Thread: E30 M50 turbo
View Single Post
Old 05-01-2007, 10:03 AM   #12
MarkD
Banned
 
MarkD's Avatar
 
Join Date: Apr 2002
Location: In Your DME!
Posts: 2,414
Quote:
Originally Posted by PedroBMW
I was implying for the DIYer, not many people can do the tuning themselves using the stock DME like they do with hondas where theres tons of guys who can do it all over for cheap. Chip tuning costs hundreds of dollars to redo if you need a retune. Either way I am in no way defending one side or the other, I for sure think chip tuning can be a great way to go. I really just wanna here Markd's opinion on megasquirt since he's knocking it and hes prob one of the most knowledgable people on here to comment. I hope he chimes in again, c'mon Markd.
I bet that in the end, using the stock ECU is not much more expensive than using MS. It may even be cheaper.

Here is what I don't like about MS:

- in most cases you have to cut into the wiring harness and replace various sensors. Maybe now it's possible to install it with the original harness and connectors but I don't know.

- you have to build extra boards to handle the coils, as Jon K did. I am sure that people don't realise that these hand-built circuits are not as reliable as a production ECU.

- Looking at MS, it does not seem to be designed for high reliability. It's more like a hobbyist ECU. I've been following it since it's inception and some of the early layouts of the PCB and the circuitry was not very good. I feel I can comment on automotive hardware as I've been involved in the design of some electronic modules for cars since 1991 and am aware of proper design practices. People have to realise that there is a difference between "something that seems to work" and a product that is qualified for an automotive environmnent or use. It's just like software, lots of people can whip up code that runs, but it often fails when certain unexpected situations occur.

- An ECU such as the ones running the M50 motors has very sophisticated hardware and software. There are a large number of tables (maps) to handle timing and fuel under all conditions such as warm up, part throttle, etc. How many maps? Over 30 in the typical 402 or 413 ECU. To properly tune a MS will end up being a lot of work, and most people don't have the experience to do it.

The software has all been VALIDATED under many operating conditions and there is pretty much no chance it will fail due to some unforseen operating condition. I can't say the same is true for MS.

- people don't realise how good the ECU's in a BMW actually are. The 4xx series from the M50's are quite powerful, and the next generation as used in the 321 hp Euro M3's and E46 M3's have very powerful CPU's. In Europe, people put twin turbos and MAF's on M20's and M30's with the original ECU's. If you move to an MS from those ECU's it's like going from a P4 to a P1 processor in your PC.


So above I have outlined some of the main reasons I don't have any interest in MS. I'm not saying this because I tune BMW ECU's. It's because I know the ECU's and what they are capable of pretty well. I've been repairing them for almost 20 years now so know what fails and also how well they are designed.
If I was turbocharging a car I'd certainly stay with the original ECU even if I could not tune it myself. It should be cheaper to do so and certainly more reliable.

Jon K has his car running pretty well but I'd be interested to know how much it cost in his time and funds. I'm also pretty sure he has more talent than most people at getting this sort of project going.

I bet that in the future, you may see more people moving away from MS and trying to use the factory ECU's on BMWs. I had told Todd at TCD to stay with the factory hardware about 2 years ago, and he said he was going to use MS. He recently posted elsewhere that he bought a Motronic editor so there must be a reason for that.

Last edited by MarkD; 05-01-2007 at 10:08 AM.
MarkD is offline   Reply With Quote