Megasquirt users thread to discuss all things Squirty

[old_mr2]

Toyo Al, I didnt know you were a MS user! I may pick your brain sometime in future (if I still cant get an RPM signal with my current plan).

PS I still havent returned your engine crane, I keep forgetting! BTW your motor Sounds hot untuned.

[thechuckster]

stu:to run the dizzi, simply: mount dizzi in a vice, fit 8mm drill into chuck, slide rubber hose (with ID 8mm) over drill, push rubber hose (side to side) up against dizzi shaft, run drill. Saves having to fuck about with centering the shaft into the drill.

Also, if you get the VR polarity the wrong way around, the ECU usually tracks the inverted signal until about 2.5k rpm, then the pulses are coming so fast that the zero cross wave is correctly tracked and your timing signal then jumps back by about 1/4 to 1/2 teeth (because you setup your trigger timing at idle speeds right )

Obvious diagnosis is via:
disable any accel enrichment
put strobe on crank pulley
slowly accelerate motor from idle to 4k rpm
timing should slowly advance as per map, then at the transition point, drop back a few degrees before continuing to advance.

The drop in timing won't be visible in the ECU logs (tho you might see some subtle MAP change).

[oldcorollas]

that would have been a better way to do it
instead i had dizzy sittin on my leg, with chuck wound in so one of the three bits was turning one of the flat parts of oil drive... sorta worked
will use tube idea next time

I will check that timing on the motor when it runs again (ie, when i pull finger)..
is that why stock toyota ECU's have a spaz at medium rpm? ie it starts having trouble determining which is which, and then rpm appears to fluctuate, and it spits internal errors and temporarily interrupts spark or fuel?

i shoudl work on new wiring tonight... and get some coolant.....

[3jcelica]

The way I usally test the vr signal is wired correctly is by trial and error (typically if you have it wrong the whole static timming setup to get the car started doesn't make sense), wire it in, rev engine to 3000rpm if you hit a brick wall with the revs you know you got it wrong, scope would be nice though. Also check out the tooth logger in Tuner Studio this can help you determine if you have it wired correclty.

[thechuckster]

I borrowed wilbo's CRO and tested it off car. Had read plenty of stories on the MS forums about folks wasting days of diagnosis and tuning trying to eliminate that ugly transition. Decided to avoid the problem all together.

No idea if it's related to a stock ECU problem - Toyota have a long history of using VR sensors so something around middle of rpm band I'd put down to a TVIS or similar induction tool having a problem?

[oldcorollas]

yah, thats why i decided to use CRO as well (have the old dual trace at home.. pictured above)


next mission is to work out how to plug ignition up... going to use Fidle pin now and want another free for thermofan.. could use 36 for that, and then run two pins each for coil and ground for BIP... but then neeed to get them to other side of board..

i suppose i could use a PCB stake in say, X11 and X14, as the stake would be able to run the current?
or... have cord going to a mic plug attached to bracket on end plate (too big to fit in end plate, unless i move injector IRFIZs...
hmmm
but the loom DB37 is annoyingly hooked up as it is
bracket is easy but messy..

[thechuckster]

suggestion? if you were using the older board attach the IGBT's to the lid, run flat flex loom up them (forming a hinge of sorts) then fit a mike plug to the the lid. I ended up running all low current stuff out of my original ECU to another case with the IGBT, fan controller, VR interface and shift light drivers.

If using the newer board with IGT driver along that back heatsick section, then another DB9 or maybe DB15 plug on the end plate and use multiple pins to carry higher current. At best you'll be sinking 5-6 amps with a coil but if using 2 pins for each coil you start to run out of pins on the original DB36 plug - or have to start disconnecting all the unused (and currently earthed) pins) from the board.

Lack of pin capacity is probably the only major engineering drawback for megasquirt.

[knightrous]

Lack of pin capacity is probably the only major engineering drawback for megasquirt.

Upgrade to MS3X then

[old_mr2]

Mechanically, your 4AGZE dizzy is the same as the 20V dizzy, it's just the 20V dizzy has the G wheel below the 24 tooth NE wheel.
From what I can gleam from the DIYPNP documentation and the links I provide in my last post, all you should have to do to make this work for you is the following:



Wire NE to VR IN +
Wire G- to VR IN -
Wire G2 to VR2 IN +

All the VR sensors in the toyota dizzy's share a common ground, which is G- in this case. After that, you should be able to jump into MegaTune/TunerStudio and setup your cam wheels as follows:

Well I have an rpm signal at last! Thanks Aaron! I am suspiscious of it though. With the drill on max speed it only registers 91 rpm in Tunerstudio. And the drill is capable of 1400rpm~ according to google...

Something is not right. BUT, if I spin it hard as possible with my hand I see over 200rpm in Tuner studio??? Not sure whats going on there...

I also noticed that I get the same results if I switch NE and G2 which also confuses me...

[3jcelica]

If using tuner stuido, go to the toother logger and see if it is counting all of the teeth, and you are getting a home signal (missing tooth or second sensor)

Roger

[oldcorollas]

getting there.. wanna solder today...

trying to work out if i need a resistor for the BIP or not.

if you put the BIP` onto one of the spare ports of the processor, then you need a 330ohm resistor, to limit current(?), and that was is shown at Msextra site and DIY (same same)

looking at the MS2 schematic, the IGN out already has a 1K resistor between the processor and the DIP20 socket (R9 here, before pin 17 http://www.megamanual.com/ms2/hardware.htm )....
so does that mean i can connect the BIP directly to pin 17 (JS10 = JP1-5) without the 330ohm resistor, as there is a 1K already there?
the V3 schematic doesn't have a resistor between IGN and IgbtIn, which goes direct tothe VB921.. so ti should be ok with no resistor?



next.. fast idle on V2.2 has a 1K resistor, driving a ZTX450 (swapped from 2n2222 for the extra current)..
so to add another ZTX450 to a spare processor output, i need to use a 1K resistor too? (99% sure on this one )

but.. do i need a diode in there? (ie like D7 and D9 in V2.2 schematic).. are they there in case power gets hooke dup wrong, or to stop floating or or??

the diode is also shown here for a "typical spare port output" circuit
http://www.megamanual.com/ms2/spare.htm

[old_mr2]

If using tuner stuido, go to the toother logger and see if it is counting all of the teeth, and you are getting a home signal (missing tooth or second sensor)

Roger

Im new so bear with me. The only way I could get a 'believable' rpm was to set the trigger to basic trigger instead of dual wheel. I then saw 1000rpm~.

When set back to dual wheel I would never see more than 300-400 rpm. And I could tell the drill was easily doing 1000rpm or so.

I also noticed that pulse input 1 and pulse input 2 mirrored each other when spinning the dizzy.

I had a go at the tooth logger, but forgot to do a screen shot and have packed everything away now... I saved the log as a csv file, but it is unreadable... Ill continue fiddling with it tomorrow..

[big_zop]

I know you are driving the cam sensor directly but does the fact the cam sensor only see half of what the actual engine rpm is make any difference?

[3jcelica]

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

[thechuckster]

I also noticed that pulse input 1 and pulse input 2 mirrored each other when spinning the dizzy.

sounds like you've connected to the two single-tooth sensors. You should be getting 24 pulses per rev. Try another output from the dizzi for input 1