Okay, I noticed that now with the stim connected to my MS the warmup LED never goes off. Evenwhen I have the coolanttemp sensor adjusted to where the fast idle LED on the stim is off the warmup LED on the MS is on.

 

Hrmm, this could be a burnt out transistor on that LED. If you're seeing the temperature in megatune go up and down but the led is always on, then I'd say that's the case.

 

Will this cause the MS to not function correctly or is it just going to cause the LED to stay on?

Also Do you have any idea why this transistor would have "burned out"?
Should I be looking for another problem before replacing the transistor?

 

Well I replaced the Transistor with a like component but the problem is still there.
Now I`m starting to wonder if I`ve wasted a shitload of money on something I`ll never be able to get right.

Is it possible that one of the modifications that I made Like the mod to use the factory air temp sensor or the mod to use the BAC would cause this problem?
I really want this thing to work. If the warmup light staying on will NOT affect the Ms operation then I`ll live with the light being on all the time. I doubt I`m that lucky though.......so.... whats the next step. And I hope its not replace that transistor again, the board can`t take anymore heat at that point.

 

I can't help you with your problem, but I can say hang in there, its VERY easy to get discouraged with this. It took me almost 2 weeks to get my properly assembled and running when I had orinally figured it would be a weekend project at most but now mines in and I have been smiling since!

 

There are a lot of things that can cause that LED to stay on... but that needs to work right.... if you're using the MS to control ignition, all three led circuits are used to drive the ignitors....

It's possible you have a solder bridge or something like that on the bottom of your board that's causing that led to always see +5v...

Like I said tho, as long as your coolant temp is above 160 degrees (and it's showing that in megatune) that LED should go off.

 

Without more specific help I don`t stand a chance correcting this.

I looked at all solder points on the back of the board and do not see any bridged.
I know this LED worked correctly at one point but somewhere along the way in modifing the damn thing to run the 7 It got screwed up .


At this point I`m rather pissed, I`ve got several weeks and almost $300 in this project and now I don`t know where to start in troubleshooting.

As an alternative is it possible to rebuild that circuit in the proto area.... or SOMETHING???????

 

OK, you didn't tell me that it was during the mods that it stopped working.

De-solder your 4.7k pullup resistor and tell me if it starts working again. You may have soldered the pullup resistor to the wrong leg of the LED, or maybe it's touching something that's inline with that leg, causing +5v to be permanently connected to the + leg of the LED.

Getting pissed won't help anything. I have to go on very little info and figure out what the problem is and help people who may not have much background in electronics or troubleshooting. Sometimes it's hard to guess what is wrong with such a small amount of info to go on.

 

Hey Ken, sorry bout the attitude, I`m not in any way upset with you. Just the opposite, I`m very grateful for all your help. I`m just upset with the situation. I can`t afford another kit

I had already thought of the pullup resistor and that is not the problem, light is still on all time even with it removed. I know 99% that light worked correctly (responded to coolant temp in sequence with the FIdle LED on the Stim) when I first completed the MS. Somewhere in the process of modding the board I had to screw something up. Two seperate transistors have been replaced Q7 today in attempt to fix this problem. Also Q4 with a 689 trans. as recommended in the FAQwhile modding for BAC control. During that same mod I replaced R18 due to a typo in the FAQ but you corrected that for me so I replaced with original. then changed R19 with 510 Ohm 1 watt metal film resistor (could not get the 1/2 watt) Besides the pullup resistors the only other thing changed was R4 with 39k resistor. I even replaced the pullup resistor with a new one just now but I`m getting nowhere still. I`ve used a jewelers magnifier to look at everything again. the only things I don`t like are the two above mentioned transistors, I got the board pretty good while removing the old ones and soldering in the new ones. But I think they are not bridged.

As aked before, can I build a seperate circuit to replace the current "warm up" ?

 

you can, build another circuit in the proto area if you want, but it'd be better to figure out what got screwed up. The processor could be damaged, the transistor, or vias that the transistor goes into could be damaged.

Now that I think about it, the transistor circuit provides ground to the LED and +5v is always supplied. So just look for shorts/etc... that keep the led grounded.

Start by putting the DMM in continuity mode, and touch the negative lead of the LED with one probe, and touch ground with the other. If it beeps then you know there's a short. If not, put the dmm in resistance mode, and check for low resistance. Do these tests with the megasquirt turned off. If you get no continuity or the resistance is very high, then it's likely that you have some kind of short that's turning on the transistor.

Do these tests and let me know what happens.

Ken

 

Ken, I don`t have a meter that has dedicated continuity, Its in the back of my service truck packed. I do have a DMM that reads DC volts, DC Amps, AC volts, AC Amps, and five ranges of Ohms here at my desk.
On the 20k ohm setting I checked the LED neg. lead to the heatsink ground.
In fact I checked all three LEDs. Both neg to ground AND pos to ground just to compare. D15 is the warmup LED that is the current trouble spot. They all (the LEDs) show the same with my meter

D14 neg to ground 5.33ohm pos to ground 1.05ohm
D15 neg to ground 5.31ohm pos to ground 1.05ohm
D16 neg to ground 5.31ohm pos to ground 1.05ohm

D14 and D16 both respond to the stim being adjusted as far as injector pulse( rpm) and acceleration (tps up)

 

ok so there isn't a short circuit there then... the problem must be with the transistor turning the LED on all the time. Since you switched out the transistor, that should rule out the transistor itself....

