Skip to main content

On "Microsquirting" the Porsche 914 - Part 6, Ignition, Wiring, and Terminals, Oh My!

Ignition, Wiring, and Terminals, Oh My!

Return to Part 5

The electrical design and the wiring...oof. Had I known in advance how much brainpower and effort this was going to take...let's just say it was a lot of work.
D-Jet ECU Diagram 1 of 3...

I began by learning all I could on how the D-Jet's electrical system was designed. "Why?" you may ask. "After all, you're getting rid of the D-Jet!" Fair point, but recall one of my design parameters what that I wanted it to be as bolt-in (or plug-in) as possible, with minimal modifications to the stock parts of the car so D-Jet can be reinstalled if desired. So I wanted to understand how D-Jet interfaced with the car and engine and possibly leverage that.

I pored over wiring diagrams and schematics writing notes and reminders, and meticulously documented in a spreadsheet what each wire and circuit does. There's no write-up I'm aware of that just offers design concepts straight-up (though Paul Anders does a great job of explaining the general concepts), so I had to learn it. But once I understood it, the electrical system seemed pretty straightforward.
914 Wiring Diagram 1 of 8...

The D-Jet system as used in the Porsche 914 is mostly self-contained, electrically. It has its own wiring harness and connects within itself between the computer and its sensor inputs and injector outputs.

Spark ignition is also isolated. The D-Jet does have a plug going to the Bosch distributor but only to get crank/cam position info from the disty's internal pickup; ignition timing is fully isolated from the D-Jet computer, driven solely by distributor springs and centrifugal weights inside the disty (helped by vacuum pots), and it has its own discrete ignition coil.

The auxiliary air regulator (to idle up the engine when cold) is not integrated; it's just a simple bimetallic-driven air valve that gets power when the fuel pump is running. The overrun valve (adds air when the throttle is closed) is vacuum-operated and also isolated.

Really, The D-Jet "engine management system" is multiple independant systems doing their own thing that happen to coincidentally work together. Elegantly simple.

The D-Jet system only interfaces electrically with the chassis via a single 4-pin wiring plug: one pin to receive power for the computer and injectors; one wire to ground the car's fuel pump relay solenoid; one to sense when the starter is engaged to turn on the fuel pump; and one more pin that is also on the starter circuit to engage the CSV during start (if it's cold enough outside).

Four electrical pins. That's it.

The D-Jet Interface Plug
For comparison, the Microsquirt (MS) system is also self-contained. It needs power for the ECU, injectors, ignition coil, and O2 sensor system. It does not need to know when the starter is engaged, since it senses when the crankshaft is turning (via the CPS). It does not have a cold start valve. It does not have an aux air valve or an overrun fuel cut-off valve. This should be easy, yes?

Theoretically, yes. I could use the same power source in the 4-pin plug to run the MS. And since D-Jet ran its injectors through that circuit, I could use it for the MS injectors, too. After all, shouldn't they both be more electrically-efficient than the 50-yr-old design? And why not go ahead and run the ignition coil through there, too?

Yes, it should be that easy. However -- why is there always an "however?" -- there is a long-standing problem with the 914's relay plate. The voltage regulator is a physical switch (a solid state replacement is available) so it gets very hot, melting the insulation below it and causing corrsion and bad connections. If I added the ignition coil current through it then that would make the problem worse. I read some reports that the IGN4-VW coil could pull up to 15A or more. Probably not good to add that load to the system.

Further, the D-Jet system had power going to the injectors any time the key was on; Microsquirt recommends that the ignition coil, wideband, and injectors receive power only when the fuel pump is engaged. So it made a lot more sense to run the low-amperage (3A?) MS ECU through the 4-pin plug and then relay the others. This a small level of additional electrical complexity, but nothing big.

My design was now to incorporate using only two pins of the factory 4-pin: one pin will provide power to the MS, and one pin will be connected to the MS to have it ground the stock relay that engages the factory fuel pump (same as D-Jet did). The other two pins will be unused.

Those factory 4-pin housings are not common; they have a unique shape. But with the proper tools they can be de-pinned without damage and re-used for the MS wires. And, I learned from Jeff Bowlsby that NAPA part number "NW 725165" are direct replacements for the terminals...just mark which wire goes where in the pin housing before you remove them in case you want to go back!

