Megasquirt users thread to discuss all things Squirty

[oldcorollas]

oldcorollas,

Looking at the msextra (led circuits), and the MS2 circuit I would be wiring it direct, i.e resistor needed, yes resitor used to limit current.

Roger

Hi Roger,
after a little more digging, it seems like..

the ZTX450 adn 2n2222 have a 1K resistor, direct from CPU, to limit current to 5ma (from VCC = 5V)

extra code recommends removing the 1K, and using a 330ohm with the VB921, giving 15ma
DIYautotune sell a 330ohm with the BIP373 (15ma)

the MS2 daughterboard already has an SMD 1K resistor on the board, between the CPU and the DIP20 socket, and assuming that the CPU only deals with 5V, it only puts out 5ma on JS10 (which drives the VB921 on the V3 board directly)


so i'm wondering if in the aftermarket/extra scene, it was found that the 1K resistor on the MS2 board made the current too low to drive the VB921? and if 5ma is enough to drive the BIP373...

hmmmm i've wired it direct and will scope it tomorrow

[oldcorollas]

oh well, i couldn't wait..
all works!! beer scoping ftw

V2.2, MS2, running latest base B&G code (as of 2 weeks ago)

mods:
12V to JP1-6 (for IAC, but not used)

ignition. Mounted BIP373 on case, jumper from JP1-5 to left leg of BIP

thermofan.
JP1-4 to X10 via a 1K resistor (JP1-4 is PAO output in TS)
JP1-1 goes to middle leg of ZTX450
JP1-2 goes to both right leg of ZTX450, and to ground (left side of unused D4 position)
right leg of ZTX450 goes to X14 and thus pin 31

fired it up on the Stew-stim, set ignition to inverted. put a pullup on the BIP output and the dwell is spot on.
as you go from low to high rpm, the dwell is fixed and spark time long, then the spark time decreases until you reach the set amount (TS has max spark, but is actuall min spark). once that min spark is reached (say, 2ms) then dwell reduces, but spark time stays.
edit: hmm, maybe the signal is the wrong way around..... will have to check today

Fidle runs as per normal

thermofan, enable PAO spare output in TS, coolant > 90, hysteresis 5 (as starting point), and works good (tested with pullup to X14)

now to add stuff to car


edit: the DIY AT Weatherpack connectors are awesome!
3x 2 pin connectors for $6? cheap!

[MWP]

So it sounds like VR is a pain in the ass

Has anyone tried converting factory VR sensors into hall sensors?
Worth the effort?

[thechuckster]

never been a PITA for me - hall effect would require custom sensor mount plus attaching the a custom-made wheel. Go for the LM1815 circuit and the only thing you have to deal with is the polarity at the sensor.

[3jcelica]

VR is not really a problem and they are very robust. As thechuckster says the biggest deal is what polarity, they only way to tell is with a scope, from the original wiring, or by trial and error (looking for that miss/wall at around 3000rpm). Any aftermarket computer is going to have the same issue. I have come across ignition modlues however that don't like the toyota sensor (GM 7 pin HEI for example).

Oldcorollas,

MS extra wires a 330ohm on top of R26, top of r26 goes straight to CPU, therfore the BIP in the extra are just running with 330ohm as you say, seems there is a difference between extra and B&G. I use the 330ohm, but I am sure 1k would work. There was some discussion on the forums about this, and different sizes used for different igintion modules. I have always used 330ohm both with BIP and VB921s and had no issues, haven't used larger but if it triggers the BIP then I am sure it will be fine.

BTW in my last post I meant yes you need a resistor but just use the 1k not 1k +330ohm. May have been a bit confusing

Good to here it is all scoped out, should be straight forward to hook that thing up in the car now.

Roger

[oldcorollas]

thanks for clarification, i assumed that was what you meant

and Charles, that idea works great! 8mm EFI fuel hose and an 8mm hex bar = awesomely easy

ok latest scope

bottom trace is the VR/ignitor/tach input. has a pullup resistor and grounds like it would a coil.
i am guessing the opto inverts it, so in this pic, i've set "ignition input capture" to rising edge.

top trace is on the BIP, which also has a pullup resistor, so it should be as a coil would be driven
coil is set to "going low (normal)"
trigger offset is 0deg, and spark advance is 0deg
the scope is hooked up to what would be the coil -ve, and to ground

from the pic, it looks to me like the coil is turning on (voltage goes high) before the input triggers, then dwelling for the set time, and then firing (voltage goes low) at the trigger point.

or have i got that backwards?????
should it be so that the BIP causes the voltage to go low during dwell, (shorting coil to ground and thus charging), and then go high when coil fires???

