Mercedes-Benz Forum banner

1 - 5 of 5 Posts

·
Registered
'72 450SL, 107.044-12-000422
Joined
·
190 Posts
Discussion Starter #1
On a warm engine, with the air temp sensor disconnected, I can adjust idle mixture to 13.5 afr (4 clicks from the rich limit). When I connect the sensor, and ECU pot turned to the range limit, mixture goes ultra lean (off scale on my wideband gauge, so above 18 afr).

I replaced the sensor with a new one, which reads 300 ohms, so seems to be correct (75 degrees OAT). I also cleaned the ground points.

Any idea what the ECU is doing internally with the sensor input?

My next thought is to open the ECU and have a look but not really sure what to look for.


Sent from my iPhone using Tapatalk
 

·
Registered
1973 450 SL
Joined
·
1,748 Posts
masand68,

Have a look at Brad Anders' work on Rennlist. Brad and some other Porsche 914 enthusiasts dove pretty deep into the science of Bosch D-Jetronic. An engineer named Joe Kerfoot produced a wiring schematic of the Porsche ECU and was able to analyze the different functional areas within. 914's used a 4 cylinder engine with D-Jet. The throttle switch on VW systems were simpler too.

Reading is very technical but helped me understand what is really going on in the ECU and how important the signals from the sensors and to the injectors are to the health of the system.

https://members.rennlist.com/pbanders/ecu.htm
 

·
Registered
'72 450SL, 107.044-12-000422
Joined
·
190 Posts
Discussion Starter #3
Thanks Brad, that’s a good read.

According to the article, the ecu is meant to enriched the mixture when the temp sensor circuit is open. So that at least tracks with what I see going on and leads me to believe ecu is functioning as designed.

So my troubleshooting continues. I’ve verified fuel pressure at 30 psi and did a cursory check for under hood vacuum leaks with carb cleaner-no obvious leaks. Is there a better method?

Anyone have any other ides on what to check? I’m thinking fuel flow capacity check on the overall system, then check individual injector flow.


Sent from my iPhone using Tapatalk
 

·
R/C107 Moderator
1986 560SL: '84 500SL: '84 280SL 5 speed: other 107s
Joined
·
31,708 Posts
Smoke test is the beast method I have seen for vacuum leaks. There are several home made smoke " machine" you tube vids. Search this forum for one with soldering iron, rags, baby oil, and pickle jar it was the best I found..
 

·
Registered
1973 450 SL
Joined
·
1,748 Posts
masand68,

What I've learned about Bosch D-Jetronic from Anders and the VW guys at Samba.com . . .

1. Don't tamper with the MPS unless everything else is in spec. When the time comes to do that, the correction involves a 1/32 turn counterclockwise to compensate for vacuum loss from a worn turning assembly (lower compression = lower vacuum too). Every single MPS I've found in salvage yards that has been tampered with has been adjusted by 1 or more full turns in the wrong direction (tighter is better mentality is very wrong in this case). The good news: I've found a decent adjustment target to bring these back into spec.

2. The weak link in Bosch D-Jetronic is the architecture of the engine harness. Approaches to solve this vary from sourcing a new harness to gold plating the terminals (Samba guys). Most lean toward originality on this because that approach is valid on so many other areas of the Mercedes rebuild world. This is a Bosch contracted system. Bosch contracted Kroschu to build the engine harness. Kroschu contracted AMP: (colon part of the trademark at the time - now AMP/Tyco) to build the terminals and connectors. This contracting train defeats the logic of adherence to MB originality. The product of this train is woefully inadequate on a V-8 stuffed into a small engine bay. Catalytic converters made this worse starting late 1974. The good news, rebuilding the engine harness to improve connectivity and integrity has been demonstrated. Until this is done, the ECU is a blind man reading engine signals by Braille.

I have a very large collection of M117 D-Jet critical spares gleaned from salvage cars. Many of these parts are much newer than the cars they were harvested from. This reveals the money pit that led these cars into my hands at the salvage yard. I like to test these parts on the bench and in operation. Old or new, when these sensors, ECUs, and MPSs are tested on an engine in spec with a good harness and a solid ignition system, they perform flawlessly. Lesson learned . . . it was never the sensors, ECU or MPS. It was in another weak area - see above.

The other huge reliability issue on M117/M116 lies in the distributor. Bearings wear due to low tech oils available in the day, leading to a micron level wobble which is enough to destroy contact points in short order. Pertronix Hall effect ignitors are the solution for this problem. I would do this mod before any other on Bosch D-Jetronic M117/M116/M110. Eliminate mechanical contact in favor of magnetic sensing. When you do this, you also eliminate the infamous green signalling wire that has been a subject of much ire and frustration in this band of enthusiasts.

When and engine runs poorly, investigate spark, air, fuel in that order. Spark on a Bosch D-Jetronic system is ignition and ECU signalling. Air in any system is intake and vacuum. Fuel is clean flow from the source, through a pressurization system, to delivery at combustion and finally to return and evaporation controls. I've learned the hard way not to dive into a single component, out of order, talking myself into a solution, throwing money at that solution and ultimately failing.

Enough big picture thoughts from my couch after a very nice bottle of Napa's finest (made by my son BTW). Please feel free to PM me about these topics. I have loaned many parts to D-Jet owners in the past - will gladly continue to assist in your diagnosis.
 
1 - 5 of 5 Posts
Top