and in with.... well... MORE MEGASQUIRT!!!!

Little back story here and then a bunch of pictures.

I have several 914s but the one I refer to as my "autocross car" (which started as my daily driver in the late 90s) has AFAIK been running Megasquirt possibly longer than any other 914 on the planet. For the last 17 or so years it has been the platform I have experimented on and in that time has seen at least 5 different Megasquirt configurations across 2 different motors. This ongoing project is one of the reasons I am so intimately familiar with both Megasquirt and D-jet as to date all the configurations I have ran in this particular car have done so using an unmodified d-jet wiring harness which required a good deal of study/reverse engineering.

At one point my focus was on making a completely stealth setup (looks exactly like a d-jet system) which I did, however as the Megasquirt feature set improved over the last 17 years and my focus moved more towards the overall performance of the system I kept hastily hacking in additional features (mainly having to do with the addition of full ignition control and O2 feedback) and an ancillary rats nest started to grow around my once stock looking system. Remembering what all extra wires plugged in where started to become tedious, especially if I would go extended periods without doing anything to them. With the desire to change/add additional elements to the system I decided I needed to just bite the bullet and build a whole new system from the ground up. Technically i have been planning (sitting on the parts) to do this for years but as the current system ran just fine it was lower on my priority list. The whole COVID situation has allowed me lots of time to get pretty much every other project done on all my other cars, so it was time to move forward with this one.

Details on the system coming out:
ECU: Completely custom (see homemade) ECU board based on the MS1 2.2 circuit design however I modified the injector circuits to supply a switched current rather than a switched ground in order to work with the D-jet harness, and then later retrofitted an MS2 CPU to expand the feature set.

Injectors: Stock 914 2.0 off a stock pressure regulator

Ignition: Full timing control currently triggered via Pertronix pickup modified with a pullup resistor in a stock distributor with a locked out mechanical advance. Output is then from the MS system to a Mallory 6AL CD box (this was a really easy way to trigger a coil from Megasquirt back in the day as it can take a logic level input)


The Goals of this project
1. Cleaning up the engine bay wiring and simplifying the connections (single harness plug to the stock relay board and thats it!)
2. Move to a 36-1 crank trigger
3. Add a 2nd MAP sensor for real time barometric correction (we have some high mountains around these parts)
4. Eliminate the distributor and CD ignition box
5. Swap the LC1 wideband controller for the newer(and smaller) LC2
6. Use a modern Fuel Injector with known voltage offset values (will get more into this later)
6.Stop using the d-jet harness (Will elaborate on this as well)


WHY???!?!
Back at the end of the 90s I was a d-jet purist, carbs to me have always been a step back and never an option for me, so when my d-jet system started acting up one of the things I did at the time was to purchase a new wiring harness made by a company called R.E.S. Systems. This was the only NEW replacement harness available at the time and is still an absolute work of art as far as i am concerned, overbuilt like no other d-jet harness you have ever seen. Needless to say it was not a piece I wanted to take off the car and replace with whatever homemade harness my limited experience could coble together at the time, so i invested time in figuring out how I could retrofit an MS system (also very new at the time ~2003) into an otherwise stock d-jet setup. Now, while doing this does result in a system that works as well as a good d-jet system I am here to tell everyone once and for all that D-jet does have some inherent faults with its design and that those faults become very apparent when you can see them in data logs and also when you realize that a lot of the known d-jet operational "quirks" follow you when running a modern ECU on the d-jet harness. The nicest d-jet harness in the world does not eliminate these issues, they are inherent in the design.
So in the interest of having the best running car possible a new design harness is needed, which is fine by me as i want to integrate everything into a single harness anyways.
Other things to come with the new build, modern 43 PSI injectors should atomize the fuel better, in addition having known voltage offset tables is mandatory to achieve the most consistant tune across all operating conditions. The change to the 36-1 crank trigger is for increased accuracy but also to improve ease of cold starts over the 4 tooth cam trigger.
Getting rid of the distributor was not mandatory but I had an idea for a new setup I wanted to try out, so its getting worked in the changes as well.

