VEMS Community Forum

Hi All.

Could someone please help me with the the ign output on the subaru?

the firing order is 1-3-2-4, and it cops fired through a external ignitor.

I can see that on the Wiki page its recommended to run wasted spark, that would mean that i have to run duelout , right?

Or can someone make me the config or megatune setting for it?
this is how the wiring is made:
i have made the the first four ignition outputs to logic level,
connected like this:

COIL 1 = EC36 PIN35 = ING 0
COIL 2 = EC36 PIN33 = ING 1
COIL 3 = EC36 PIN34 = ING 2
COIL 4 = EC36 PIN36 = ING 3


This is a scop trace from the other day.

(http://img.photobucket.com/albums/v41/Denmark/Subarutriggerpicture.jpg)


Thanks,
Skassa

Quote from: Denmark on April 16, 2008, 10:59:33 AM
the firing order is 1-3-2-4, and it cops fired through a external ignitor.

COIL 1 = EC36 PIN35 = ING 0
COIL 2 = EC36 PIN33 = ING 1
COIL 3 = EC36 PIN34 = ING 2
COIL 4 = EC36 PIN36 = ING 3

Wasted spark pairing would be 1&2, 3&4 so you have the wires connected correctly there.
Just select Settings->Ignition Settings and set Dual Output Mode to Enabled.  And Alternate banks from h[0] to 01..00
Then in Settings->Ignition Outputs Dual set
0 to 2+3 (EC36 pin 34+36)
1 to 0+1 (EC36 pin 35+33)

Thanks,

I just had the coil on cylinder 1 out,so i could adjust the tdc delay, but there is no spark at the coil,
So i tryed to measure the 5v direct at the ecu, but i cant measure anything,
But that is probbely just becourse it such a short pulse, that i DVM cant measure it,

So i just think i give up, using the logic level outputs, as i dont think they can fire the ignitor(was that the problem you have had on a Nissan, Rob?)

Tomorrow i will try the stepper output instead, hoping that  will work..

All info would be great :)


Thanks,
Skassa

That was the problem I had on the Nissan - and it's the same ignitor on both the Nissan and Scoob.

You can get around the problem using the stepper driver as the coil driver, although I am not doing this because I am thinking of a VEMS on my MY99 Scoob which needs the stepper driver to drive the stepper.

Thanks for your answer, i also plan to install it into some V5-V6,so i have a friend who have made this setup to make the logic level work better,
Can some of you please rewiew this.Ans please make a comment, as iÃ,´m going to try tomorrow, if there is no other clever or smarter ways of doing this,



(http://img.photobucket.com/albums/v41/Denmark/highercurrentlogiclevel.jpg)

Hilly, has already done this on his Audi engined Lotus:
(http://i67.photobucket.com/albums/h286/s3esprit/IMG_0731750x562.jpg)
I don't see any mention of the buffer chips, I guess you could in theory use a stepper chip to do the same work.

Rob

We did try the above, but it did not work, then we found out that there is no signal comming from the 74HC259 chip,so we canÃ,´t get it working, there is power supply, but not gate signal..

Hmm,
Maybe i need to drive the Ring with oe ecu, but that would be a disaster, as there is no power, and  the oe lowend mapping is really lacking, fuel and ignition :(


/Skassa

I will later this evening, try to use the stepper output to see if i can get a spark..

well right now i have just had it with VEMS >:(.

The power supply is constant?  Have you tried inverting the signal?

The power supply is constant,

What is the H2table supposed to look like , if i use A-B-C-D from the stepper chip,

And Cyl1 is A,
cyl2 is B
Cyl3 is C
Cyl4 is D,

ItÃ,´s duelout h0 = 01-00
IgN Output in megatune :
00=stepper C+D
01=stepper A+B

Waterpump is activated with S259CH7


Thanks,
Skassa   

All that looks correct to me - still having no luck?

When we tested the Nissan igniters (the Mitsubishi DIS4 IIRC) which is the same igniter pack as you have.  It looked as if we'd be able to fire two channels from one stepper output, which would give wasted spark and leave two stepper channels free.

No itÃ,´s not working,

But when i download the config and table, the H2 table looks strange, something like :
H2 37 27 07 17 7f 7f 7f 7f,

So i think i manuel have to plug the right numbers in, before uploading it again,

But when using all 4stepper outputs like like writen above, isnÃ,´t it supposed to look like:
H2 20 10 00 00 00 00 00 00, when running wastedspark?