I will update with a second pic, with "going high (inverted)" in a few mins... now that i think abotu it.. going high is correct?


pic for going low (normal)



pic for going high (inverted)


come to think of it.. both traces should be the same orientation, sinc eboth are as though they are firing coils
DOH.... S.M.R.T

i'd be buggered if i didn't have a scope!

going to play around with trigger offset and advance to see what does what..

[oldcorollas]

shits n giggles time

everythign hooked up.. and second pic is breakout board. can switch any input between either the stim (stewstim ) and external.. so you can test each sensor/input while keeping everything else stable

in this setup, only the tach input is swapped to external... so is OFF on left switch block, and ON onright boack, then going threw screw terminal

[oldcorollas]

umm.. so i'm assuming that since the ignitor is dumb, and from the earlier scloping, the start of the dwell on the ignitor matches the zero-crossing of the VR... then using the falling edge trigger will be most stable?

ie, with the way it'S timed above, the coil fired at the zero crossing = most reliable timing method?

[thechuckster]

Stu: run the MSextra code for your board/ECU. Heaps more features than the B&G code. Would also suggest getting Megatunix on the laptop as it too has a good trigger logger.

Also - what is the timing map for your test ECU? For the above assumption to work you'd need all the cells to be 0, the cranking and post-start to be 0 and the trigger angle to be 0 - which the ECU really doesn't want to see (as you know MS prefers the signal to come in fairly early (e.g. ~50BTDC). It needs the trigger coming before it thinks about charging the coil and firing it.

Also 2 - I would suggest you disconnect trigger input testing with spark output testing and treat them as 2 isolated things to get right.

fwiw: I drove the VB921 via one of the board LEDS (can't remember which but can pull the box apart if needed) and the ECU was set for an inverted output.

[oldcorollas]

i decided i don't need the features of the extra.. the base B&G code is enough.
(edit: this is just for the shitbox K, and was running ok on base B&G MS1 code before only upgrading to control spark..)
just fuel, spark, thermofan and idle up..
hmm, trigger logger would be easier than hooking up to scope

timing map is 0, trigger offset is 0, is above cranking rpm, and after afterstart.

yeah, the timing of the trigger is of concern. doesn't it just go to calculating the next spark rather than the current one?
i can advance the trigger via moving the vac advance plate in the dizy, but not sure how much yet.
will it be really unhappy working out next spark instead of this spark??

i'll set up the dizzy so the rotor is centred at maybe 25 or 30deg advance, and then vac advance unit should be able to move it another 15-20 maybe?

i tested the VR, then the VR and toyota ignitor, then the toyota ignitor and BIP output... the timing of the last tw was of interest in working out which edges etc... i would've thought?, otherwise you have no time reference of what is doing what?

yup, is looking like inverted is the way to go.. was just my brain getting mixed up

edit: here talks abotu setting up the trigger at a good cranking advance...
http://www.megamanual.com/ms2/configure.htm

[thechuckster]

Ah, i keep thinking older ECU firmware - which needed the trigger way ahead of the spark event. No idea about the MS2 B& G code but when in doubt, follow the MSextra manual WRT setting up a VR dizzi (and I'll go there now to have a look). I never bothered with the toyota igniter - was easier to include a tacho booster circuit in the box with the VR/ignition/fan/shift-light circuitry.

dizzi: I removed the vacuum advance but left the connecting rod in place - then rotated the contact/sensor plate to an approximate position and used a long grub screw (in place of the short screw that attached the vacuum can to the dizzi) to lock that connecting rod. Once i was happy with it, I encased that bit of the dizzi with h/d epoxy.

fwiw, the rotor will cope with highly advanced trigger angles giving you a bit more flex.


edit: from some searching 'Below 20, MSnS fires ignition in "next cylinder" mode' - need to confirm whether this applies to B&G code.

[knightrous]

In anticipation of putting my MS2 onto my 4AGE 20V in a couple of weeks, I decided to splash out and buy a Dual VR 2.0 from JBPerf.

Saves me having to reconfigure my MS2 for VR and having to build a second VR circuit.

Next up on my list is to purchase a MS3X board to run with a custom MS2X adaptor to connect it to the MS2 (Not MS3)

[oldcorollas]

one of these...
http://www.racepak.com/Components/EF...e_newsletter18

connects your megasquirt directly to one of these? cool
http://www.racepak.com/Instrumentation/IQ3_Display.php


i'm not likely to get either, but very nice

[oldcorollas]

was the freekin idiot that filled my DB37 with hotmelt glue?!?!!?!!?!!!!!!


oh, that'd be me FFS


edit: then rip off the bottom plug tiny piece by tiny pice.. in the dark.. to find all the spare wires loomed up... and that the fuirst 3 feet of loom are totally detachable....
thats good but....... why didn7t i see that earlier

[MWP]

was the freekin idiot that filled my DB37 with hotmelt glue?!?!!?!!?!!!!!!

Hah. That kinda thing always seem like a good idea at the time