Programmable ECUs

[Hiten]

Hello, I am a masters student studying at the Nelson Mandela Metropolitan University in Port Elizabeth, South Africa and would like to do my master thesis on the Formula student car, particularly the ecu. I would like to find out if there have been any teams that have developed their own programmable ecu's or do teams just purchase aftermarket programmable ecus?

Thanks

[Hiten]

Hello, I am a masters student studying at the Nelson Mandela Metropolitan University in Port Elizabeth, South Africa and would like to do my master thesis on the Formula student car, particularly the ecu. I would like to find out if there have been any teams that have developed their own programmable ecu's or do teams just purchase aftermarket programmable ecus?

Thanks

[Erich Ohlde]

3 main methods for fsae/fstudent:
1)purchase a programmable ecu, (AEM, Motec, Performance Electronics, Electromotive, Bosch)
2)purchase a 'piggyback' unit, modifies the stock engine computers fuel and ignition signals, e.g. Power Commander
3)build your own, check out mega-squirt.

[Hiten]

Thanks for your reply. Have there been any successfull custom built ecu's?

[screwdriver]

Yes, check out
http://www.megasquirt.info/

There aren't many DIY ECUs out there, I know about. The only team I can think of that did that was TU Munich, which only sort of worked. I also don't know the details about their system, but I suppose that they didn't build it from the scratch.

[Tech Guy]

I believe Cornell has built their own ECU for many years.

[Biggy72]

We got our first engine up and running last week with megasquirt II. Sounds pretty good and we haven't hardly messed with the base map at all.

[jrickert]

I don't remember the specific teams, but i distinctly remember seing several teams with custom built ECUs last year that were not Megasquirts. Some years ago, we built an ecu, but abandoned it because of the ease of use of software. Its one thing to build an ECU that works. Its much more difficult to build one with good mapping sofware that is easy to use with all the luxeries of say a Motec or Pectel. If you are not familiar with Engine management, i would highly suggest trying playing with an off the shelf commercial prodect to see how it works before building your own. The embedded hardware is only a small portion of the battle. Many people seem to overlook this. A usable ECU needs to be robust and human friendly.

Also, i often wonder what use a custom built ecu is. I wouldn't think it would be of much help in design unless if fulfills some need that an off the shelf unit could not. It would be like going up with team built spark plugs. Its really cool and you might have put alot of effort and engineering into them, but in the end how do they stack up against their commercial counterparts? It might help you in cost. You can probably easily build an ECU for under $100. Thats the only advantage i can see. Then again if its cost you are concerned about, use an ecu off a production car.

I have long thought about building my own ECU. I think it would be really cool, and i know i would learn alot, but in the end i don't think i would provide much advantage. Is the educational value alone worth it? Possibly.

[Biggy72]

I'm not saying MS is for everyone, but I think they are something that should definitely be considered. I would not use a MS setup as a primary setup until a team is comfortable with it. Until then get a performance electronics or something similar that's not too terribly expensive to get everything up and running. We have a couple guys that have built several so this year we decided to swap over.

With that said you can get quite a bit numerous features and options with a MS board. Almost every bit as what I have heard and seen on some motec boards that cost 10x what we paid for ours. We built two ecu's for this year, one with sequential fuel and spark, and another with batch fire. The batch fire should be pretty reliable since our guys have quite a bit of experience with it. The sequential will be a bit more of an experiment, but we should be able to go from one to the other in the same car with teh same harness pretty quickly.

From my experience the MS software is much much easier to use than the PE. It has quite a few more features and much more adjustability.

[exFSAE]

A handful of teams run Mototron ECUs. They looked cool. I recall looking at pics from 2003 comp, maybe earlier, with Wisconsin being sponsored by them. Think Oklahoma and Colorado may also run it.

They provide the hardware / harness stuff.. and you basically code your entire ECU in Simulink. Get to add in whatever functionality (TCS, LC, brake bias distribution) you want.

[Michael Hart]

A number of years ago (I think 7 or 8) some folks from Dartmouth built their own ECU - it was a ton of work and in the end, it was pretty much just comparable to an off-the-shelf unit. We used it for a year or 2 then switched to a commercial one because of the tech support I believe (my understanding is the designers of our unit graduated, leaving us with very little support compared to a full-tinme company)

For what its worth, we currently run the Performance Electronics box - its software is very user-friendly, its got a lot of options/inputs/outputs, which is nice, and most importnatly, Brian at PE is incredibly knowledgeable, helpful, and patient. A few guys on the team have kicked around the idea of building our own ECU again, but i don't think that it will happen because that last part is hard to overstate - it's really helpful to get an expert's opinion when the stuff hits the fan and to know that a company is there to back up their product when you're running a strung-out race motor with it

[B Lewis @ PE Engine Management]

We built our first ECU for the University of Cincinnati FSAE team as Masters students back in the 90's. I would agree that developing a system from scratch is quite a lot of work but it can be done. If you do plan to develop your own system, my suggestion would be to stage the project over 2 years. By the time you get the car running to develop some of the less static features of the ECU, you are already hurrying to get finished for the competition and to shake out all the other major vehicle systems. Also, make sure that you have someone who has a strong hardware background and also a strong programming background. Usually, these qualities do not exist in the same person http://fsae.com/groupee_common/emoticons/icon_smile.gif

Good luck and if you need a sounding board for your ideas, drop me an email.

[Hiten]

Many thanks for all the replies and info. I have developed a programmable fuel injector controller for my Bachelors degree and now will just develop that further incorporating the ignition timing control with additional features for the FSAE car itself. Its great to see how many other functions the teams use such as electronic gear shifting and traction control just to name a few. I am hoping to get this project approved and will you guys posted of my developments.

[j bakker]

The University of Guelph is currently building a custom ECU for our 2008 car. It was the controller for our AWD system in the past and should have all of our data-logging functions built in once completed. Hopefully it is done by the time competition rolls around, come by and check it out (i hope).
Our team is mostly systems and computing engineers so they eat this kind of stuff up, it is actually a 4th year design project.