Thanks,
Skassa

Using 20 10 ... in H2 would drive the IGBT outputs (or logic levels in this case)
I'll check a working H2 this evening, take a look in the wiki to see if you can find anything of use on the Stepper outputs.

ItÃ,´s really strange,
later today, we will measure the the input to the stepperchip, but i actully think, that there now is something wrong with the ATmega,
Have there ever been a board were all things work, but no ignition outputs.

Otherwise we will try to put a 36-1wheel on a drill, and the use that input to see if that can make a output.

I have tryed both the 4 logiclevel the 4 igniton fets, and the stepper , but there is no output


Any other suggestions?


Thanks,
Skassa

If you are getting a stable RPM then you should be getting an ignition pulse.  Are you scoping the output to see if there is an ignition pulse?

The Rpm is stable, and itÃ,´s cranking around 220rpm, and yes it is a scop, that we use to measure the output,
But there is no output at all,

This evening we mounted a 36-1 ford wheel on a drill, and held a ford vr sensor in the hand, to see if this could work, as this is a known trigger setup,
We could again get a stable input rpm, and the injectors are clicking, but there is no ign output..
We scoped both the stepper output, and the 4 logic level, and the 4 ignition fets,

So right now, iÃ,´m really pissed, and very close to put the vems print in the garbage can, and buy a working setup , Apexi power fc og some simple ecu,
I can see that even megasquirt got the subaru trigger working, off the shell.

IÃ,´m out of idears to make it work, other then the board is broken..

When you set it to work on the logic level outputs do you set the H2 table in the configs?

You've had the VEMS working fine on other cars.

Send me your config and tables and I'll put it on my test bench.

Have you installed the 5V supply wire to the stepper ic. If not read this build instructions (step 2) green wire
http://www.vems.hu/files/MembersPage/PeterJensen/wems_hardwareinstall.doc

Regards Peter

Quote from: peter_jensen on April 21, 2008, 11:26:32 PM
Have you installed the 5V supply wire to the stepper ic. If not read this build instructions (step 2) green wire
http://www.vems.hu/files/MembersPage/PeterJensen/wems_hardwareinstall.doc

Regards Peter

Yep, this is not the problem,
becourse there is also no output to the ign fets, and yep i also set the H2 table , when trying the Ignition outputs.

Quote from: [email protected] on April 21, 2008, 10:13:57 PM
When you set it to work on the logic level outputs do you set the H2 table in the configs?

You've had the VEMS working fine on other cars.

Send me your config and tables and I'll put it on my test bench.

Okay, do i set the subaru trigger, or the 36-1 trigger, for you to try?

Thanks,
Skassa

There is nothing wrong with the triggering from what I can see, if the RPM is stable then there should be ignition events.
So when you send it over just tell me what triggering you've got it setup for.

Okay,

My friend is today making a trigger generator, to make a trigger imput , and that will be 36-1, so i will make a config for 36-1,
But it will not be before tonight, as i have to work to make, to make a living,  ;).
But will give it ago again after work...

Strange strange, there must be something wrong with the board, but we will see..

Why not use the SignalGenerator program in the tools folder, and use your soundcard?

Thats what I do, it gives good RPM sweeps.

What are your H2[] entries looking like?

I have to work to make a living too you know, thats never stopped me fault finding though ;)

Its possible that there is something up with the board - although that would be a very rare occurence, it is far more likely to be something wrong with the configuration or output tables.
Are you getting an activity from the injector FET drivers?

Rob

Yes the injector fet`s  fires, as they should

Peter

Well,

Last night we got it working on the bench, Peter jensen came over with another board to try,
As we had 2 boards were there were no igntion outputs,
We then tryed the new board with the 1.0.38 that was already in it, and it worked perfect,
We then on that board uploaded the 1.0.73 firmware and got the working, by changing the H2 table in the config,
We then uploaded the same config to the 2 boards that was not working, and it was working again  :)

So there have been some kind error in the config file,
The only change, that we did not do on this new board, was to write the long string,regarding the h2 table, that is written on the Wikiside, that have to be interred, that was interred on both the none working boards,
So we suppect that this is the course to the problem....

So DON`T write the string....

Tonight i dont get the time to put it in the car, as i need to mount new brakes, and some AGX shocks all around

Thanks,
Skassa

Quote from: dnb on April 19, 2008, 08:09:15 PM
That was the problem I had on the Nissan - and it's the same ignitor on both the Nissan and Scoob.

