Ms3 with ms3x k24a2 engine (from acura tsx)

[Barton]

Hello, I'm trying to start a k24 engine with a ms3x, I get good spark and injection secuence (fully secuencial), the injectors and coils are working fine, my problem is I cannot get the timing right, the tsx/d17 code only let me to modify the offset to +-20 degrees, but I need at least +60 degrees more to get the fixed timing value to match de crank marks.


the MS3 recognizes the 2 position signals (ckp and cmp) just fine, the rpm graph is smooth when cranking (see the attached files msq, datalog etc etc), I have installed a several megasquirts, sometimes when I can't see the toothed wheel (like this case) I just change the tooth #1 value until I see the crank pulley mark, in this case I need more offset to get to match the timing light with the marks.

Cam sensor I'm using:


Videos:

offset 20, fixed to 0, timing light 0 (here you can see the timing mark just on top of the crank pulley): http://www.youtube.com/watch?v=RPEnOAwO5xY

offset -20, fixed to 0, timing light 0: http://www.youtube.com/watch?v=jE7Ebuvi3Rc

Spits flames: http://www.youtube.com/watch?v=02_96AAIdz0

I've been chating with dfzuntor (he has a k20 engine running with megasquirt) , but there is something we are missing.


Thanks for your help.

[jsmcortina]

What's the other sensor to the left?

The TSX mode is coded based on user supplied data and as far as I'm aware they are running it fine.

James

[Barton]

the left sensor is the vtc sensor, I think that one has a different pattern because if a plug it the rpm does not behave right.

The MS recognizes the triggers perfectly but it need just to add more offset range for the crank pulley marks to match the timing light, the user dfzuntor told me he offset uses +19 but even in +20 the light is far from the 0 mark.

the engine has the same compression in all cylinders, I tested the injectors and the coils even de secuence and they are ok! but the spark is retarded and I cannot fix it.

If you have any idea just let me know.


thanks

[billr]

Are you using a simple timing-light, one that flashes just as the spark occurs, or is it a "dial-back" type?

[Barton]

I'm using a innova 3568+, but the engine behave perfectly, I mean the spark is retarded and the engine spits flames like it should.

this one: http://www.equus.com/Product/3568/INNOV ... ming-Light

[billr]

That is a "dial-back" type, they can give false readings if the settings (on the light) aren't used correctly. I'm not saying that absolutely is your problem, but you might try testing with a simple light. At least, check operation of the dial-back on another engine with known timing marks.

[Barton]

thanks but I'm sure it is not the problem, 2 days ago I use the timing light in a mustang with a megasquirt and worked fine.

I pulled out this picture from the video: +20 offset, 0 fixed timing, 0 in the timing light.

[Barton]

hello, some help???

[jsmcortina]

So, aside from the offset, everthing else is identical?

How many degrees is that off? Another 15-20?

James

[Barton]

Yes, well the spark timing is perfect, if fix it to any degrees the timing changes and stays at that value like any other ignition pattern, the offset is at least 20 degrees off.

Thank you very much.

[jsmcortina]

Can you confirm to within 5 deg how far off and which direction. i.e. What offset number would you like to enter?

James

[Barton]

i just meassure and I saw 25 to 30 degrees out of fase (retarded) I need to be able to use +45 to +50 degrees of offset.

Thanks for your help james.

[jsmcortina]

Test code sent by email.

James

[jsmcortina]

I'll look to add a new wheel mode to the next alpha code with the 40 degree offset built in.

James

[Barton]

Thanks James, the engine started smoothly I used -5 of offset.

[neslider]

Basic question, but pardon my ignorance because my previous experience is with Porsche CIS systems. I have a new project utilizing a K24A2 from an 04 Honda Accord Coupe to be retrofitted in a stripped down 911 Coupe chassis with the stock 5 speed transaxle. Which Megasquirt version (1,2,3) is "required" for the engine to run stock. No turbo, no supercharger, just box stock. I can't utilize the Honda ECU because it's for an Automatic and I'm trying to keep the cost down and include the minimum amount of Honda harness and electronics out to make it clean like stowing the immobilizer unit and having to use the Honda ignition and key. I need for this to function, basically as a stand alone unit in the chassis. Any help is appreciated.

[Barton]

neslinder: hello, you need to use the ms3 + ms3x to drive everything COP, injectors, vtec, vvt, etc etc.

well here are a few logs from the car, I getting a few problems,

1) when the engine starts, is necessary to keep it runing 10 to 15 seconds at idle because if you rev it it losses sync, but after those 10 to 15 seconds it works perfectly.
2) I cannot use COP mode, the engine does not start, I've tried a lot of things it just simply does not work. (I'm using wasted cop mode).
3) We made a few mods to the engine so the vvt worked perfectly using the maximum angle and the engine made a lot of power there, but a few days ago we change the vvt pulley (the oem pulley is for 30° and the civic si is for 50°), but the angle changes just before the vtec engages (5500rpm) and the engine losses a lot of power there.
4) the vvt system freezes at 7k rpm.

ms3+ms3x, firmware 1.3.2

[jsmcortina]

Yes, I can see the VVT1 spikes there in the log.

I've checked through my records (a bit) and I don't seem to have clear data on what the VVT cam pulses look like. Do you have data on that?

Do you have the registered version of TunerStudio? If yes, please record an "Engine Log" as that will capture the crank, cam1, cam2.

Please also try the current pre-1.4 alpha code. Some changes were made to VVT angle handling along the way that might change the behaviour you are seeing.

James

[jsmcortina]

Any updates?

James

[Barton]

Hello James I Finally could make the datalogs, I sent them to your email.