Author Topic: Automotive ECM comparison  (Read 5225 times)

0 Members and 1 Guest are viewing this topic.

Offline spencerlinnTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Automotive ECM comparison
« on: February 10, 2015, 01:20:22 am »
Hi everyone! I'm new here so I apologize in advance if this is the wrong section, but basically I just wanted to get you opinion on some automotive stuff. I've done a few projects with my cars and with friends cars, and that's what's really fueled my interest in electronics. I'll explain the background behind this post if anyone wants, but basically I want your opinion on the quality of construction and thoughtfulness of design that went into two different Engine Control Modules. They're both from similar vintage cars, around the mid 90s I believe, but the construction inside is totally different - one of them (exhibit A) is almost entirely SMD, the other (exhibit B) is a mix with through hole components. My basic beginner knowledge of this stuff tells me that exhibit A certainly seems much more quality for the mid 90s, but I'm also thinking that not only is exhibit B rather old fashioned, but they appear to ignore the fact that a car is a harsh environment for what looks like SIP hybrids and other rather flimsy-mounted parts.

Again, I'm just looking to get your opinion. I enjoy seeing stuff like this where two different devices were devised presumably around the same time, to do the same job, but the construction is totally different!

 

Offline luky315

  • Regular Contributor
  • *
  • Posts: 244
  • Country: at
Re: Automotive ECM comparison
« Reply #1 on: February 10, 2015, 08:32:15 am »
I am interested in the digital and analog input circuit. Do you have more inforations on this?
Personally I don't like the construction of the "B" ECU with its unsupported THT devices, but I'm sure they tested it properly and it works fine....
"A" is a Bosch ECU with a nice thermal design (using the case as heatsink).
Is there any conformal coating on the units?
 

Offline spencerlinnTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Re: Automotive ECM comparison
« Reply #2 on: February 11, 2015, 03:38:46 am »
OK guys, appreciate your input. I realised that there are some differences in the ECMs that my car could have used. I believe Exhibit A in my original post was the one used for the V8 E39 models, it was called a "Bosch ME 7.2" DME (Digital Motor Electronics, which the Germans use as a synonym for ECU/ECM etc), and I don't have a spare laying around (since they're coded and generally not swappable for simple diagnostics like most Japanese cars!) but I did find another BMW ECM laying in my junk box.. It's from, I believe, the inline 6 E39 models, and it's… surprise.. not even a Bosch?! It's a Siemens DME "MS 41.1" which I suppose is the type/version. It's got quite the thermal management going on on the DPS926748 ICs, which look like audio amps? lol there are 6 of them held to the chunky aluminum chassis of the case (which both sides comes off of using little E-torx bolts.. dang Germans!) and on the other side, six smaller STP8055's in TO220 packages also held to the chassis. These appear to be ignition related, and since BMW loves individual coil-on-plug, it makes sense they use 6 chips to make those coils fire, rather than one if it were a distributor, or a few if it were wasted spark. So I'm guessing the DPS926748's are fuel injection related, but couldn't come up with a datasheet on a quick search. Either way, it's curious that not only do the V8 and I6 motors use different types of DMEs, but from different manufacturers entirely?! (I also noticed the ST… what appears to be a micro perhaps, made in.. Malta?? OK!)
 

Offline RJFreeman

  • Regular Contributor
  • *
  • Posts: 207
  • Country: au
    • Australian Technical Production Services
Re: Automotive ECM comparison
« Reply #3 on: February 11, 2015, 03:54:45 am »
I do remember hearing (although had no direct knowledge, so tale is anecdotal only) that one of the reliability issues with the VN Commodore (VN stood for Very Nasty) and the Holden Camira was that the ECM for them used a multi layer board with very fine traces and over time thermal expansion and contraction was enough to cause failures in the PCB.
 

Offline ion54

  • Contributor
  • Posts: 34
  • Country: ca
Re: Automotive ECM comparison
« Reply #4 on: February 12, 2015, 08:10:45 pm »
Can I make a guess? The Exhibit B is a Japanese ECU.
As a matter of fact is OK what they've done in regards with the through hole components. The ECU location is not exposed to hard mechanical shocks because it's not attached on a door or wheel but on body, which is sort of isolated by suspension.
The thing you liked the most, the heat management of the Bosch/Siemens ECUs, shows in fact lack of efficiency. Newer  designs (shall I say better?) do not need that much heat sink. That is money you spend extra when buying BMW, and they are not too reliable either.

 

Offline wagon

  • Frequent Contributor
  • **
  • Posts: 289
  • Country: au
Re: Automotive ECM comparison
« Reply #5 on: February 12, 2015, 08:25:14 pm »
You won't find much information on anything used in automotive ecu's.  I've fixed hundreds (thousands?) over the years, and just lucky most of the problems are caused by capacitors leaking their guts.  The second picture in the first post looks like an ecu out of a Mazda, probably a 4 cylinder.  The Germans tend to run ignition coils direct from the ecu, so they quite often have big heatsinks and stuff.  The Japanese (back then, at least) run an external igniter for the ignition coil.  All run the injectors direct. 

The problem with VN Commodore, etc. units was not bad design or pcb quality : it was actually really, really piss-poor soldering.  Most of the reco 'Delco' units are just stripped of their conformal coating, resoldered, tested and recoated.

I've got a Cummins N14 'Celect' engine ecu in to fix at the moment (mid 1990's), and also a transmission controller for a 1984 Steiger tractor.   The tractor controller looks really old-school, with what looks like 1k eproms in it.  If there's any interest, I'll take some photos as I repair and run up these units.
Hiding from the missus, she doesn't understand.
 

Offline spencerlinnTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Re: Automotive ECM comparison
« Reply #6 on: February 12, 2015, 11:52:15 pm »
Can I make a guess? The Exhibit B is a Japanese ECU.
As a matter of fact is OK what they've done in regards with the through hole components. The ECU location is not exposed to hard mechanical shocks because it's not attached on a door or wheel but on body, which is sort of isolated by suspension.
The thing you liked the most, the heat management of the Bosch/Siemens ECUs, shows in fact lack of efficiency. Newer  designs (shall I say better?) do not need that much heat sink. That is money you spend extra when buying BMW, and they are not too reliable either.

It makes sense as was said, that the Siemens ECU controls the individual coils itself, and it makes sense since they use individual coil-on-plug, so my car has 8 coils to run, whereas most Japanese cars I've seen run either single coil (with distributor) or wasted spark. And since they love 4-bangers, that's only two coils also, so easier to run with external hardware I guess?

Also regarding the mounting and shock, the Japanese ECU is bolted to the passenger floor, no waterproofing etc. The issue I had was that saltwater from your shoes in the winter seeps through and corrodes the ECU housing, which also isn't sealed. No indication that it corroded or shorted anything inside, but looked a bit.. dare I say.. how ya doin'?  :clap:  Anyway, the Bosch ECU is mounted in a sealed compartment, on isolated mounts, and also includes a blower fan that appears to heat and cool the compartment as needed, perhaps to control drift? I don't know if anything in an ECU is that temp sensitive, but I'm also thinking that cars do see pretty extreme temperatures. It appears they certainly spent more time designing the thing. I've never had or experienced any problems with the Bosch or Siemens ECUs, and I haven't either with the Japanese ones but I guess seeing how little effort the Japanese put into them is sort of disconcerting. Maybe it's unfounded. Maybe they work fine. Maybe the German ones do fail more often? But it's hard not to think that if the Japanese ones are reliable, it's only because they're relatively unsophisticated?

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf