Using LBZ injectors and pcm on detroit diesel

[boothybunch]

Got a good question for you guys.

Someone i chat to on another forum is planning to try to make a detroit diesel 4-53 mechanical injection into a common rail system using LBZ duramax injectors and PCM.

Just discussing if you can disable injectors using EFI or would it be possible to fire 1 injector twice by wiring the injector ground to the PCM twice? Was thinking of putting a diode on the grounds so it will not backfeed the PCM.

lets hear your thoughts

[LB72004]

Sounds like an interesting project.

With the scan tool you can temporary disable one injector at a time but you are probably lookin for a more permanent one right? I have diabled my LB7 injectors before by replacing them with a small inductor. Had to try a lot of inductors to find one that wouldn't put it into limp mode.

I don't think you can just use a diode to keep the signals separate as it will likely also block the ecm from monitoring the injectors too.

What about using something like what VW uses that is already 4 cylinder? Or since it is a 2 stroke with a custom cylinder head (which he will likely need anyways) maybe two injectors per a cylinder will work. If the timing is right it would just alternate between the two injectors for each power stroke.

Sounds like a lot of work to me but would be awesome if it worked

[schwoch1]

I am the guy that boothybunch speak of!!! I am crazy enough to try this, and hopefully I can make this all come together!!! Basically the plan at the moment is to tie the injector grounds from two opposite cylinders to one injector and hopefully I do not piss of the PCM too badly. The injector are already powered in pairs, so power should not be a huge issue. My goal with this project is to try and accomplish all this with off the shelf parts. After talking with some local electronics guru's here, worst case scenario is I end up with some sort of driver module to fire the injectors and keep the PCM happy and out of limp mode.

I plan on using the stock Detroit head, there should be no reason a custom unit would need to be made. Custom injector sleeves will need to be fabbed up (think like LB7 injector sleeves but bigger!) and some minor clearancing will need to be done on the exhaust rockers to get the injector to clear an acceptable amount! Biggest downfall at this time is there will be no provisions for a Jacobs Engine brake as the injector rocker arm will not be used anymore :yucky:

I looked into other common rail systems out there, but a few problems arose. Using a 4 cylinder setup meant doubling the engine speed to the PCM to get the injectors to fire at the correct times to satisfy the needs to a two cycle engine. Secondly, and most importantly, it came down to being able to tune the thing. I am not expecting this thing just to start right up and run perfectly, I am expecting a ton of tuning ahead of me!!! I have used EFI Live for years and I am comfortable with it and seems to be the most thourough tuning I can find for the Duramax! The Duramax and the Detroit are roughly the same displacement as each other per cylinder, so at least that is a step in the right direction as far as tuning is concerned, but I am sure that is all they will have in common!

My next step is getting some PCM connectors from Mike@ eficonnection.com so I can power up my newly aquired PCM and see what it can do!!!

I am sure I will have some questions along the way that hopefully ya all can help me with when I get lost! I will post any progress I make on the general diesel forum as it really doesn't fall into the LBZ specific page! Thanks for showing some interest!!!

Mike

[LB72004]

I will definitely be following this one. Sounds like a lot of fun.

I still think if you can figure out how to get two injectors into each cylinder that would be awesome but your approach if it works would be the quickest rout to a running engine.

I have always thought that it would be neet to rig up something like this with a large Cummins or CAT motor, v8 of course. Would probably need larger injector nozzles and maybe twin cp3s but just think of the power you can make. Your engine should make crazy power if you can get it running right

Good luck

[THEFERMANATOR]

The injectors aren't powered in pairs, but in banks. As I understand it the ECM has 2 drivers that power the injectors in 2 banks of 4. Essentially it uses a tick tock pattern to allow one driver to fire while the other is charging. You could use 1 group of 4 to run it, but not sure how happy the ECM would be with 1 driver set dead. Also it should be of mention that if the ECM detects a problem or backfeed of any kind on an injector driver, it shuts it down. SO i don't think your plan of tieing to injector leads together will fly very well. One to talk to about this that would be able to give you a much better description would be Ben "duratothemax", he understands the electrical side and can describe it ALOT better than I could ever begin to even try to. He has also done some crazy one off swaps for others. And also like you pointed out, keep in mind you will need an ECM capapble of controlling twice the RPM's you will actually be turning since you will be firing an injector on each up stroke instead of every other one. The 2 injectors per cylinder method would take care of this, but I don't see how you would ever fit 2 to a cylinder. You might want to consider something along the lines of what teh diesel cruze uses for a control system and see if it can be used since it is a 4 cylinder to begin with.

