Isuzu Trooper 4JX1 Heui Diesel

[JonoD92]

Hi guys
i have a 2001 Holden Monterey (the same as a Isuzu Trooper) with the 4JX1 engine
the 4JX1 engine was one of the first so called common rail diesel engines but has a few things unique to itself
mainly the injection system is a Heui system developed by Catterpillar and licencsed to Isuzu and a few others
it uses high pressure oil to fire the injectors instead of high pressure diesel

The injectors are not Piezo injectors and i believe that they only fire once per cycle not multiple times like most modern diesels
they are driven by solenoids and require around 100v to fire
the amount of fuel injected is decided by the oil rail pressure control valve and it uses a solenoid to open to the right amount
the ecu reads the oil pressure from a sensor and then adjusts the control valve to get the desired presure

i would need to be able to tune rusefi to adjust the control valve based on the oil pressure and throtle postion for example to get the desired fuel delivery after that it is the timing of the injectors.

it is also a turbo but i cant see that making things any harder

i will be setting up a test bench with a oil pump from one of these engines and all the required sensors to flow test the injectors and grade them with the stock ecu as i am currently gathering information and parts to rebuild the injectors as they are the major problem with these engines and if we can rebuild them and get or make new solenoids for them it will save a lot of people lots of money (mainly the solenoids they all tend to die from the heat and age for example i currently have 3 sets of injectors with bad/qustionable solenoids yets internally they are good

anyway by doing this i shoild also be able to do a lot of testing of rusefi on the test bench instead of in the car for a start i will probably be setting up another engine on a test stand aswell so that could be used to test it with a complete engine

Please let me know if you think it can be done im not afraid of doing the work and i am very good mechanacally and also pretty good with electronics both have been hobbies of mine for a long time
im guessing the code will need changing a fair bit to get this to work but i can see if i can do that i have messed with micros a fair bit in the past but by no means am i a professional

Thanks
Jonathon

[AndreyB]

Welcome!

A flow bench would be great, I believe we would need to use something new to us in order to control these injectors, maybe MC33816 http://www.nxp.com/files/training_pdf/FTF/2012/americas/WBNR_FTF12_AUT_F0098.pdf https://octopart.com/search?q=MC33816

[kb1gtt]

Welcome along, and sounds like a cool project.

Have you found the wiki? Just in case you haven't here's a link. http://rusefi.com/wiki/index.php?title=Manual:Hardware_Frankenso_board

I've been dreaming of doing DI, but time and life get in the way. I once drafted up this simulation which might be able to do the push with out special chips. I have posted a QUCS simulation which might be of interest as well. However I also understand that purchased chips now exist, so this circuit might not be worth building these days.


This circuit basically charges a coil, then dumps that coil's energy into the injectors. When it dumps this energy, it spikes the voltage at the injector, then decays to a normal 12V drive.

I believe that's a 4cyl, is that correct?

I think it's possible. I think there are some physical issues to over come. Do you know the inductance of your injectors?

Do you know the analog signal from the pressure sensor? Is it 0-5V, perhaps frequency, or something else?

Also do you know the signal required to drive the oil pump? Is that PWM and 1A like an injector or something different?

[JonoD92]

First of all thank you both for your time and knowlodge

Welcome!

A flow bench would be great, I believe we would need to use something new to us in order to control these injectors, maybe MC33816 http://www.nxp.com/files/training_pdf/FTF/2012/americas/WBNR_FTF12_AUT_F0098.pdf https://octopart.com/search?q=MC33816

Thanks i will have a look at this chip the good thing with all this ofcourse is i dont need it running right now and it will all be tested on the bench and then probably on an engine on a stand because i need the test bench for the injectors now to do a good job at rebuilding and testing them and i might as well make it a complete engine simulation/test bench while im at it

and as for the engine test stand i need to build that anyway for another project that i have planned at some point but no timeframe yet and that is i will be building a BMW M50 engine i have a complete 2l m50 and also a 2.8l M52 engine minus the block as the threads for the head bolts had stripped and i was going to use the iron block from the m50 to replace the alloy block from the m52 but no longer have the car that it was in so now it is a project for the future to build most likely a track car with the m50/m52 engine and turbo it.