Had a week of vacation time to burn after Christmas that seemed like the perfect time to get this all done so...

on with the show...

Will start with the "before" pictures, this is the rats nest of years of hacks im looking to clean up. The ECU there is not d-jet but contains my custom MS board internally (note the blue vacuum line running into the bottom of it.

Click to view attachment



Other than that the system is very "stock-ish looking" though on the latest motor I changed things up a bit. On my previous 2.0L motor I was running with the stock 2.0 intake step, the current built 2056 has modified 1.8 heads so im using a 74 1.8 plenum modified to accept a 50mm throttle body from a 2.1L Vanagon... almost looks like it was supposed to be there!

Click to view attachment




Weeks prior I had "borrowed" the 2.0 injectors off this car to put on a 74 bumblebee im bringing back to life so figured the injectors and pressure regulator were a good place to start in order to have the fuel system buttoned up before messing with the wiring.

Out with the old....

Click to view attachment



and in with the new...

New high impedance injectors (to eliminate the need for resistors) also have a 3 bar minimum pressure so the regulator has to go. New piece from Mario at TheDubshop that accepts modern bosch regulators going in its place with minor modification to the engine tin.
Click to view attachment

Click to view attachment

Click to view attachment

Factor contributing to the rats nest, coming out!

Click to view attachment


And the ridiculous amount of wiring connection the LC1 wideband needed

Click to view attachment

and now the part that makes me sad... pulling the beautiful RES Systems D-jet harness. Almost 20 years later this thing still looks practically brand new. just a tank of a wiring harness with every connector and junction properly sealed to water proof it and every connector labeled.

If i don't wind up putting this in another car it will probably go up on my garage wall as art!

Click to view attachment
Click to view attachment

New ECU is just going to be an off the shelf Microsquirt. No more 12 hours of labor custom building circuit boards for me!

I had a gutted d-jet ECU case from another one of my previous MS builds, so i re-purposed that as a bracket for the ECU, both MAP sensors and the new (much smaller) LC2 wideband controller.

Click to view attachment



My old MS 2.2 custom board d-jet harness compatible ECU next to the new Microsquirt setup, not quite as trick looking but I may revisit that later.

Click to view attachment

And now the REALLY fun part... putting this in there... and trying to make something out of this new mess!

Click to view attachment


Laying it out as an initial test to get a rough idea of spacing from the ECU bracket to the relay board.
Not looking complicated at all at this point

Click to view attachment



Making the runs for the injectors and MAP sensors
Not using pigtails for this build, everything is color coded un-spliced wires out to crimped pin connectors. Minimizing the the junctions minimizes the potential problems. NO SOLDER!

Click to view attachment

After i got the spacing relative to the engine I pulled it back out to mock it up on a spare engine on the floor.

Click to view attachment
Click to view attachment
Click to view attachment


Decided to leave all the unused wires in the Microsquirt harness intact at full length in case I wanted to add something else later (I tend to do that) so they got sealed up with liquid electrical tape and cover in heat shrink. A couple wires im not currently using i still ran out to where i may use them, as I already have some addition upgrade in mind

Click to view attachment


Pulled back on the bench to get everything sheathed and connectors installed.
Click to view attachment

Just about ready to go, just need to add fuses inline on the power circuits at the relay board and finish sizing the wire lengths to my coil experiment. I was going to go with a familiar setup there but i just cant help myself and had an idea for something new!!
Not nearly as pretty as the RES harness, but i think i did ok. May revisit it again later as im not super happy with the braided sleeve.

Click to view attachment

And now for the next surprise idea(s).

I want to keep the plug wires all running through their stock locations so I took an idea (and a spare distributor) down to Ye local house of 914 art (AKA PMB) and had them Frankenstein this piece up for me. They did this way prettier than I ever would have been able to.

