OK my coolant temp sensor reading in erratic. It goes from like 180 and climbs steady and really fast to about 300. It does this on the stim and on the car it self. Sounds like some problem in the ecu it self to me, but this is my first ecu. And that makes a common problem. With the ecu the only constant. Should i look for an open? Or could it be progaming? Some times it likes to hang around the 180-190 area but soon starts changing. Any ideas? Thanks

 

Sounds like a circuit problem. Did you follow the instructions at www.megamanual.info when you were building it?

Can you post a short datalog of it happening?

Ken

 

Ok tried to upload a log but it wont take it. Ill try again after work. And could flux cause this? Just an idea maybe i didnt clean it well enough?

 

I've never seen Flux cause this sort of issue, but a short of some sort could. You should definitely make sure you clean off the flux after you're done.

Ken

 

Hmm could an open on a trace cause it?

 

Think that might be it looks like it comes from the clt resistor to the db37 on the clt pin out? How could i fix it?

 

ok found r4 and r7(coolant) resistor have different resistances and they are supposed to be the same right 2.5k?

 

Can I use a 2.2k resistor?

 

Ok well last night the resistance of r7 was 3.6kohms. Tonight I rechecked before removing it and it was1.37k. So I removed it and checked the resistor, the value was correct 2.49k. So I put it back in and rechecked and i got 1.37k again!So, since I forgot to check across the points without the resistior, Guess I need to do that. I also tried a 2.2k resistor and saw the same 1.37k. Starting to wonder About the board at that spot? This sucks, (for now) I just want it to work. Oh well Guess ill hope for the best.

 

OK!

You can't usually measure the resistance of the resistor in the circuit because there will be other things decreasing the resistance across those 2 points. If the correct value is 2.5K (I can't remember sorry) then clean the area and put the part back in the board. I would suggest getting, a flux pen, a toothbrush, some iso alchohol, solder and iron, and reweting each and every joint on the board until all the joints are perfect. Flux helps LOTS! Next clean the board with the iso.

You should also try to learn to read resistor bands so you can check that you have all the right resistors in the board.

 

I did clean the board with acetone and tooth brush. Yeah I know your not supposed to check resistance on a part connected in a circuit but If I remenber right the schematic is identical to the IAT circuit and it reads 2.49k. Just threw me off. Im going to check all the jointsw again. Thanksb

 

Where did you ground the temp sensor?

 

I have gounds next to the ecu and then from there they are also grounded on the block. The problem also occurs when pluged into the stim as well.
Anyone else having problems accessing anything megasquirt related on the net? The megamanual comes up and wont load any pages except the home page. Nor can I get up the forums for it. What a day. Oh well hope it works later.

 

You probably should not ground both on the chassis and the block, but if it's happening on the stim, I still think the most likely culprit is a short or some other problem.

Ken

 

OK so Guess ill track the culprit in the diagram.

 

Ok, I swapped all the resistors and caps from the iat to the clt circuits and same results. I pulled the ms2card and plugged in the stim, and watched the voltage change when moving the clt pot. I checked the resistance in the circuit while the card is out and I see 2.49k ohms and 1.37 with the card in. The iat circuit still has 2.49k all the time. So what about the card is messing things up? Could the flash cause problems? Like this anyway? Also in megasquirt if I go to themisitor cal. and set the ferinheight scale opposed to celsius the reding stays at 180. Any ideas? Thanks

 

A damaged ADC in the processor is a possibility. Try measuring resistance between the (non-ground sides of C5 and C7) and ground. For clarity, that would be C5<->ground and C7<->ground. I'm sure there's other ways for the internal circuits of the processor to exhibit damage, but the 2 I've seen damaged from folks (the same guy!) accidentally placing +12 to VRef exhibited similar behavior on the damaged ADC input. Ken is probably much more literate than I on the inner workings of that processor's ADC pins - protection diodes, pullups, etc.

 

ok so I have been playing with it. I reflashed ms2/extra2.0.1 . And the reading looks little better. But is bouncing from 170-200. Should the reading be steady and not changing like that? Im going to try flashing the standard ms2 code and see if any thing changes.

 

The temp readings should not be bouncing around. They should reflect the resistance in the temp sensor/stim pot, depending on what it's plugged into. When the CLT or IAT inputs are completely disconnected the observed readings will slowly (over 2-4 minutes) "float" towards a higher temp; that would be the exception to the previous statement.

The resistances are equal and OK with the MS2 card out but exhibit a difference with it in, correct? If so, I think the problem has been logically localized. If not, I'm misunderstanding the problem.

-Mike

 

Logically localized? hmm so operator programming error? Or operator error?lol So. I have a bad card?

 

Don't even bother with 2.0.1 code, use the one ken put up, or give me your email, and i will send you 2.1.0d which I have been using for a few months with no worries.

 

badasshonda2001@yahoo.com Cool think that code 2.1.0 could be a problem? I was looking at your contribution thread and was wondering about the tables for ve and ect.. The throttle vs rpm colums and such look different from what I saw with the 2.1.0 code? will that change when I try the update to 2.1.0? And would the table settings be ok to start a 13bt? Just start? Thanks in advance ill try the code asap. Im so close I can smell it!lol

 

You can try my fuel table, I would suggest starting with a req_fuel value at least scaled up ot the 13B displacement of 2616cc, from the 12A's 2292cc, so 14% bigger, so scale it up to req_fuel=8.0ms, assuming roughly the same fuel pressure and 550cc primary injectors...

You will need to retune your throttle pump settings, they will be NOTHING like mine!

I will forward the original email from ken (maui) to you. This code works... I haven't tested the latest code posted here, which is newer, but until it has some KM put on it, and is tested by someone other than ken, basically assume he knows how to get it working 100% but no one else understands the nitty gritty bullshit until it doesn't work for them.

Its the nature of megasquirt, if you find a core variant that does everything you want, then don't bother reflashing your ECU with newer code because ken will change a whole lot of things to improve the quality of the ECU, but that does mean that frequently enough stuff will change that your car will run bad until you spend serious time on a retune....

[EDIT]

Right now I am telling you to move forward from 2.0.1 release, to 2.1.0d because I had problems with the 201 release, once I was able to explain the Ken the serious problems with it, he found them and fixed them, and sent me new code which works flawlessly in my eyes. Without critical review from a rotary user, assume that new code is basically *untested* alpha code.

I can tell you with 100% assurance that the code I'm about to email you does work with no known serious problems... If you can't get this working it is your own fault :P

 

Just for clarification, NO this new code will not fix your coolant temp sensor problem... I told you what to do regarding that, check the circuit solder etc, check there is continuity between the ECU pin and the coolant sensor pin in the loom.

The updated code will fix a few problems you haven't encountered yet

 

Circuit solder seems good, and I cleaned the board 2 times. I also looked over every inch to look for bridges and spatter etc. I can remove the ms chip and check voltage on pin 25 of the 40pin socket. With the stim pluged into the ms I can watch the voltage change in accordance with the coolant pot. That tells me everything with the board(-the ms2 daughter card) is working fine. Right? And when the ms2 card is put back in everything with the coolant circuit goes to crap. Im going to update to the 2.1.0 code anyway.
Ps. When I put the card back in earlier the pins for js4 and js5 broke off. How bad could that be? Looking at the chip schematics they go nowhere until I jumper them and have a use for them?
My apologies for being a ms newb. And Thanks for giving up your time.