ECU software and testing

[Aly]

I'm new to working on ECU's
I wonder if there is a known procedure for testing the ECU performance...
Is it useful to simulate the ECU using certain software?
I didn't find any such software to simulate the ECU ...
please help
any ideas how to start on a good procedure for the ECU
thanks

[Aly]

I'm new to working on ECU's
I wonder if there is a known procedure for testing the ECU performance...
Is it useful to simulate the ECU using certain software?
I didn't find any such software to simulate the ECU ...
please help
any ideas how to start on a good procedure for the ECU
thanks

[Mbirt]

If you're using a member of the Megasquirt family, it is useful to have a "stim." Megasquirt Stimulators (giggity) (http://www.msextra.com/doc/general/stimtesting.html) The Megastim and JimStim are both useful for simulating sensors and testing the ECU's response in the tuning software. They're especially helpful if you've modified the ECU or are diagnosing a problem.

Some controllers can be calibrated with Matlab/Simulink and I'm guessing that setup would allow you to "simulate" everything pretty easily.

[wweissin]

The real question is why? What are you hoping to accomplish. Have you built your own ECU and need to test it. To you want to see if a manufacturers ECU does what it clams to do?

[Aly]

wweissin

the issue is that we have an ECU of a team that their car never worked
so I need to know weather it's working well or not

besides they want to upload different maps and test to see which map is optimum and use it

[Aly]

by the way our engine is a
GSX-R600
of a suzuki race motorcycle coming with its own ECU

how can I work on such an ECU
what tests should I conduct
http://fsae.com/groupee_common/emoticons/icon_confused.gif

[wagemd]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Aly:
by the way our engine is a
GSX-R600
of a suzuki race motorcycle coming with its own ECU

how can I work on such an ECU
what tests should I conduct
http://fsae.com/groupee_common/emoticons/icon_confused.gif </div></BLOCKQUOTE>

If its stock, hook the ecu up to the motor, connect power, fuel, etc. Hit start.

Otherwise, you can collect a bunch of spare sensors, injectors, and coils and lay everything out on a bench. Simulating the cam and crank signals is the hard part which is why doing it on the motor itself is probably easiest. I don't know how robust the starter on the motor is, but if you are low on resources, you can pull the spark plugs and fuel pump and crank the motor over to get all of the required inputs (Just don't crank for long periods of time: long enough to check for spark, fuel injectors, etc, nothing more). I'm not going to explain everything, you should be able to figure it out from there with a little creativity.

[Aly]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">. I'm not going to explain everything, you should be able to figure it out from there with a little creativity.
Daniel Wageman </div></BLOCKQUOTE>
Dear wagemd,
What you're talking about sounds pretty logical and easy (no doubt we would have done similar tests ) but ..........
what I meant by my question is to know if there a reliable known procedure conducted to test an ECU like for example connecting it with all its ports to a certain module and then connected to the computer .... where a certain software simulates inputs and output as if the ECU is really running on the engine.
A safe test ,,, a test without messing with anything on the engine .
An important fact that our engine and ECU are not stock at all,We can't take the risk of ruining anything..at least for the coming 2 months.
Any way I guess from your answer that the ECU testing procedure goes by improvisation meaning that there is no fixed reliable procedure that is usually followed on these cases.
Anyway thanks and any help would be appreciated

[murpia]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Aly:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">. I'm not going to explain everything, you should be able to figure it out from there with a little creativity.
Daniel Wageman </div></BLOCKQUOTE>
Dear wagemd,
What you're talking about sounds pretty logical and easy (no doubt we would have done similar tests ) but ..........
what I meant by my question is to know if there a reliable known procedure conducted to test an ECU like for example connecting it with all its ports to a certain module and then connected to the computer .... where a certain software simulates inputs and output as if the ECU is really running on the engine.
A safe test ,,, a test without messing with anything on the engine .
An important fact that our engine and ECU are not stock at all,We can't take the risk of ruining anything..at least for the coming 2 months.
Any way I guess from your answer that the ECU testing procedure goes by improvisation meaning that there is no fixed reliable procedure that is usually followed on these cases.
Anyway thanks and any help would be appreciated </div></BLOCKQUOTE>
The kind of equipment you suggest using does exist, often known as hardware-in-the-loop (HIL):

http://thedailywtf.com/Articles/The-Test-Machine.aspx

However I would think it completely inappropriate for FSAE.