Click to view attachment


Why you ask....
Click to view attachment


I had a very specific coil pack in mind. Looks almost like it was designed to go exactly there.
Click to view attachment

The coil pack is the biggest unknown/experiment of this current build as I was not able to find a lot of documentation on it or reports of people using it.
I have previously used VW 2 channel logic level wasted spark coils on build i have assisted people with and good results however wasted spark coils have a short coming in that 2 of the plugs fire with reverse polarity resulting in increased and uneven spark plug wear. As i am going for maximum precision and reliability here (and i don't enjoy regular plug changes on Type 4s) Newer cars run dual platinum plugs to offset this wear, but i wanted a solution where i can run the stock copper plugs. I figured i would experiment with a different potential solution.

This experiment comes by way of another VAG coil pack, only this one is a 4 channel (CPP) logic level pack. First issue there though is that without additional harness modifications the Microsquirt only controls 2 coil outputs however being a logic level coil in theory I can tie the input pins together in pairs and batch fire them, essentially firing the same as a wasted spark but with all plugs using the proper polarity. Only concerns I have are is the logic built into this coil going to allow me to fire off 2 coils in the pack simultaneously without issue, and as this motor spins to 7k RPM and ill be firing the coils at essentially twice the rate they would in a factory application, can they take it or is there protection built into the logic circuitry that wont allow this???

TBD

It sure does look cool though!

Click to view attachment

The entirety of everything that came off the car in this endeavor

glad to be rid of the mess


Click to view attachment

Pre flight checks on the new 36-1 crank trigger looking good!
This should help a lot.

Click to view attachment



Static tests from Tuner studio confirmed injectors are firing properly and
I can indeed batch fire the coil pack as hoped.

Only issue during pre-flight was a fuel leak at the new regulator during the fuel pump relay and pressure test. This was at the plugs where you can install pressure sensors. Had to take some extra measures to seal them up, but all is well now.


Always nervous as hell to fire a new MS build, but happy to report she fired up on the first attempt.

Now to wait for warmer weather so I can get a proper tune on her.

James, that's excellent. Thanks for posting this. I've got a small rat's nest in my setup as well, and I'm looking to clean it up with better wiring connections and routing. But I'm in the same sort of boat -- there's a tension between wanting to test something and improve it versus cleaning it all up and declaring it done and final!

One question: what model injectors are you using?

Again, looks great and thanks. Hopefully you do get some warm weather soon

Incredible! Your skills are impressive.

Sounds like we're brothers from another mother...

I can't say enough how happy I am with the Microsquirt system I did last winter (with "bespoke" harness).

However, I'm now building an AW11 Toyota MR2 race car, and I am even more impressed at DIYAutotune's PnP system for it. I mean, it is seriously one well-thought-out bit of kit; I'm shocked that not all AW11 cars are running these ECUs. If we could only design and fab a PnP system for the 914 DJet harness... On the other hand, I can't leave well enough alone either: I've already ditched the AFM, added wideband, and I have COPs and a 36-1 coming in for it...

Subscribing! - GA


Edit: for future consideration, I am now using the 14Point7 Spartan 2 Lambda controller for my installs. Tighter install and has a much heartier lifecycle reputation than the LC2 (I've already replaced the O2 sensor in the LC2 in the street car). I get mine from a buddy, Conover Motorsports: https://www.raceconover.com/shop/spartan2

I. LOVE. THIS.

Care to make another?

@Montreal914

Wow! I'm looking for the Google translator to put this into dumba** English for my consumption. Impressive. Thanks for sharing.

Nice !!! work James !

You would have to ask @Eric_Shea about that one. His guys at PMB did the fabrication work for me as I didn't have the necessary tools to do it myself. I brought them a core distributor and an idea and they returned that beautiful piece. In the end it is just a hacked up cheap distributor (i wouldn't waste a d-jet one) with a plate welded on it but the result is exactly what I was looking for. Really nice for a one-off part but im not sure it would make sense to do it this way in a production environment given the labor and need to sacrifice a distributor. If someone wanted to produce these in any volume I suspect it might make more sense to just machine the whole thing.

Understand.

