MS1 extra V3 board revival...

[robocog]

Long story, many years ago now I fitted MS1 with the v2.2 board in my 4 pot car
Mate had a V6 cossie in his and went with MS1 with a v3 board
We were both running EDIS

A few years later my mate sadly passed away - but before this another pal bought a "spare" v3 board from him and did nothing with it for more than a decade or so
Now he wants to fit this in his car running an identical engine as mine...

I of course offered to help (always easier if someone who has trodden that path casts an eye on things)

Got the ecu on the bench and it appears that L1 (1uh inductor) was not fitted,,,and is nowhere to be found
I cannot find one online to buy and I reached out to a seller of megasquirt here in the UK and have drawn a blank...

SO carrying on regardless I cast my eye over the mods that the original owner had done with regards to EDIS
I have compared with the notes in the manual - and they do not tally up

I have undone the mods and done the ones suggested
It Was
D14 > IGN
XG1 > Tach select
JS9 > SPR1 (not sure this was related, but I undid it anyway)
S12C > C30/D2/R12
Tsel > Optoout

It is now as per the manual
IGN >JS10
XG1 > XG2
TSEL > Optoout
Optoin > Tach select

It also appears that R12 got replaced with a 1k rather than the 390ohm suggested...not sure if that also needs swapping back to the suggested resistance to work nicely...

Also R4 and R7 got swapped from the suggested 2.49k resistors to 27K items (I am guessing as he would have been using Ford sensors for air and coolant) so I will have to swap those as we are sensibly using the GM suggested sensors.....

Anyway - I NEED to source L1 inductor before I can go any further with it (as it currently does nothing as there is no 5v present at the ECU)
Any ideas where I can get one?
Part as far as I am aware is 495-5611-1-ND ...which is bringing up no hits when I search for it
I'm in the UK and only need one of them ...which possibly doesn't help my cause

[bobxyz]

For the inductor, you have a Digikey part number. It's the same as:
https://uk.farnell.com/epcos/b82144a210 ... 20inductor

Temporarily, you could just use a jumper wire across L1 -- it won't filter the electrical noise like the proper part, but would let you continue with bench setup/testing.

[robocog]

I have briefly jumper'd it - to make sure it "lives"
It does power up and most stuff on the stim seemed to make things happen in tuner studio (wrong values for temps and no tach)
...and I didn't want to push my luck playing too long on the bench so am on hold for now

Phil from extraEFI has very very kindly managed to find me one - and has it in the post to me, so not long now!

I have got the right resistors for GM sensors here - I know the original builder had Ford sensors and had fitted resistors to suit
and the right resistor for R12 - which I believe is the cause of no RPM being detected
I did quickly use a scope and traced the pulses - up to R12, where it goes deaf
(originally it had some odd wiring mods - including this resistor, so am putting it back the way the instructions say)

Hopefully this is going to be fairly drama free from this point on...

[robocog]

Now I am completely lost ....

for EDIS....
my stim (megastim V2.2) is not giving any RPM on the V3 board

Have verified it against my MS1 (V2.2 board) and I get RPM shown as expected

Now starting to wonder if I have followed the wrong path when making the V3 board ready for EDIS....

Currently it is built up without any VR circuitry (50a only...skipped 50b as assumed not needed)
Tachselect to Optoin
XG1 to XG2
Tsel to Optout
JS10 to IGN
R12 is 390 ohm
D1 and D2 have been removed and have a jumper in place for each

One thing to note - the board pad for D2 at the C30 (unpopulated) end has been burned off (wasn't my doing!)
It has a jumper to one end of the jumpered D2 to make up for the lost pad
(was like this when I got it)


I have traced the circuit based on the schematics and it seems OK like this
Got continuity to pin 1 of the optocoupler from R12
Optocoupler pin 2 goes to ground
Optocoupler pin 4 (emitter) goes to ground

Did a basic check of the optocoupler (in circuit though)
LED side seems to check out as good
Transistor side also seems to check out as good
(get 0.6v forward voltage base to collector and emitter...nothing in reverse)

Running out of ideas as to where to go from here

Starting to wonder if I should put the mods back to how they were before I got my hands on it
optout to tsel
tachselect to XG1
1k resistor in place of R12
link top hole of C30 to S12c

[robocog]

Have rewired as per the extra documentation - and assumed that EDIS is a 12v
rather than 5v
(seems I may have been following the wrong one and it was quite possibly originally wired right for extra?)