[LB72004]

The injectors aren't powered in pairs, but in banks. As I understand it the ECM has 2 drivers that power the injectors in 2 banks of 4.

yep, cylinders 2, 4, 6, 8 or on one bank and 1, 3, 5, 7 on another. if the ECM detects any issues, too low or high resistance/inductance, short to ground and possibly injector to injector shorts it will shut down the whole bank


You might want to consider something along the lines of what teh diesel cruze uses for a control system and see if it can be used since it is a 4 cylinder to begin with.

now that is a good idea. how far out is the cruze from full tunability? is it just a matter of getting the files like the LML? i also wonder if the ECM could drive the LBZ injectors, actually i think it could drive the LML injectors as the cruze and LML are both piezo injectors right?


hay Schwoch1

if you havent already seen this site, here is a wealth of info on the Detroit 53 series engines

http://www.4btswaps.com/forum/showthread.php?10876-Common-Detroit-53-Series-questions-answers-and-info-Check-here-first!

[schwoch1]

Here is the wiring diagram that for the moment I am basing my assumptions on. I could be wrong but it looks as if the PCM provides 4 power leads that are grouped with two injectors. Curret thought at the moment is to tie them to the four injectors that I have and figure out what I may need to do with the ground side of things. Like I said, I am in the design/dreaming phase of this one at the moment. This could be a total bust for all I know at the moment. I have a PCM that I will be wiring up in the next few weeks and we will see what happens than!

17680

I thought of using the Cruze system also, but I would have to double the RPM that the system see's, which wouldnt be a huge deal but one problem I see at the moment is there is no way to disable EGR, DPF, SCR UREA etc for use of this system. I am told that the higher ups at EFI Live have no intentions of making them deletes avaliable to the motoring public, so hence that one is a dead duck for this project! WOuld make life a bit easier if the injectors would cooperate I guess!

I am apprehensive about approaching anyone about my ideas, as I have been called everything except late for dinner when mentioning this project to some people. Unfortunately some people cannot see outside their little box they live in and look around. I am glad that I am getting some ideas without someone telling me to "just use the whole Duramax" or"just stuff a Cummins in it and be done" or "leave the Detroit all mechanical so when we have the next zombie attack and they use an EMP, my truck will still run ". Again, thanks for the ideas! Keep any ideas coming!!!

Mike

[LB72004]

just checked my schematics for the LBZ ECM and that is the same wiring diagram i have. i dont see any way to use diodes to block back feeding on the low side drive. i would just join them together and cross your fingers it works and doesn't put it into limp mode

how many RPMs do you want to redline the engine at? if you only use four of the injectors and trick the ECM with resistors for the others then you will be limited to around 2k, maybe 2400 RPM. if joining the lines up works then you can probably spin it fast enough to launch pistons :cucumber: of course it would be easy to limit this in the tuning to whatever you wanted.

will the LBZ injectors fit between the valve strings? i still like the idea of having two injectors in each cylinder. fill in the original injector hole. bore out two new ones as close as possible and mount the injectors back to back. the LBZ injectors seem like they might be tall and narrow enough that it just might fit. having a double fired common rail 2-stroke detroit would be the baddest motor around

[schwoch1]

The 4-53 in stock form turns 2800 or so RPM's under load, a little bit higher under no load. I thought about just doubling the RPM on the LBZ system, but if I remember correctly, the tables stop at 4800 RPM which ain't quite enough to make it to my governed RPM!

I have test fit the LBZ injector in the Detroit. It will fit between the valve springs no problem, just getting it in between the rockers and bridges is a little tight!!! It clears.........barely!!!I think I will need to do some creative shimming or a little clearancing on the rocker bridges to make sure there is enough room to keep everything from kissing!