I use the VAG IGN-4 ignitor/coil and it work great. However, my initial installation was on a steel plate I fabbed that sat across the disty retention stud and two engine tin screws. As the weather got hotter this Spring I started getting a lot of ignition cut-outs...short answer was the IGN4 was overheating, sitting on that plate mounted directly to the engine block.

Short term I moved it to sitting on top of the cooling fan, and I've yet to come up with an elegant replacement solution. I like yours and suggest that may keep the coil far enough away from the engine heat to allow air to circulate around it.

I've got an old 009 hanging about, might try fabbing something.

Greg, what about a phenolic block spacer, kinda like what was used under the runners at the heads...

It is hard to tear something out again when its working just fine. Little easier for me in this case in that the shortcomings of the d-jet harness were just becoming way to apparent. Once i get a tune on the new setup ill have to post a comparison of datalogs from the old and new systems as they are a night and day difference.

As for the injectors, they are an off the shelf part from FiveO Motorsports so no need to go sourcing anything used from a wrecking yard.

https://www.fiveomotorsport.com/a280-a380-h...-fuel-injector/

Lots of advantages over stock d-jet injectors, New(not 50 year old NLA parts) High impedance, Ev1 connectors, standard 3 bar pressure (try finding a good 2 bar d-jet pressure regulator these days) more reasonable flow rate (d-jet 2.0 injectors are pretty oversized) and probably my favorite part is that they come with documented flow data and more importantly, the voltage offset data. I was actually surprised at how low the dead time on these was when running at 3 bar. Knowing the proper dead time and voltage correction is critical for keeping the mixture consistant under varying system voltages. If you have issues with your idle dropping due to changing mixture when you turn on your headlights this may be why. I was so excited to test this that turning on the headlights was one of the first things i did once i got the idle mixture mapped in. No more idle dip! In theory someone could deduce the proper value for the d-jet injectors through trial and error, i just was never excited enough about doing that work.

After working on so many MS systems for myself and others I have developed a preferred list of hardware that I recommend, and these are my pick for use with stock intake runners if you want to eliminate any compromises.

Available in higher flow rate to if needed, or you can up the pressure as 3bar is the minimum spec on these.

Click to view attachment

Heat is one concern I have in the back of my mind for sure with this coil setup but wont know what the end result will be until it warms up and I can get it out on the road. Like i said this coil pack setup is probably the most experimental part of this build so I cant yet speak to its reliability.


Assuming the coil you are running is the VAG 032 905 106B wasted spark coil? In the past I have had good results with when working with that pack however never tried it mounted where the stock distributor goes. The one issue I did observe though (and one of the reasons for trying something new) was some unusual/high rate plug wear. The plug had completely eroded the center electrode in less than a couple thousand miles. At the time I was not sure if it was due to a defective plug or the ignition system itself but after doing further reading it sounds like this is the nature of wasted spark coils due to the reverse polarity on half the plugs, band-aided by manufactures with the use of platinum plugs.

Curious if you re-checked your plug gaps after running the coil for a while now?

Also curious what your total dwell is when you are running? Wondering if excessive dwell may be contributing to excess heat in the coils? Wasn't able to find much specific information on either one of these packs around dwell requirements but information I was able to find on other VAG coils that was extracted from VW ECUs seems to indicate VW in general runs their coils with far less dwell than the suggestions I have seen on MS forums and manual.

Given this, I aimed on the lower end of the spectrum for the initial setup and am not seeing more than 2.8ms dwell when running after voltage corrections are applied and this value may even still be high based on the VW ECU tables I have seen where they are running roughly 1/2 of that. But again, no idea really as I have not been able to find any concrete documentation. Time/experimentation will tell. So far though the coil has been cool to the touch, granted this was only ran for about 20 minutes with no load with weather in the 30s. May be a different case when autocrossing when its 105 out.

One thing I considered doing when installing the pack in the first place was using spacers on the bolts between the coil and the mounting plate to allow air flow under the coil as i believe that is how some are mounted in the factory applications. Didn't have spacers on hand but i may go back and revisit that as well.

ill let you know how it turns out.

