speeduino to the ECU headers? Or do you go to a scrap yard and buy an ecu with the plugs attached and chop away from there?