Regards, Ian

[Aly]

Dear Ian ,
Thanks alot for trying to help .
Unfortunately the article is not informative and has nothing to do with our technical work .

I really appreciate your reply and thank you another time for trying to help
regards
Aly

[TMichaels]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">The kind of equipment you suggest using does exist, often known as hardware-in-the-loop (HIL):

http://thedailywtf.com/Articles/The-Test-Machine.aspx

However I would think it completely inappropriate for FSAE. </div></BLOCKQUOTE>

Well, I would not say that. There are teams out there testing their complete electronic setup like this including gear shifting systems etc. It really helps you to sort out your electric gremlins before putting it all into the car.
Additionally there are teams developing their own engine management units. They also benefit from using HIL systems.

Regards,

Tobias

[murpia]

I agree that some teams, especially those developing their own ECUs, would benefit from a HIL setup. But teams at that level of competence already know that, and won't be posting here asking for help. The point of my link was to suggest that nothing is as easy as it seems and that one should not underestimate the complexity necessary to benefit from this approach.

It sounds to me like Aly is asking for a turn-key solution to his problem (and by implication a cheap and quick one). He has rejected the previous posts suggesting simple ways to bench-test his ECU, yet the leap from there to a full-blown, validated, HIL is huge.

Regards, Ian

[Aly]

Dear Ian and Tobias
First I really appreciate your intention to help
and I apologize for any inconvenience from my side
.......
any way ,, I started to figure out now that ((Mbirt's)) comment was really useful
In fact I never knew what's an HIL till hours ago.
Any way if you're the first one ever to test an ECU in your faculty you'll know what I mean .
Reading,I've found that using an HIL is perfect and might be essential for our case (it's essential for all teams modifying their ECU's as I understood)
So I'm asking you now
What HIL software or package FSAE students usually use ?
any help in the HIL issue would be appreciated
regards
Aly

[wagemd]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Aly:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">. I'm not going to explain everything, you should be able to figure it out from there with a little creativity.
Daniel Wageman </div></BLOCKQUOTE>
Dear wagemd,
What you're talking about sounds pretty logical and easy (no doubt we would have done similar tests ) but ..........
what I meant by my question is to know if there a reliable known procedure conducted to test an ECU like for example connecting it with all its ports to a certain module and then connected to the computer .... where a certain software simulates inputs and output as if the ECU is really running on the engine.
A safe test ,,, a test without messing with anything on the engine .
An important fact that our engine and ECU are not stock at all,We can't take the risk of ruining anything..at least for the coming 2 months.
Any way I guess from your answer that the ECU testing procedure goes by improvisation meaning that there is no fixed reliable procedure that is usually followed on these cases.
Anyway thanks and any help would be appreciated </div></BLOCKQUOTE>

If you wanted to try the really gheto approach, you could cut out replica cam and crank trigger wheels (that match the layout in the engine) and spin them in a fixture front of their respective sensors. If you lay out everything else on the bench, you should be able to see everything firing as it should. It also gives you the opportunity to o-scope things. I usually try to start with the simplest and cheapest solution... If that doesn't work, I find a CSE guy...

[Jon Burford]

A different approach,
Save yourself a lot of time, stress, and agony and buy a stand alone ECU,
DTA are brilliant units at really good prices with brilliant support.
http://www.dtafast.co.uk/

[TMichaels]

<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Aly:
Dear Ian and Tobias
First I really appreciate your intention to help
and I apologize for any inconvenience from my side
.......
any way ,, I started to figure out now that ((Mbirt's)) comment was really useful
In fact I never knew what's an HIL till hours ago.
Any way if you're the first one ever to test an ECU in your faculty you'll know what I mean .
Reading,I've found that using an HIL is perfect and might be essential for our case (it's essential for all teams modifying their ECU's as I understood)
So I'm asking you now
What HIL software or package FSAE students usually use ?
any help in the HIL issue would be appreciated
regards
Aly </div></BLOCKQUOTE>

You do not need HIL for this kind of test. Just go with the suggestion of wagemd. Stimulate the crankshaft and camshaft sensors and see if the ECU tries to fire. Make sure that the sensor signals are plausible.

Regards,

Tobias

[mech5107]

Cut a motronic trigger, get a VR sensor, plug everything, attach trigger disc to a lathe, mount the VR sensor, spin the lathe to 1000rpm.

See if there is spark, see if there is inj. pulse. (connect injector and feel it vibrating)