That's the one, and I have not re-checked spark gap. I wasn't aware of that erosion problem, but knowing that I'll check them; I've got about 2500 miles on the system so far.

When I was troubleshooting the cut-out issue I ended up backing off the dwell settings from where Mario (DubShop) recommended. I'm currently running Standard Dwell, 2.2ms Nominal, 1.0ms spark duration. Runs/starts quite good; I'm figuring it would work fine with less.

I have not dinkled with the battery corrections; it's at 88%@16V, 102%@14.0, 128%@12.0, 168%@10.0V. I'm running a Lithium battery so field voltage tends to run high, around 14.2-14.5V running.

It runs warm to the touch after dirivng for some time, but not so hot that it will burn. But it is in the engine compartment so hard to gauge. It was wicked hot when it was mounted to the block; my infrared indicated 180* at one point. Can't really say if that was due to higher dwell or being mounted to the block, but the cut-outs stopped immediately after I moved it.

@Porschef , I think it needs airflow, like James describes. @Montreal914 is building his in the same location I put mine but with spacer offsets.

indeed brother! I kept an eye on your build when you posted it, I was impressed as a lot of people struggle with MS builds but it looked like you knew your stuff. Very clean.

Ill have to crack open that ECU I removed and show you what is inside. PnP to D-jet is basically what it is though, should plug into any d-jet car with minimal effort. I added other features later on hence the additional wires out the bottom, but it originally looked 99% like a d-jet ECU from the outside.
One issue with it though is it wouldn't be legal to sell as in its current state it violates the MS license agreement by running a custom PCB. The Microsquirt module used in the all the PnP systems these days wasn't around back when i built this thing so i just had to wing it. Somewhere on an old hard drive I have the gerber files for a PCB I designed using the MS 2.2 circuit diagrams that would have been a direct replacement replacement for the board in a d-jet ECU. I think I may have posted pictures somewhere on this board of it back in the day. Never went so far as to get it printed though due to all the legality questions

A similar thing could be done with a lot less effort and legally these days using a Microsquirt module on top of a breakout board to the d-jet harness, but knowing the limitations of the design of the d-jet harness itself and the fact that most peoples harnesses are pretty trashed at this point, I question if it would be worth it.... If you have to replace the harness anyways due to damage may as well replace it with something of a better design. Just my current feelings on it anyways after having ran with a d-jet harness for so long.

I have had various ideas and designs for PnP ECU replacements as well as PnP systems that involve harness replacement in my head for years, Im not in the parts supplying business though and really just get enjoyment out of coming up with this stuff. Sometimes I seed the ideas to others to let them run with it. There may or may not be a shop local to me now working on some other parts to facilitate easy Megasquirt conversions, at least for their customers cars, but i don't want to spill the beans and will let them speak to that if they want

Just read the whole thread! Super nice installation!!

I did a pause in my Microsquirt conversion to address rust repair before getting my car painted. I am definitely in one of these "while I am in there" thing...

I am looking forward to get back into it, maybe while the car is at the body shop. The power train is currently removed and complete. Probably fairly easy to make a temporary setup to try starting it. But before that i will need to complete the Microsquirt install.

Here is my coil bracket I did out of ~16ga steel. It also supports the MPS and does mount on the tin and distributor stud. Raised, it incorporates three studs so it supports the coil like it was intended. The coil is off of the bracket base by a small gap, but after @GregAmy 's experience, I fear it might overheat. Also, since then, I went with @JamesM recommendation and got another MPS for barometric pressure, so that will need a location too...

Click to view attachment

Click to view attachment

Click to view attachment

James, will there be any cover on your ECU or it is just a support structure? I have been debating as to where to put the ECU. Greg cleverly installed his underneath the relay box, I was thinking maybe inside the cabin, away from the heat. The stock location is appealing though.

Definitely following this thread closely

That's where I was, too. Once I committed to needing a new harness, the possibilties opened up...and I suddenly found myself surrounded in wires and wire bits...