The next decision was, "where do I mount the ECU"? It's a compact little box, around the size of a couple deck of cards -- or a couple packs of cigs, if you're older. The MS unit is listed as "splash resistant", but the case halves are not sealed. I read reports of users removing the cover and sealing the case halves with silicon sealant to make it more water-resistant (I heard the MS was used in some kind of recreational watercraft.)

The MS manual specifically states:
[The ECU] is not designed to be installed in the engine compartment. Typically it will be installed under the dash in a car or under the seat on a bike - but away from direct engine temperatures. The temperature must not exceed 185°F (85°C.)
Some 914 owners have installed Megasquirts and mounted the ECU and relay box(es) in the rear trunk; that's what was done in the one I assisted in getting running and idling. I also considered mounting the MS ECU under the relay box in the engine compartment. "The engine compartment?" you ask? Sure, but I suggest the 914 engine compartment is somewhat unique: it's an air-cooled engine with sheet metal and seals separating the top cool air intake side from the bottom hot side where the air-cooling exit and engine exhaust are. Since everything on the top side of the engine is cool air, and presuming your sheet metal and seals are in good shape (mine are new), then you really should not see temperatures approaching 185°F up there; if you are, then something is really wrong. After considering the rear trunk, maybe under the relay plate, and maybe even somewhere inside the car, I decided to mount the ECU on the left side frame rail right in front of the rear trunk. It's isolated from the hot side of the engine and minimizes the length of the MS wiring loom.

Left frame rail, from inside engine compartment looking toward the driver's side.
MS ECU is at the far left


3/28/22 Update: I've experienced zero problems with the ECU being mounted there. It does not reach 180* on the top side of the engine compartment. - GA

Side note: you 914 owners will catch a few unusual things in that photo. First, my return fuel line is on the driver's side instead of the passenger side. I suspect that a previous owner installed the stainless fuel lines and decided to route the return to this side. That actually worked out good for me as I don't have to route the return fuel line all the way back over to the passenger side.

The second item of note is the red power terminal there at bottom middle. My car clearly had some "Hell Hole" cancer (and it has evidence of less-than-professional repair) and it no longer has the battery box installed; it's just a big empty cavern on the right side. My battery is now installed in the rear trunk, left side, with the positive cable going directly down through the floor to the starter; from there positive goes through the vertical sheet metal to that power terminal. All of the car's smaller positive battery wires were moved to the left side to this power terminal (that I added).

Other useful items in the photo: the wiring harness (with the yellow wire) from the T12 plug (starter, oil pressure switch, backup light wires, tach input to car). From that plug comes the T12-12 wire, which used to go to the D-Jet's auxiliary air valve to turn it on. That wire is hot when the fuel pump relay is energized. This is the wire that I chose as an output to energize the secondary relay, top middle in the photo. That relay provides power from the red battery terminal to the engine connector harness, and it will supply 12V power to the injectors, ignition coil, and wideband controller -- but only when the fuel pump is running. That all got tidied up into a single 8-pin Deutsch connector to have one connector to the engine (more on that later).

Whew! And we've not even gotten to the engine part yet!

There were a few other design considerations I was tossing about at this point:
  • I wanted to avoid -- if not outright eliminate -- butt connecting or splicing wires; everything should terminate inside a pin housing.
I'm bothered by seeing a lot of spliced and butt-connected wires. I'd rather the wires extended and crimped directly into their proper terminals. If I need to put in a full connector to make it "disconnectable" that's fine, but I don't want single-wire terminals all over the place.

Toward this, I bought the Microsquirt with the 8-foot lead instead of the 3-foot lead. I probably could have gotten away with the 3-footer, maybe, but I am a lot more comfortable having to trim wires away instead of adding length.

I did not use any connectors with pigtails spliced into the Microsquirt wiring harness. Instead, all connectors were bought with bare terminals, crimped into the MS harness, and inserted into their empty pin housings. No butt splices!
  • All wire functions must have a discrete color
Don't use the same-color wire for multiple functions. When I see a purple or a red or a white-with-black wire, I want to know exactly what it's for.

