Home made ECM ? wiring diagram errors.

[Great work!]

 

 

I'm working on a home made ECM for my truck to go along with my VP44 project. I almost found an ECM today on a burned up truck

but the wiring was burned all the way to the ECM so i figured it was toast. The guy wouldn't sell it, I would have to buy the motor and he said the VP44 was bad with hot start issues.  He had a V10 (blown) dually that he was going to put the cummins in. He wanted me to buy both. If he had a manual tranny in the deal I would have took it. But I skipped it. 

 

The usual diagram that many of us use has some errors in the pin names and functions.  From all my VP research I know many of the details

of these pin functions and thought I would share them.

 

Pin 33 Knock sensor is wrong. Instead it is a strange bi-directional signal. its pulled down internally in the VP pin 5.

When left alone it is a current mirror of the injection solenoid current, looks like its chopped or ADC sampled.

When pulled up to 12 volts by the ECM it shuts off the vp. Some kind of safety function and it can monitor the solenoid current blip.  It's called MAB (German). The injection 'Knock' does happen shortly after the blip. I believe the 

injection line delay and combustion delay is sent to the pump in a CAN message for fine tuning the timing calculation. However there is no Knock sensor.

 

Pin 45 is output by the ECM to PCM it turns on ALT, AC, TRANS. It is a 5volt square wave at 2X the RPM. 33 Hz at1000 RPM, 16,6 Hz. It is not sysnced but free runs, but on my VP project it is synced to the crank sensor.

 

Pin 18 is the engine position reference signal. DZG(German) it is the timing ref for the VP so it can calculate its cam donut position relative to the engine crank position. It is a 12V square wave. The trailing edge of the signal always occurs every time a piston is at 10 degrees after top dead center. It is decoded from the CKP and or CMP sensors.

 

Pin 34 is an idle valid signal from the ECM to VP. It is 12V at idle and 0V when pressed. it is  pulled up to 12V in the VP on pin 4.

 

The APPS has 2 idle outputs that are open collector that are always in opposite states.  VP pin 2 is 12v =idle,  0V = pressed. APPS pin 6 is inverted.

 

The ECM monitors the idle signals and pot and if its happy. the ECM pin 34 changes state it is sensed at VP pin 4 LGS (German)

the VP then changes a byte in the CAN message going to the ECM as a cross check  then The ECM increases the fuel command byte.  If the pump is Hot wired, grounding pin 4 makes the engine run at a higher RPM.

 

CAN message from VP :  22 40 8 04 00 85 23 xx  yy 79 13.

The Byte in red  is:

0C HEX = not running

04 HEX = idle valid state (as detected via ECM pin 34 to VP pin 4)

00 HEX = above idle operation      "" 

yy HEX is the pump RPM MSB

xx HEX is the pump RPM LSB

79 13 is the fuel temp in Kelvin encoded ?

 

 

Edited December 31, 2023 by Great work!
clarity

[DanTheMan]

Perhaps you'd know what my issue is, first of all, AWESOME write up of the VP44 pinouts. I've made several ECU's for small engines before, it's definitely challenging. Have you come up with a good control system for it yet?

Secondly.. My issue. I just bought my first 24v cummins, its in an 01 dodge ram 3500 dually, with the nv4500 transmission. I am having high idle issues, where the truck idles fine on startup, then idles up to 1000 (to be expected) but then continues to climb as it gets hotter, to around 1300-1400 RPM.

 

I have replaced the Coolant temp sensor, the TPS (And calibrated), the MAP, and the IAT sensors, all without any luck..

 

I upgraded and repaired all grounding issues, new terminals, new batteries, new alternator.

 

I checked all harnesses, made sure the 5v reference signal was clean, going to the TPS (APPS), and other sensors... 

 

I have literally checked everything, maybe you can give me a glimmer of hope. I wouldn't imagine its a mechanical issue with the VP44..

[Great work!]

The ECU is loosing it scratch pad EEPROM

So that's where the tps calibration is. You can't fix without a special tool and software. I put a 1k pot on my tps signal signal to fool it and idle it down but your ecu will die soon. My VP44 project doesn't need ecu just CKP ,tps and map.

[DanTheMan]

So even though the voltage is properly calibrated, I’m assuming the resistance is what the ECU listens for to idle down? I could try to throw a potentiometer on there and see if it idles down. 

 

If you don’t mind sharing, how did you go about this project utilizing those 3 sensors? I am fairly versed in Arduino and the like; if the ECU is indeed shot, that would be a neat project. 
 

I’ve also got several devices for reading EEPROMs, do you have any information on reflashing the memory? I’m assuming that’s what you meant by the ECU is losing its EEPROM.

[Great work!]

The ecu scales the fuel from tps resting to wot those values are stored. If the get corrupt then the idle changes. You could try to read the EEPROM repeatedly with arduino and get a good read then copy to a new one and solder it in. Very tricky.

 

I'm working on a githup repository but I'm old dog and don't know what I'm doing. Lol

 

If you have delayed WTS light then the the ecu processor is going out too.

[DanTheMan]

I will be doing some more research! That’s very interesting information, I’ll look into extracting the data off the EEPROM. Once you get your GITHUB setup I’d love to take a look!