If we could find someone to fab up these harnesses - I have mine fully documented - these could be easy mostly bolt-on installs for someone else. But, as with you, I'm not in the parts business and I don't want to get into the biz of fabbing harnesses, I just couldn't do it at a price that would make sense to others.

- GA

Yeah, for sure check all the plug gaps then. It may have been a fluke thing we encountered with a bad plug but after seeing it occur I did some more reading/research on the subject and turns out this is a known side effect of wasted spark systems so worth keeping an eye on.

In your case though it does really sound like you have it isolated to being heat related.

I believe there is actually something in the works (or at least discussions of such) between a couple well known vendors on this front. I think some people are starting to realize that in addition to all the benefits, after market injection of some sort will eventually wind up being a necessary for most to keep their cars drivable.

Already seeing similar things in the Vanagon world where complete bolt on systems are available from vendors. I think once a 914 specific system is widely available and base maps exists for a standardized system out of the box that will have them already 90% dialed in after bolt on we should see a lot more interest and adoption. Interest for sure has already been picking up but the steep learning cure and fab work is a blocker.

2nd map sensor is for sure recommended if you have some extreme elevation changes, my weekend drive has me going up about mile in altitude and the absense of live correction was noticeable at the top. Without it the MS still does barometric correction however the correction is based on a set altitude where you start the car rather than a constant correction based on your current altitude.

From what I have read though, due to vibration and heat you dont want the MAP sensors mounted on the engine itself. I had considered just putting a plate where the stock d-jet pressure sensor mounts on the passenger side of the engine compartment and mounting both sensors there but as I had room in the ECU case i just mounted them there to get them more out of the way. As for covering the ECU assembly, not at this time. I would have liked to do as i had previously and have everything buttoned up inside the stock ECU housing however there isnt enough depth in the housing to allow the Microsquirt to fit unmodified, so for now its just a bracket. Had a couple ideas but still noodling on it. Its not bugging me to much yet though as it is still massively better than the rats nest that came out.

Great read James. Excellent results. I applaud your work but mostly your desire for improvement. I remember watching for your updates before I went to the Dark Side several years back. I think your pioneering efforts will allow many to experience all that the "Type 4" can be. Keep up the great refinements. Hopefully I'll see you at RRC 2021. Kent

Someone needs to speak up against what some may misperceive from some of the potentially misleading rhetoric of this thread. The passion shown for wanting to improve is interesting and admirable but comes with words of caution: Many things in this thread can be misinterpreted by those that may be lesser informed. They read words like these implying defects to the OEM systems and components:

"...Lots of advantages over stock d-jet injectors..."

"...but knowing the limitations of the design of the d-jet harness itself..." and "...Just my current feelings on it anyways after having ran with a d-jet harness for so long...".

And then surprisingly with caveats like "...One issue with it [the suggested 'improvements'] though is it wouldn't be legal to sell as in its current state it violates the MS license agreement by running a custom PCB..."

There is nothing fundamentally wrong with OEM injectors, harnesses or other parts...although they may need maintenance replacement now or in the future like any car does. Millions of cars from that era were manufactured using these simple systems and components with many of these cars still roadworthy today.

Come down to reality gents. Acknowledge that you are merely exploring nuances, refinements and subtleties at a much more intricate level than D-Jet or L-Jet was ever intended to be, and that it does not functionally need to be. You can reengineer and refine and throw lots of R&D time and money into creating one-off 'betterments' that are interesting and may offer subtle improvements but are not essential to a correctly functioning, reliable, easy and economical to maintain TIV engine unless you just are interested in the exploration and process. Hats off to you for all it takes to pursue these efforts but don't pose them in the light of being desirable or essential to the masses, or worse that D-Jet/L-jet are somehow deficient, or less worthy than the improvements proposed.

Countless VWs, 914s, MBs, and other cars run these stock FI systems daily even today with their stock components and they work as intended. Maintenance replacements for everything are available unlike one off solutions that may need specialized or custom replacement parts that may not be available in the future - and someone knowledgeable to work on/maintain them.