With a scope...and using my megastim
Can see the square wave input getting to the optoisolator on pin 1
No square wave at the base collector or emitter....or to IRQ1 on the ms chip (so understandably TS is not reporting anything happening)
I can only assume it is a dead optoisolator

Have found a couple of new sealed 4n25's in my box of bits(possibly over 20 years old, so was buried deep)...and have ordered some solder in IC bases to make future part swaps less hassle, pretty sure I used to have a pile of them, but could only find ones for IC's with more pins
Hopefully they be here tomorrow and I can do more tests...
Fingers crossed that was the issue all along

[robocog]

Inductor arrived, soldered it in
4n25 6 pin socket arrived, soldered in, fitted the new optoisolator...no difference

Is is that the stim I have (megastim v2.2) is not capable of triggering the unit?
Can see a nice square wave going into the opto isolator from the stim...nothing coming out

currently got it wired as per the extra documentation (no output wired as yet)

[bobxyz]

Check that R13, the 4.7K pullup to +5v on the opto output, is good.

You could measure the resistance between the DB-37 pins and the opto:
390 ohms crank in to pin 1 (if D1 and D2 are jumpered)
0 ohms ground to pin 2
0 ohms ground to pin 4
Also measure
0 ohms opto pin 5 to CPU pin 15

[robocog]

Board is currently wired as per the extra manual
Tachin > XG1
R12 changed to 1k
between D2 and R12 is linked to S12C


R13 measures 4.7k - left side R13 goes to 5v reg (0 ohms)
Right side R13 goes to opto pin 5 (0 ohms)

DB37 > opto
bottom row 5th pin in from left (pin24 tachin.)

0 ohms to pin 2 on opto
all others are open circuit to the opto


14k for opto pin 5 to cpu pin 15 (PTD1/MISO)
0 ohms for opto pin 5 to pin 14 (IRQ1)

[bobxyz]

Sorry, I misread IRQ1 as pin 15. You're correct, it's 14.

You've gone back to the alternate opto isolator wiring, where the tach signal grounds the opto LED and R12 is changed to a 1K pullup to +12volts. I don't have the old instructions handy, but do you still have a pullup resistor to +5Vref on Tachin? If so, try removing it.

Edit: Also check the stim board to see if it has a pullup. On the newer JimStim boards, there's a black 2-pin jumper to enable the pullup. You'll need to look at the original Stim docs to see what it has. With the tach grounding the opto isolator, you don't want any pullup on the tach signal.

[robocog]

Massive leaps forward here
It seems the megastim V2.2 is not capable of a +12v to gnd square wave
(seems to only give a 1v square wave - so when VBATTis 12v I only get an +11v to +12v square wave)

I downloaded and printed the Jimstim schematics and compared to the megastim V2.1 schematics (cannot seem to find one for the V2.2 which has differences - as it has a switch to toggle between PNP and NPN) but possibly similar enough a circuit to not matter for what I want

I recreated the jimstim type NPN output and used an arduino (blink sketch) on a breadboard and got it to spit out a good clean +12v to GND square wave - hooked to the meagsquirt pin 24 and got RPM!