My goal is to get this all to work with no major machining to make this happen, so at this point I am going to rule out the extra injector, although that would be hella cool!!!. I am going to try and just get the LBZ injector to double fire, as it would be the easiest solution!

I like the thought process, keep it coming!!!

Mike

[LB72004]

in the duramax the injector only has to fire once every two revolutions not counting pilots. so at 4000 RPM it is only firing 2k time a minute.

with the detroit at 2800 RPM the injector is going to have to fire 2800 times not including the pilot injections. can it do that? how quick is the response time of the LBZ injector when doing a main shot?

[LB72004]

Schwoch1,

is this what you are thinking for the wiring?17687

[schwoch1]

Yes, that is EXACTLY what I am thinking of doing, not sure if it will work though at this time!!!

I need to procure a few more injectors and a 2:1 gearbox to get the crank and cam signals to the PCM to do some accurate testing on the bench.

You got me thinking on the injector response time, I would imagine that there should be enough time to fire the injectors, may have to forgo pilot pulses after a certain RPM though was my thinking. I am positive that the noise of the engine would mask any combustion noise that would be present past 1800 RPM!!! For such a small engine, it sure is noisy!!!

Mike

[LB72004]

I need to procure a few more injectors and a 2:1 gearbox to get the crank and cam signals to the PCM to do some accurate testing on the bench

wouldn't a reluctor wheel(s) with the correct the number of teeth be easier and more reliable?

[schwoch1]

I have to have some way of replicating the cam sensor signal on the bench, which is 2:1 of crankshaft speed, hence the need for a ratio adapter of some sort!

On the actual engine, I have the same problem also as nothing on a Detroit turns 1/2 speed of the crank. I will cross that hurdle when that day comes. If everything else works as planned mounting a cam reluctor should be a cake walk!!!!

Mike

[LB72004]

ok, i got you. just bench testing.

so, what do you plan on doing with this engine if it all works out? putting it in anything or is it going to just be a floor model/proof of concept type deal?

[schwoch1]

I plan on putting it in some sort of older truck and use it to haul my 6000 lb travel trailer where ever I need to go with it!

I got my eye on a '69 GMC 3/4 ton 4wd truck that someone took a Suburban and a truck and made a crew cab truck. Fab work is excellent and this would be the perfect truck for this swap. For now if the deal works out, I am going to put my 4-53T that I have in it, hook it to an Allison 1000 and drive it till I get the common rail conversion figured out!

I am tired of driving what I call 'me too' trucks anymore! Me too refers to that I can go and sign my life away for 6 years on a truck and have something that looks the same as 98% of the rest of the trucks on the road! I want a functional conversation piece I guess is what I am trying to say!!!

Mike

[THEFERMANATOR]

The RPM issue can be overcome with a CAX file as I know a few are turning close to 6000 with LBZ's in racing applications retaining the BOSCH ECM. The problem would be in the fact the injector may not be happy living at those kinds of duty cycles to the injector coils for those lengths of time. I'm not trying to be a buzzkill as this sounds like an innovative project, but I don't want to see you go through all this trouble and run into issues down the road.

[schwoch1]

The engine will get to 3000 very little, most of this time it should be in the 1900-2300 RPM range, so the duty cycle of the injector shouldn't be too bad, as it won't be spending too much time at the 6000/3000RPM mark. I plan on bench testing this all before I even attempt to adapt it to the engine to save myself alot of time and trouble. Worst case scenario is I have a $750 pile of LBZ parts that I will try and liquidate on ebay!!!


I'm not trying to be a buzzkill as this sounds like an innovative project, but I don't want to see you go through all this trouble and run into issues down the road.

THEFERMANATOR, dont feel you are a buzzkill, it is nice to hear others opinions and bring up problems that I may not thought of. Chime away if you feel the need!!!!

Mike

[THEFERMANATOR]

The engine will get to 3000 very little, most of this time it should be in the 1900-2300 RPM range, so the duty cycle of the injector shouldn't be too bad, as it won't be spending too much time at the 6000/3000RPM mark. I plan on bench testing this all before I even attempt to adapt it to the engine to save myself alot of time and trouble. Worst case scenario is I have a $750 pile of LBZ parts that I will try and liquidate on ebay!!!