You can get around the problem using the stepper driver as the coil driver, although I am not doing this because I am thinking of a VEMS on my MY99 Scoob which needs the stepper driver to drive the stepper.

We measured the output voltage of the stepper chip,and the input is 5v , but the output is 12v,

so i wold like to know if you just connected it tothe ignitor or if you changed the output to 5v, as i dont belive that the ignitor will like to get 12v, or do any of you know better?

Thanks,
Skassa

The Hitachi DIS4-04 transistor (igniter) pack does not seem to operate at 5V, we're using the stepper chip to fire off the Nissan S14 and S14a coils which uses the same igniter pack as the Subaru. 
The tests I did showed something like a 70mA current draw, and the coils started to fire happily.

Rob

Hmm,

IÃ,´m sure that we scoped the volt signal from the oe ecu to be 5v, ím just not sure what model ignitor thatÃ,´s on my car, i have just had a quick look at it, and itÃ,´s a alu one, were i have seen some have a plastic one.

put if itÃ,´s a 5volt signal, cant we just also change the supply voltage to the stepper chip to 5volt?

Here is a complete manual over the Legacy turbo, but i can open it here at work, itÃ,´s section 6.1, were this is shown, if i remeber correctly.
http://www.rslibertyclub.org/forums/showthread.php?t=70542


/Skassa

Take a 5v feed and repeatedly touch it to one of the control lines on the coil pack - if you get a spark then 5v with a good current will trigger the pack and you'll need to look at ways of doing that, maybe the stepper chip or maybe some of those driver chips that Hilly used.
The Nissan S13 had all aluminium igniter packs - these fired off from the IGBT driver chip.  Have you tried the IGBT driver chip now that you have a signal?

I will give it ago again tomorrow or saturday, as i dont have the time sooner,
I have not tried anything since we got i working on the bench the other night

I then hope the car will start ::)


Thanks,
Skassa

It is a 5volt siganal, that is comming from the oe ecu, so if using the stepper output, this have to be modified to output 5volt.

I think we will try the ignitor outputs again.

Thanks,
Skassa

Well,

I got the car started today, first running on 2 cylinderes, as the outputs were wrong!, itÃ,´s now run on the stepper output,And the B and C is switched(so runningA+C, and B+D) duelout,
and the suply to the stepper is changed to 5volt,
we did this after we had some trooble with the logic level outputs,misfire,
But We then found out that there is something wrong with the firmware, that is 1.0.73, when itÃ,´s running the Subaru trigger setup, it took a long time to figuere that out,
And that is as soon as the fuelpump is switch on in megatune, running p259 ch5, then there is something wrong with both the input and output signal(trigger in and ign out)


I did not have the time to go mapping today, as i had to had the cars alliginment done today, as this was the only day i can,before next week Ring trip,
So i had to plug in the oe ecu.

I does seem that i have no control over the IAC valve, even though i can hear it clinking, when i turn it on,
I cant regulate up og dawn at all, it just idleÃ,´s at around 1600rpm.
I have duelvalve , Inj6 EC36-pin6 as close end, and inj7 EC36-pin17 s open,
that is right, right?
With the same setting as the 20v AE86, there is no control at all.


/Skassa

I thought you had the dual IACV working nicely on the 20V?

Obviously its going to need a lot of tuning, so surely getting it working on On/Off type will be the way ahead.

I've never heard of the fuel pump causing problems with the trigger - has Peter Jen put anything on the wiki about it?

Rob

Yep,

I did have a perfect idle on the 20v,
therefor a think itÃ,´s a bit strange that i cant control it atall on the subaru, and i dont know were to start with the adjustments, so i will map it, without the idle control first, then i must see if i can get it working.

Peter told me to write it on the Wiki, but i ave just not come aroud to doing it yet,

Btw, we measured the oe dwell time with the scop, so we now know that this i 3ms :)
I have now scop pics of the oe tacho output ,cam and crank sensor, and the output for the oe ignitor,
This is usefull, if you need it, some time, when this get working ::).

Oh wel, i have lots of time, itÃ,´s only saturday morning, iÃ,´m going the 800km to the ring, 3days fun, and then 800km home again ;D


Thanks,
Skassa

Excellent work, have fun at the ring.
A scope trace of the VR signals when the fuel pump turns on would be very useful.

I dont think We saved the scop picture, of that trigger event, but it was just random, and not the same all the time, there could be 2 event so close, that you could only see it as a wider signal, then it could mis the input/outputs a few cycles, then again some random firing.