The stock D-Jet (or L-Jet) are reliable, well-designed and built, and durable for what they were intended to do given the state of technology of the 70s for a production vehicle. Its much less expensive and time consuming to just maintain the stock system with stock parts for the average 914 owner who wants to just simply drive and enjoy their car.

Chasing the extreme may be OK for some to pursue but its not necessary or even desirable for the masses.

By the way...I agree with you on one thig that you are not comfortable with....loose that corrugated slit harness casing unless it is just for temp mock-up purposes, it does not protect against water getting into the harness and its electrical connections. The poly braid is no better and not suitable for our open engine bays. Use a sealed, heat-resistant casing...like the factory OEM heat shrink tubing or better.

Hey Jeff! figured I would see you here eventually.

Was thinking about you while I was fighting to piece together a working D-jet system on my bumblebee a few months ago. Cant count the number of times I said "I wish i could just throw a Megasquirt system on this", but some things are to sacred.

D-jet is a fine system for what it is and has held up surprisingly well for 50 years now. I didn't start this post to debate that point. My commentary on this comes mainly from an analytical point of view and i dont believe anything stated here to be missleading Honestly I could probably spend hours going over the improvements in injection technology since d-jet but I'm not here to prove anything, just share in 914 adventures and discovery (though i will probably wind up going over a lot of the points eventually just because i love talking about this stuff)

You seem somewhat defensive every time this subject comes up and im not sure why as there are some non-debatable points on this matter that are just undisputable realities that we need to face. For instance:

311 905 301
280 160 001
311 906 081 A
022 115 542

Those part numbers and many others present a problem as to sustaining the operability of our cars long term if we remain on d-jet, that is just a fact of life. Aside from the car I enjoy experimenting on I have kept my other 914s all running d-jet to this day, but that road is getting real bumpy and we may be approaching the end. This isn't just about improving, there are many factors to take into account.

For instance (and i don't mean to get anyone riled up by i know this next statement probably will) Some people (particularly those selling engines and servicing customer cars) are starting to see d-jet as a potential liability at this point. Consider the fact that a Type 4 motor will run surprisingly well across a fairly wide range of fuel mixtures and timings, so well in fact that many wouldn't even realize something may be amiss with their car if they have not added multiple additional gauges to monitor. Unfortunately the actual AFR+timings needed that keep that same motor from melting down over the long run is a much narrower window. Its pretty well known at this point that due to the analog nature and design choices of d-jet your mixture is already drifting around quite a bit just in day to day operation, add to that 50 year old components to which new replacement parts are NOT available and you are basically gambling with your motors longevity. I wouldn't be speaking about this stuff if I had not seen it first hand. Was driving behind another member out to RRC a few years ago and got my car bathed in oil from his exhaust after his faulty 50 year old injection (that happened to be running well enough not to notice there was an issue) burned a hole through his piston in his recently rebuilt motor

I personally dont enjoy driving with one eye on the gauges at all time, I want to know my injection is doing what its supposed to be doing without me babysitting it.
You may call that nuances, i call it piece of mind. Really the only reason I am comfortable driving my d-jet cars across the country, is because I am completely comfortable with the fact that my motor may very well blow up before i get there. Not everyone can say that.

And all due respect but saying things like:



tell me that 1. You have absolutely no clue what we are talking about and 2. You just seem to be looking to troll this thread

And yes, d-jet does have fundamental deficiencies that go beyond the fact that its 50 years old and new parts are NLA, This is evidenced by ALL THE IMPROVEMENTS THAT WERE MADE TO BOSCH INJECTION SYTEMS OVER THE LAST 50 YEARS, not to mention the obvious fact that Bosch was attempting to chase down and correct issues with it throughout 914 production. Why do you think they added that knob to the ECU in 72 among other system corrections over the years? If you want to educate yourself on this matter you could start by reading up on why critical sensors shouldn't be grounded directly to the engine. I may get more into all of the other specific issues eventually but really not interested in spending any more time defending against a trolling at the moment.