THEFERMANATOR, dont feel you are a buzzkill, it is nice to hear others opinions and bring up problems that I may not thought of. Chime away if you feel the need!!!!

Mike

My concern is your dealing with an injector designed to be fired at a max of 3250 RPM's from the factory, and on average didn't spend any real time above 2500 to speak of. For you this would only be 1250 peaking out at 1625 due to the doubled firing from 4 to 2 stroke.

[LB72004]

The RPM issue can be overcome with a CAX file as I know a few are turning close to 6000 with LBZ's in racing applications retaining the BOSCH ECM

Holy H#ll, how do i get one for my LMM. i like the sound of that :banana:

[LB72004]

how hard would it be to make your own ECM that only uses four injectors out of something like an arduino module? seems like the wiring and sensors, etc would be easy enough, it is the programming that seems like the biggest hurdle.

you would only need one crank or cam sensor but can wire both for redundancy if desired.

Use a PWM circuit and some SCRs fired by the arduino module to run the injectors. with the PWM you would be able to control the saturation point on the injector coils and turn them on and off with cheap high powered SCRs. minimal heat with maximum efficiency that could be build into a small package and programmed on your computer.

[DURAtotheMAX]

how hard would it be to make your own ECM that only uses four injectors out of something like an arduino module? seems like the wiring and sensors, etc would be easy enough, it is the programming that seems like the biggest hurdle.

you would only need one crank or cam sensor but can wire both for redundancy if desired.

Use a PWM circuit and some SCRs fired by the arduino module to run the injectors. with the PWM you would be able to control the saturation point on the injector coils and turn them on and off with cheap high powered SCRs. minimal heat with maximum efficiency that could be build into a small package and programmed on your computer.

IMO thats not realistic at all...

Think about it, you would literally be building a complete ECM and tuning from scratch, using an extremely basic hobbyist 8-bit microcontroller. In other words, you would be recreating/building something that GM has spent millions of dollars on, hundreds of thousands of man-hours on, and run using modern 32-bit RISC processors......

The timing and closed-loop monitoring control of getting a single-cylinder common-rail diesel engine to even at run/idle (let alone high RPM, different power/load settings, multiple cylinders, etc) with an arduino would be a pretty amazing feat in itself.

[LB72004]

well, here is a few examples of people that have actually done it

https://www.youtube.com/watch?v=r_Xhwcqrcy8

http://blog.atmel.com/2014/02/25/arduino-mega-drives-custom-ecu-for-diesel-engines/

http://www.reddit.com/r/arduino/comments/1r12bz/arduino_based_engine_management_fuel_and_ignition/

obviously the programming would need to be tweaked but it is not a insurmountable million dollar life long project. of course it is not the perfect solution it was just another thought to keep the ideas going.

[DURAtotheMAX]

well, here is a few examples of people that have actually done it

https://www.youtube.com/watch?v=r_Xhwcqrcy8

http://blog.atmel.com/2014/02/25/arduino-mega-drives-custom-ecu-for-diesel-engines/

http://www.reddit.com/r/arduino/comments/1r12bz/arduino_based_engine_management_fuel_and_ignition/

obviously the programming would need to be tweaked but it is not a insurmountable million dollar life long project. of course it is not the perfect solution it was just another thought to keep the ideas going.

Yes, I've seen those before...but none of them are common rail Diesel engines. And none of them are multiple cylinder engines (the tdi doesn't count, because the VP37 pump does all of the hard work, the Arduino is just giving it a throttle signal).

The power supply alone for a 4 cylinder common rail diesel would take some thinking. The lbz/LMM ECM has a fairly complex power supply in that it actually uses the flyback from the closing injector solenoids to charge the cap bank for the next combustion event, and lots of other trickery as well. At idle when the injectors only have about 400uS of on time, that requires a fair amount of precision/timing and processor speed to handle.

jmo.

Ben

[schwoch1]

It seems the more I look into this, the less feasible this is becoming.... :(
Any systems out there that may work better possibly that have the tuning capability that the Duramax's have?

Just curious!
Mike