Welcome along, and sounds like a cool project.

Have you found the wiki? Just in case you haven't here's a link. http://rusefi.com/wiki/index.php?title=Manual:Hardware_Frankenso_board

I've been dreaming of doing DI, but time and life get in the way. I once drafted up this simulation which might be able to do the push with out special chips. I have posted a QUCS simulation which might be of interest as well. However I also understand that purchased chips now exist, so this circuit might not be worth building these days.


This circuit basically charges a coil, then dumps that coil's energy into the injectors. When it dumps this energy, it spikes the voltage at the injector, then decays to a normal 12V drive.

I believe that's a 4cyl, is that correct?

I think it's possible. I think there are some physical issues to over come. Do you know the inductance of your injectors?

Do you know the analog signal from the pressure sensor? Is it 0-5V, perhaps frequency, or something else?

Also do you know the signal required to drive the oil pump? Is that PWM and 1A like an injector or something different?

Thanks
yes i have seen the wiki lots of great info on it

Thank you for the circuit i will keep it in mind for testing however as you said now that there are chips to do the same thing i might as well start with them depending on cost

Yes it is a 4cly engine

The inductance of the injectors should be about 8mh for a brand new injector but they do tend to die for example i have 1 set reading 4.9mh on all injectors and another 2 sets reading 5.6-5.9mh
from lots of reading it seems anything under 6mh is not really worth using so by using all the information i have gathered from other people on forums and such i have ordered some Catterpillar injector solenoids that i might be able to modify to work with my injectors and we are also looking at 3d printing some replacement solenoids and winding them to replace the stock ones

the oil pressure sensor i do not know the signal from it yet but i can find out easy enough and if need be i can change that to another sensor as i will most likely be relocating it to the main high pressure oil feed pipe instead of under the rocker in the oil rail as they do tend to fail and with it relocated it is more like a 5 minute job to replace instead of an hour or 2

ok so the high pressure oil pump is run off the timing gear on the engine and the pressure to the injectors is controlled with a valve to let more or less oil to the injectors this is controlled with a solenoid on the back of the valve but i dont know what signal is required for this yet but i will be finding that out soon


as for both the control valve and the oil pressure sensor i have a couple of spare ones to play with so that may help and i might need to hook up my scope to the sensor and solenoid to get a better idea of what is needed


Thanks
Jonathon

[JonoD92]

I forgot but i should also mention i have a Tech2 so i can run diagnostics on the engine with the current ecu and i can get data such as what oil pressure the ecu wants for the injectors at different rpms etc
and also i believe it shows what the injector pulse is aswell when looking at live data on it

[JonoD92]

Just one more thing here is a link to a page that has a brief overview of how the HEUI system works it is from the ford powerstroke engines but most of it aplies to the 4JX1 aswell
http://www.trucktrend.com/how-to/expert-advice/1304dp-heui-how-high-pressure-oil-injection-systems-work/

[kb1gtt]

If you consider a chip, perhaps that MC33816, I can offer some help in making a proto board. I would do it in KICAD and then you could either have someone populate it, or if you are good with soldering, you could populate it.

How are you measuring the mH of the injectors? Do you have a piece of test equipment?

Often these DI designs have a high voltage generating circuit, which is then used to generate the push part of the injector

If you generate a 3D model of the solenoid, my daughter (8 years old) has a 3D printer and I could make you a part. We use ABS and PLA plastics.

On that page I see it claims the ICP (oil pressure sensor) is an analog 0-5V signal.

The IPR (oil pressure regulator) might be a bit more complicated. It's probably a PWM signal, and it might desire a dithering frequency added to the PWM to prevent hysteresis issues. I do not know what PWM frequency it might want. Do you have a scope? Can you sniff an existing signal? Is this solenoid low side drive, such that 12V is connected to the solenoid and the PCM is between the solenoid and GND? Also can you measure it's DC ohms with a multi meter?