That means that either 2 of the leads of the transistor are somehow shorted together, the processor is broken and turning on the transistor when it shouldn't, or there's another short to +5v somehow turning on the transistor.

To test for +5v going to the transistor when it shouldn't, you can put your multimeter on volt mode in the +5v range, put the negative probe on ground, and the positive probe on the proper pin on the processor. I'm not sure which pin it is off the top of my head, so I would just measure from the processor side of the 1k resistor that goes to the base of the transistor. If you have +5v there, then something is shorting +5v to the transistor, or the processor is broken.

You may be able to tell which situation is occuring (broken cpu or short circuit) by doing the above test first to make sure that there's really +5v on that line... then turn off the MS and pull the CPU, and turn it back on. If only the one that was staying on before glows, you know that there's a short circuit, and you at least know where to look for the problem.

 

Alright, on the MS site the schematics show that CPU pin 9 is "warmup, 8 is accelerate and 7 is "squirt" (injectors). I have a constant 5 vdc pin 9 to ground no mater where the Stims coolant temp sensor pot is set. pin 8 responds to the Stims TPS pot perfect and pin 7 responds to Stim RPMs. I removed power from the unit and pulled the CPU then repowered the MS. no LEDs are on without the CPU. I can make the LED in question come on by jumping two pins on the transistor. Looking at the transistor from above, the flat is at the top. by jumping the pin on the left and the pin in the middle the LED will light up. Based on what you have helped me learn so far then it appears there is NO short on the board causing this. It looks like the CPU is bad (based on the tests) Now just one question, is it POSSIBLE the Stim could be the problem? Could it be showing its own LED for FIdel responding to temp but NOT sending the same info to the board? Just trying to cover all the bases.

 

My main question is:

does the coolant temperature gauge in megatune respond to changes in coolant temperature on the stim?

If it does, when it goes above 160, pin 9 should turn off. If pin 9 does NOT turn off, you need another cpu. First you need to figure out what happened to burn out that pin on the cpu though. I'd figure that out before putting in another one. It's possible that if you didn't remove the CPU while doing your mods you short circuited something while soldering or something like that.

If that guage is NOT responding to changes on the stim, or the guage doesn't go above 160 then yes, it could be a problem with the stim, or with bias resistor changes... (maybe you changed the wrong bias resistor?)

Ken

 

Ken, The coolant temp gauge DOES respond to changes in coolant temp on the stim at ANY time when the CLT pot is turned.
The Stim FIdle LED goes out only after the RPM has been raised above idle and THEN raise the temp to about 194` or more then the Stim FIdle LED goes out. Below the 194 mark the Stim LED comes back on.

When the Stim FIdle LED is out and the guage in Megatune shows above that 194 mark there is STILL 5vcd present at pin 9.

I did forget to remove the CPU at one point during mods so maybe thats what happened.

Thanks again for working to help me get this straight.

Sorry I almost forgot. The only resistors I changed are as follows
R-4 to 39k at 1/4 watt
R19 to 510 at 1watt
R-18 was changed due to a typo in FAQ but I changed it back to original 270 at 1/4 watt.
Also added the pullup resistors. thats it.

 

Which firmware are you running?

I've never seen the CPU break in the manner this looks like... it still seems like maybe software is causing this. Usually if a pin on the CPU stops working, it gets stuck off not on.

I still think it could be a broken CPU but I want to be sure before you buy another one. (Luckily another one should only cost you $20 or less.)

 

Ken I`m running your o29J firmware and Megatune 2.25 from the links provided by yourself.

I already ordered another CPU $18 plus shipping.

 

ok, how do you have the software configured? I assumed you were still running the old standard ms firmware (mainly because with the newer firmwares, the LED functions change depending on your settings).

You should check your Codebase and Outputs dialog to see if you have any of the LED's set up as anything special.

If I would've known you upgraded your firmware I could've avoided all this

Having another CPU isn't a bad idea anyway. It's good to have a spare in case you do screw something up. So far in all my MS builds (20 maybe?) I've only screwed up one CPU way back when I was doing the original test megasquirt to get the rotary stuff working.

 

Well ain`t that a bitch. I checked the Codebase and Outputs dialog and sure enough the LED in question was not set for "warmup". I corrected the setting, burned it to the ECU and suprise suprise........the warmup LED now cycles off at about 162-164 degrees. hopefully there are no more "issues" to deal with for a few days. I`m a little burned out with this one.

Ken there sure is a lot to learn about setting up this software. No hurry though. I have my CAS with 2 teeth cut ready, just need to get a baseline setup (S4 NA) and try to get the 7 started with the MS. I belive I`ve have the wiring adapter finished, just need to re-check all my connections against the pinout. I`m getting there slowly but surely.