Now i just need to get it mapped right, and get the boostcontrol working, then the other can wait, til later,

Boost is the most importen thing ;D


I have writen a bit on the wiki, and updated the config and table

/Skassa

Well, i was out and drive a bit last night, and adjusted a bit on the fuel and ign map,

The car is having a hard time starting, and trigger error shows up when cranking,but it humps like cracy, but ones it starts itÃ,´s fine,
just like the 20v engine when running the oe trigger, so itÃ,´s probbely noise, itÃ,´s just not present, when we scoped it the other day.

I have some trobble withthe idle, as it runs 1300-1500rpm, and will stall when the fan kicks in, this i cured with setteing a much higher ref table, so when 71degrees it says 60 procent, the it would not stall,
but i dont suppose this is the right way,
Also a bit difficult to drive when there is no engine brake, becourse the idle is holding the rpm up!


/Skassa

Put a ramp in the ignition timing at the lowest RPM bin (probably 600 or something) so that idle falls into a V shape.  As RPM drops, the timing will rise and RPM will build again.   This should help stop the stalls, but it is only a temporary fix.

If you want another set of eyes on your datalogs, then let me know :)

David's idea is a sound one, which I always use when running a car for the first time.
As you smooth out the map you may find that the idle will smooth out - it will also get smoother with the correct injector sequence as now you may find that theres a lot of unburnt fuel passing out of the exhaust, Cliff demonstrated the effect with a peice of white paper - if you hold the paper 10cm from the exhaust and its covered in wet dots then you are not lighting off the charge properly, this can make you chase your tail slightly because the wideband wont read the unburnt fuel so it may show a lean mixture, which you'll add more fuel to and end up with filthy stinking gases coming out the back of your car...

To get over this you need to address the injector phasing as long as you keep the sequence correct you can try them in four different positions:

0 -> 8 4 2 1
1 -> 4 2 1 8
2 -> 2 1 8 4
3 -> 1 8 4 2

Today we measured with a scop ona my99 Impreza, to get some data , other then from a manual,
And some strange thing showed.

The ign output from the oe ecu, to the coil/ignitor  on the wastedspark coil , is 12v outputs, so i will use the ign fets(inverted) with a 12v pullup to drive the 2channels,
We also measured dwell, and there was a big surprise, as when cranking there is 58ms dwell, when on idle there is 8ms dwell, and at 4k rpm there is 4ms dwell,

The rpm output needs a 12v pullup, just like my own Legacy,

That was what i did today regarding the Impreza V5 install.


Any comments to that info?


/Skassa

I'd not be too worried about the dwell on the ecu to ignitor signal because it's probably not the same as the dwell applied by the ignitor itself.   All the same, it's a bit wierd when we are used to having fairly constant dwell systems these days...

Good that you're making progress.  I have used the ign fets to drive a MY96 ignitor on mine - not ideal, but it works on the bench...  Still not got the car finished because I'm still waiting for the the mechanical chap to nail the rest of the car together :(

Okay,

On the my96 that is a extern ignitor right?

On the my99 there is no external ignitor, that is built into the coil.

here is a great side with online manuels

http://www.cannell.co.uk/Manuals.htm

The My99 GT fit the diagram of the P1, with minor changes. there is can be seen how the coil is driven from the ecu.

also great to being able to see the idle stepper connections.

The cranking voltage on the cam and cas is, 4v peak-peak,
so the same as my old Legacy, i just hope it dosent have the same noise, but i hope to being able to tell during the next week..


/Skassa

Yes MY96 is an external ignitor.  It's only being left for ease of wiring.  I will remove it the second it causes trouble... I know the internal IGBTs work well. 

Nice link :)  Will be useful for my MY99 RA.

Yep, that link and the this wiring digram, is whatÃ,´s needed for the v5-v6 impreza

http://www.linkecu.com/support/dlandswupdates/Downloads/MY99Manual



But i just hope it runs with the legacy settings, with minor changes,
I really just hope that the trigger noise is not present, so i dont have to use all that time to figure it out on this car, as this is a customers car!


/Skassa

Had the car started today, run right away on the legacy settings,with the ign outputs inverted and with a pullup,
Also got perfect idle, like oe, with the stepper on the difficult Impreza stepper, and it seemÃ,´s to start much better then the Legacy, i just need to open the throttle a bit and it starts right away.

I have to work some on that this following week, as i also have to install some bigger injectors, so i did not want to put a lot of work into it before they are changed.



/Skassa