To me it sounds like the firing the injectors is very much like your spark drive. So that should be fairly easy to tune, just claim they are spark and tune them like spark. You have a dwell, and it varies the timing for when to inject. Then to control the amount of fuel you change the oil pressure, which I don't think is a huge change. Basically you need a control loop for the oil pressure, and the fuel table needs to control the target pressure of that loop instead of commanding the fuel pulse.

[JonoD92]

If you consider a chip, perhaps that MC33816, I can offer some help in making a proto board. I would do it in KICAD and then you could either have someone populate it, or if you are good with soldering, you could populate it.

How are you measuring the mH of the injectors? Do you have a piece of test equipment?

Often these DI designs have a high voltage generating circuit, which is then used to generate the push part of the injector

If you generate a 3D model of the solenoid, my daughter (8 years old) has a 3D printer and I could make you a part. We use ABS and PLA plastics.

On that page I see it claims the ICP (oil pressure sensor) is an analog 0-5V signal.

The IPR (oil pressure regulator) might be a bit more complicated. It's probably a PWM signal, and it might desire a dithering frequency added to the PWM to prevent hysteresis issues. I do not know what PWM frequency it might want. Do you have a scope? Can you sniff an existing signal? Is this solenoid low side drive, such that 12V is connected to the solenoid and the PCM is between the solenoid and GND? Also can you measure it's DC ohms with a multi meter?

To me it sounds like the firing the injectors is very much like your spark drive. So that should be fairly easy to tune, just claim they are spark and tune them like spark. You have a dwell, and it varies the timing for when to inject. Then to control the amount of fuel you change the oil pressure, which I don't think is a huge change. Basically you need a control loop for the oil pressure, and the fuel table needs to control the target pressure of that loop instead of commanding the fuel pulse.

Hey
Thanks for the offer about the board design that would be great that is one area where im not very good i wont have any trouble populating it and could either hand solder it or use my reflow oven

for the inductance of the injectors i am using my Der EE DE-5000 to read the inductance of the solenoid while it is on the injector it has not been calibrated but should be close enough i have never had any problems with the meter

the ICP i missed the bit about 0-5v on that page but that is good if it is

as for the oil pressure regulator i do have a scope but it is only an old 20mhz analog scope i have been looking at getting a new digital one at some point so might have to move that forward as for it being low side i havnt checked yet but will be in the next few days and i can check the inductance of the solenoid and check the ohms aswell

with firing the injectors that is pretty much what i thought for a start anyway it should really be the same as a spark setup in a lot of ways it is a bit crude but should work
the best setup would be having the ability to set trim codes for the injectors so they all stay open for a differant amount of time to get them all as even as possible and also being able to have the ecu adjust the pulse to the injectors over the map so they are not all being fired the same at all rpm ranges etc

as for the 3d printing thanks for the offer but i should have my own setup in the next month for a work project and i was going to try printing the solenoids with nylon because it should hold up to the oil and high tempretures better than abs or pla
(asumming that the replacement ones i have on order cannot be made to fit or are not a close enough match coil wise)

as for the high side for the injector there must be something to generate it so what i might even do is track down an ecu from a wrecker and strip it down to see how they are doing it

[kb1gtt]

I think a 100V supply can be made easy enough. Either that or if you can get one salvage that could work as well.

You have a re-flow as well. I'm thinking a breakout board of this size could be easily soldered by putting a standard toaster on it's side, zip tie the toast ejection lever down, then plug it in until the solder melts. Once it's re-flowed, unplug it and let it cool. Of course this all assumes you have solder paste. If you do not I've had good luck with this stuff, but would recommend purchasing extra needles. It like to dry out and plug the needle.
http://kd5ssj.com/solderpaste/smt-tools-and-process

I believe @russian has one of these china hot air stations, they are also very nice. This one is $42usd to your door step.
https://www.aliexpress.com/item/700W-Hot-Air-Gun-858D-ESD-Soldering-Station-LED-Digital-Desoldering-Station-Upgrade-from-858D/32643771374.html?spm=2114.10010108.1000014.10.nVqmkj&scm=1007.13338.50051.0&pvid=c11c80a1-5a66-42d1-be64-e4cf22e1f540&tpp=1

Wow, you get allot for $100.
https://www.aliexpress.com/item/220V-110V-Saike-852D-Hot-Air-Rework-Station-soldering-station-BGA-De-Soldering-2-in-1/1989604879.html?s=p&spm=2114.30010308.6.7.uPRnIH

I'm sure your dealing with kHz signals, so the 20MHz is more than enough. If you get a scope, I've been happy with the RIGOL DS1054. I also like my DSOquad as it's very small and lots of alternative software out there for it.

