190e with 5.0 M113
Posted: Sun Jan 16, 2022 10:15 am
I have an M113 50 that will be finding its way into a 190e eventually there are two projects in my line of fire before this one gets done, but I am looking to get the engine running by itself sooner than later. The engine is from a 2002 S500, which is useful to note as it is a facelift engine. The engine was cheap, and also came with the 722.6 which will be reused on another project. This one will receive the 6spd manual from an m112. The disadvantage of the cheapness is that I never so much as saw the engine running or the car it came from and it is covered in oily grease making inspection of what it does tricky. Apparently it was just too rusty to MOT again, as a W210 owner I can believe this.
Pinning a link to the shared album I'll use to keep tabs on stuff here
I’ve come to rusEFI after doing initial research on Speeduino because:
1) I wasn’t sure how well the waste spark configuration needed on speedy would work with the twin spark coil packs,
2) because rusEFI supports electronic throttle (I like the magnesium intake manifold and the airbox integrated in the vanity cover on the m113 so keeping this is a plus), and
3) because I can see the work being done in the area of getting a standardised 128 pin ECU for the engine
This is my first standalone project, and will be my first tuning project. I’ve got a fair bit of knowledge to catch up on, most is generalist stuff that I am just working on consuming as fast as I can, while there is certain domain specific information I’m chasing.
Ignition
Thanks to the hardware_mercedes Slack channel this morning, we've pretty much confirmed that the factory coil packs are twin dumb coils per cylinder, one per spark plug -> 221 503 035 # bosch
Q1. I’ve removed an ignition pack from the engine, on the reverse it looks like each HT lead has its own coil pack, and the three pins on the input connector suggests that each is individually powered. If this is the case, can these be wired parallel from a single IGBT or would it be better to have two IGBT per pack? One seems better to me, in that I don’t think we want half spark if there is an issue, i think I’d prefer to have the whole cylinder stop firing?
Q2. I'd like to be able to break the processes down into smaller chunks, so for the firing of the ignition coils, I have an idea that I will supply pulse width modulated signals to an ignition coil sat on the bench/kitchen table to simulate this IGBT arrangement using a DC power pack, some method of formulating PWM current (not sure on this one yet) - simulate the ignition output for a pin on proteus, an IGBT (useful for testing sizing, as this is all new to me), leading to the coil pack to observe sparking of plugs. Are there any practicality constraints with doing this?
The ECU
The wiring for a prefacelift m113 43 can be found here
Sharing some links to my own pictures,
Factory ECU pin outs
It has been suggested on the slack channel that I make up a breakout adapter from Proteus to my ECU connectors -> this will leave the factory engine wiring harness, which looks reasonably complete in place with their factory connectors, I think future me will appreciate the use of the factory wiring colours and positions for debugging. Some leads on this have been suggested; firsty that similar adapters have been done in the past for Honda engines (I will get searching), and that the M54 BMW plug is similar to the M120 plug on an S600 at least. I recently attended an automotive wiring seminar online so I feel semi confident I can make one of these up
The workbench
As I sort of alluded to at the start, I am working on some other projects in tandem. The common theme to all is the concept of single board computers. I'm building up a list of tools I think I will need, staring the ones I don't already have - please let me know if you think I am missing anything
* Soldering station with hot air for desoldering* # have an old soldering iron that I will replace for one with variable temp
* DC Power Supply (you know the ones with twiddly knobs in science labs)* # currently been using jump packs and ctek to delivery 12V to test stuff, to varying degrees of disaster
* ratcheting crimp tool
* Oscilloscope, the Rigol DS1054z being well regarded for its value and hackability* (if only some would come up second hand on ebay!)
* maybe some arduinos, bread boards and an array of components I can use for simulating inputs such as CPS, IAT etc -> the idea being that I have some canned data I can play to the rusEFI so that I can test it before putting it on my engine*
Q3. does bench testing work? is it ever representative of what is observed when plugging into engine?
Pinning a link to the shared album I'll use to keep tabs on stuff here
I’ve come to rusEFI after doing initial research on Speeduino because:
1) I wasn’t sure how well the waste spark configuration needed on speedy would work with the twin spark coil packs,
2) because rusEFI supports electronic throttle (I like the magnesium intake manifold and the airbox integrated in the vanity cover on the m113 so keeping this is a plus), and
3) because I can see the work being done in the area of getting a standardised 128 pin ECU for the engine
This is my first standalone project, and will be my first tuning project. I’ve got a fair bit of knowledge to catch up on, most is generalist stuff that I am just working on consuming as fast as I can, while there is certain domain specific information I’m chasing.
Ignition
Thanks to the hardware_mercedes Slack channel this morning, we've pretty much confirmed that the factory coil packs are twin dumb coils per cylinder, one per spark plug -> 221 503 035 # bosch
Q1. I’ve removed an ignition pack from the engine, on the reverse it looks like each HT lead has its own coil pack, and the three pins on the input connector suggests that each is individually powered. If this is the case, can these be wired parallel from a single IGBT or would it be better to have two IGBT per pack? One seems better to me, in that I don’t think we want half spark if there is an issue, i think I’d prefer to have the whole cylinder stop firing?
Q2. I'd like to be able to break the processes down into smaller chunks, so for the firing of the ignition coils, I have an idea that I will supply pulse width modulated signals to an ignition coil sat on the bench/kitchen table to simulate this IGBT arrangement using a DC power pack, some method of formulating PWM current (not sure on this one yet) - simulate the ignition output for a pin on proteus, an IGBT (useful for testing sizing, as this is all new to me), leading to the coil pack to observe sparking of plugs. Are there any practicality constraints with doing this?
The ECU
The wiring for a prefacelift m113 43 can be found here
Sharing some links to my own pictures,
Factory ECU pin outs
It has been suggested on the slack channel that I make up a breakout adapter from Proteus to my ECU connectors -> this will leave the factory engine wiring harness, which looks reasonably complete in place with their factory connectors, I think future me will appreciate the use of the factory wiring colours and positions for debugging. Some leads on this have been suggested; firsty that similar adapters have been done in the past for Honda engines (I will get searching), and that the M54 BMW plug is similar to the M120 plug on an S600 at least. I recently attended an automotive wiring seminar online so I feel semi confident I can make one of these up
The workbench
As I sort of alluded to at the start, I am working on some other projects in tandem. The common theme to all is the concept of single board computers. I'm building up a list of tools I think I will need, staring the ones I don't already have - please let me know if you think I am missing anything
* Soldering station with hot air for desoldering* # have an old soldering iron that I will replace for one with variable temp
* DC Power Supply (you know the ones with twiddly knobs in science labs)* # currently been using jump packs and ctek to delivery 12V to test stuff, to varying degrees of disaster
* ratcheting crimp tool
* Oscilloscope, the Rigol DS1054z being well regarded for its value and hackability* (if only some would come up second hand on ebay!)
* maybe some arduinos, bread boards and an array of components I can use for simulating inputs such as CPS, IAT etc -> the idea being that I have some canned data I can play to the rusEFI so that I can test it before putting it on my engine*
Q3. does bench testing work? is it ever representative of what is observed when plugging into engine?