To get this thread back on an more interesting note...

Cracked open the insides of the MS to D-jet Franken-ECU that I have been running on the car for the last 10 or so years to take a picture for everyone, was a lot prettier before the additional hacks i made to it. Thinking ill probably wind up doing a little clean up and re-purposing this.

I actually have a 2nd board I had built back then intended to improve upon a couple things i didn't like about this one but never wound up installing it in a case

If anyone wants specifics on how to accomplish this I can go into it, but I wouldn't recommend most, or really anyone, take this route.

Click to view attachment

Jeff, I think you misinterpret the Injector and the harness topics.

There is nothing wrong with the original low impedance Injector if you run D-Jet, but they are not ideal to deal with if you use them with a modern EFI system. Same with the wiring Harness, it's working perfect with D-Jet but with a modern EFI system you do not want to coupple in any noise (as digital systems are much more sensitive to noise) and therefore you have a different grounding philosopy as well.
That said, for modern EFI James way of working is the right one.

At the end everybody needs to decide himself if it is better to stay original or upgrade to something else, and some people out there even call it an upgrade if they remove D-Jet and mount carbs instead... All about personal preferences...

Take care,
Frank

James,

I did not work with Microsquirt so far and am curious how you set the voltage correction for the injector dead time as it is not linear. Does Microsquirt have a none linear correction table like MS3, or how do you deal with the problem?

Thanks and best Regards,
Frank

Hi Frank,

I don't think it has a non-linear voltage correction for dead-time. It does for dwell, but for dead time settings I think you get one shot as per picture below:

Click to view attachment

Well, I hesitate to freely offer my tune, simply because of the risk of someone simply bolting on and walking away.

But my engine is a dead-stock "Lord knows how many miles" engines and mine is a solid, safe tune. If a vendor were to offer the same exact components as I am, I'd be comfy giving them this tune as a baseline for their products, even without wideband feedback for EGO correction.

The aftermarket for something like this is an oddball, as you always risk someone taking componentry made for a different config, bolting it on, and damaging something. Then they come back blaming you.

I'm having a 2056 built right now, so no more developement on the stock engine.

Thanks for clarification Rob, it's identical with the options/settings I have available with MS2

I agree with Jeff's premises. And of course I agree with James' as well.

Both viewpoints were part of the internal discussions I went through before Microsquirting last winter. I summarized it in my opening blog post:

https://tgadrivel.blogspot.com/2020/03/on-m...914-part-1.html

D-Jet was ahead of its time...in 1970. That said, there's still actually nothing "wrong" with D-Jet; it does what it needs to do, and does it quite well...as long as it's running correctly. If it is running correclty then there is absolutely no reason to replace it, as any potential efficiencies gained by "modern" fuel injection is unlikely to be recovered in fuel economy (though driveability could be improved).

But when things start to go wrong, D-Jet starts to get expensive. And it's still black magic voodoo to large percentage of the 914 population, which is why many simply chuck it all and throw on carburetors.

Right now we have a big hole in between the two paradigms. Our D-Jet community is in "replace it and see" mode of troubleshooting (my car is doing this, what do I replace?) Getting into an aftermarket replacement EFI system takes skills and understanding, requiring a bolt-on solution that the aftermarket cannot yet provide.

The gap will narrow. I suggest our current knowledgebase on D-Jet will always remain available to those that wish to retian it, and we'll see increasing gravity toward aftermarket EFI solutions. This thread (among related others) is just one tooth in that cog.

Yeah, the issue with supplying a tune to anyone now is the wide variance of hardware being used in builds. Even supplying just the VE table has its issues if everything else hasnt been done perfectly or some hardware variables (such as injector dead time) have been guessed. Im in the same boat as expecting smeones shared tune file to work out of the box at this point is a long shot. With a standardized hardware configuration on a stock motor or at least a motor that was built with the same specs as the one donating the tune, this would be more feasible.

Well said!