Except for 12V power (red) and chassis ground (black) all functions were guided by MS's wiring color schematics.
  • All sensors/components should be easily replaced without having to solder or repair terminations
Almost all the sensors have a plug that will accept a pin housing; those pin housings would easily disconnect and reconnect to a replaced component. I can do that with all sensors except the DubShop CPS (but then again, replacing that one may require removing the engine anyway...)
  • Use industry-standard pin housing and terminals, and standardize on one style
This one was pretty much impossible to meet 100%. The TPS is Bosch so it is likely MetriPack; the VW IGN4 coil is probably Metri-pack, too; the MPS, IAT, and CLT are Delphi (GM) parts, so they used Weatherpack-style pin housings.

On top of that, I'm a big Deutsch fan, I really love their connectors. Small/compact, efficient, weathertight. I built a harness for the race car using Deutsch barrel pins and connectors using a tool I borrowed from a buddy; I love that tool. But if I made this and someone wanted to replicate it then it needed to be done using common tools. I'm presuming that most people have a "w" crimper (like what's used with Weatherpack) but not so much the Deutsch. Further, as recall, the Deutsch connectors aren't that cheap...so I was leaning toward standardizing my connectors on Weatherpack. 

Then I found this kit on Amazon. Deutsch connectors with pins and seals for $35 delivered. What caught my eye is that it uses stamped connectors, not the barrel connectors; I was not aware of those! Stamped uses the standard "w" crimper (that everyone maybe has, right?) plus they also strain-relief on the insulation. As with the barrel connectors, these pin housings are fully weatherproof. Of course, if someone prefers the barrel connectors, the design will work with those, too.

Bought two sets. I now have my connector standard: Deutsch.





Comments

Popular posts from this blog

An Interesting GTSB (Greg Transportation Safety Board) Report

An Interesting GTSB (Greg Transportation Safety Board) Report (Subject to revision) On August 29, 2024, at approximately 6:30PM (EDT), the engine in the #33 Porsche 914 suddenly stopped as it was slowly driving away from the town green of Falls Village, CT. Hearing a large noise (discernable by nearby onlookers), the driver immediately disengaged the drivetrain via the foot clutch and car coasted to the side of the street. Initial attempts to restart the engine were met with a starter that would not rotate; attempts to push the car while the transaxle was engaged were met with full resistance. It was quickly determined that the engine had locked up. After minimal roadside investigation, it was determined that the dry sump tank oil valve, which supplies oil to the engine oil pump, was in the closed position and it was obvious that the engine had been run without a supply of oil. As a result, the car was "flat bedded" back to Lime Rock Park and retained as a paddock display for...

On "Microsquirting" the Porsche 914

Bosch D-Jetronic The Bosch D-Jetronic system is pretty cool, especially when you consider it was designed in the 1960s. "Computer"-controlled electronic fuel injection with manifold pressure sensor, intake temperature sensor, crankshaft (well, distributor) angle sensor, and throttle position sensor/switch. It uses constant fuel pressure and flow, so only injection duration needs to be modified to control air/fuel mixture. It measures incoming airflow by monitoring the intake manifold pressure; engine speed, temperature, and other factors are monitored for the purpose of fine-tuning injection duration. Ignition is by a standard cam-driven distributor with an internal D-jet-specific pickup for the crank/cam angle position. This "speed-density" D-Jet system was used on many cars of the period, including Volvo, Jaguar, Volkswagen, and of course, the Porsche 914 (1.7L and 2L engines only; the 1.8L used L-Jetronic -- "L" for "luft" or "air...

On "Accusumps"

(This is another blog in a series of thoughts/ideas that I commonly see, created as a placeholder for future reference so I don't have to re-write it each time. I totally understand the limited audience and for such a post, but I invite feedback from those who have interest in this arcane subject...) -------------------------------------------------- This was posted in response to someone inquiring about PDM control, pressure sensors, switches, valves to manage an Accusump in his race car... You're making this wicked too complicated...and significantly failure-prone. An Accusump (AS) is a passive hydraulic/mechanical oil pressure accumulator (thus, "Accu"sump). Its purpose* is to provide oil flow when the pickup tube in your engine becomes unported due to side loads oil "slosh" or flow away from it. First, note that an AS is a "BandAid" for a design flaw in your road race engine. That design flaw is the unporting of the oil pickup. The ultimate res...