IMG_20240127_152238.jpg (98.79 KiB) Viewed 985 times

IMG_20240127_152708.jpg (200.67 KiB) Viewed 985 times

so ...it is not the megasquirt at fault - but the megastim V2.2 I have

Now looking at how to drive the NPN jimstim type output from the megastim's AD654, but drawing a blank...not sure it is even possible based off what my scope is telling me - so brings me back to using an arduino to trigger the 2n3904

I have already got some perf board and an arduino on the stim (used it to spit out wheel decoders for a speeduino)...along with a couple of LEDS - again for the speeduino setup

I could set that to 4 cyl dissy output and then create the "jimstim" transistor setup that /does/ work and hook that to the arduino...

IMG_20240127_152918.jpg (177.21 KiB) Viewed 985 times

BUT running out of real estate on the perf board and would be a lot neater if I could get the megastim to put out a clean +12v to gnd square wave and use the RPM pot that is on the stim...

[bobxyz]

I looked at the Stim v2.1 schematics and the problem is that the 2N3906 drives the tach signal hard to approx +12volts, but the tach signal is only pulled down weakly to ground via stim R4 10K. The MS2.3 board, with your latest mods, pulls the tach signal up to 12volts through a 1K resistor. So stim+ms, with the 2N3906 off, has 12volts across 1K + 10K, which only swings between 12v and ~ 11.0v.

If you replace stim R4 with a 1K resistor, or just add a 1K from the stim tach signal to ground, it should be enough to turn on the LED in the opto-isolator.

[robocog]

I tried adding a 1k between tach and gnd (left R4 in place)sill not enough to trigger the ECU
(but switching between NPN and PNP on the stim did bring the LEDS on for the 3 second prime)

Still not close to getting it to ground level according to the scope
(still gives square wave between 7 odd volts and 6volts)
(this image is with a 1k between tach and gnd)

IMG_20240127_194123 (Medium).jpg (157.84 KiB) Viewed 971 times

Looking at the scope, the only time I get close to it coming to gnd level is when I flick the switch between NPN and PNP

[bobxyz]

^^^ Where is the scope connected?

If you disconnect the stim board, do you see a 0 to 12v square wave on the Stim-to-MS tach signal?

With it connected, I'd expect the tach signal to swing between ~5volts and 12volts.
(5v low swing voltage would be: +12v to MS R12, 1K to opto LED, ~2volts across LED, MS tach pin to Stim pin, 1K to ground)

[robocog]

With the stim disconnected there is nothing (as it uses the 5v coming from the megasquirt to power it- DB37 pin 26 - VREF)

scope at the tachin (pin24 BB37) with the stim connected to the MS

[bobxyz]

I couldn't find stim v2.2 schematics either, but pictures are still available at: https://www.diyautotune.com/product/meg ... bled-unit/

I'd try removing your changes from the stim board and then capturing scope pictures of the tach signal in both NPN and PNP positions. I'm guessing that the v2.2 NPN setting will actively drive the tach signal low (PNP actively drives high), and should be compatible with your MS tach/opto wiring.

[robocog]

Given up on the megastim 2.2 board being able to work right

set the arduino (running ardu-stim) to output a 4 cyl dissy pattern (so output is a 5v square wave)
removed the transistor Q1, R4, R5 from the stim to "disable" that side of stuff
made up the same transistor and resistors as Jim stim base fed from the arduino (giving a nice clean +12v - 0V square wave) - and fed into Pin24 (tachin)
It works,so is doing what I need - am now able to sweep and set RPM's and check functionality

Onto the task of trying to fit the seqeuntial shift lights to the proto area, test it works and send it to its new home

[robocog]

I have found the solution!
The transistor in the original megastim v2.2 is obviously not turning off fully (which explains why such a small square wave and always positively biassed based on the vbatt voltage)
so a bit more digging ...

Spotted this
https://www.miataturbo.net/megasquirt-1 ... ion-27953/

megastim_mod.jpg (12.78 KiB) Viewed 890 times

And there it is ...the right answer - putting a 1k ohm resistor on the base of the transistor has solved it, transistor is now actually fully shutting off and turning on
I now have a square wave that goes down to zero volts!
It triggers the ecu as intended, and is nice and clean
All is good!