[JonoD92]

I think a 100V supply can be made easy enough. Either that or if you can get one salvage that could work as well.

You have a re-flow as well. I'm thinking a breakout board of this size could be easily soldered by putting a standard toaster on it's side, zip tie the toast ejection lever down, then plug it in until the solder melts. Once it's re-flowed, unplug it and let it cool. Of course this all assumes you have solder paste. If you do not I've had good luck with this stuff, but would recommend purchasing extra needles. It like to dry out and plug the needle.
http://kd5ssj.com/solderpaste/smt-tools-and-process

I believe @russian has one of these china hot air stations, they are also very nice. This one is $42usd to your door step.
https://www.aliexpress.com/item/700W-Hot-Air-Gun-858D-ESD-Soldering-Station-LED-Digital-Desoldering-Station-Upgrade-from-858D/32643771374.html?spm=2114.10010108.1000014.10.nVqmkj&scm=1007.13338.50051.0&pvid=c11c80a1-5a66-42d1-be64-e4cf22e1f540&tpp=1

Wow, you get allot for $100.
https://www.aliexpress.com/item/220V-110V-Saike-852D-Hot-Air-Rework-Station-soldering-station-BGA-De-Soldering-2-in-1/1989604879.html?s=p&spm=2114.30010308.6.7.uPRnIH

I'm sure your dealing with kHz signals, so the 20MHz is more than enough. If you get a scope, I've been happy with the RIGOL DS1054. I also like my DSOquad as it's very small and lots of alternative software out there for it.

hey
yes i have a reflow oven (well its an old toaster oven converted to run off a pid)
and yes have used solder paste lots but do need to get some more

and i also already have a 858D hot air station they are cheap for a reason but you cannot beat them for the price

the one with the soldering iron attached is probably not bad for the price but i have found most of those chinese soldering stations dont hold the temp very well and cant react to the tip cooling quicky

i have a few cheap units i get by with but they dont really keep up with what i need for example if im cleaning some pads from a board after removing a bga chip you run the risk of it pulling pads because it cant keep it tempreture
so because i am starting to do a lot more work with electronics for my job (i run my own company with my business partner mainly leasing printers to other small business but i am getting more into the electronics side of things and need to be able to trouble shoot the cirtuit boards better and analyze the signals)so i am looking at getting a new soldering station aswell most likely a jbc station or maybe just the jbc handle and build one of these
http://dangerousprototypes.com/forum/viewtopic.php?f=56&t=7218&p=61175#p61175


and as for the scope im sure my old school 20mhz analog scopes will be ok but with what im doing at work i would like the option of being able to record data and look back over it
the ds1054z is the one i have been looking at getting from what i have seen you cannot beat if for the price and if i need something better i will have to pay over $1000 for it and that could be closer to $2000
but for what i would use it for it is just not needed

[Aussielegend]

Hey guys im from australia i have read most of this post i have been repairing the local 4jx1 population for awhile now and helped many i dont consider myself an expert but i do alright

Wondering did you manage to get anywhere with this? I have an ecu apart at the moment and am looking into what i need to get whwn i pull the chip from the board to see if i can get a map off it.

I will say tho there is a factory port on the board hidden inside the ecu casing.

[kb1gtt]

I haven't heard much about this thread lately. I designed a board found below. At this point in time it has not been assembled or tested. I'm short on time, so it will likely be a while before I get to make much more progress on that board.
https://rusefi.com/forum/viewtopic.php?f=4&t=1337