My car runs again :D (Some megasquirt stuff & informatio

Website · Post by **antiekeradio** » 22 Mar 2006 05:21 pm

some oscillation between the MS and the O2 sensor, then...

perhaps something to do with gas flow in the downpipe. where have you put the O2 sensor?

think WB O2 sensor would totally cure your problem,anyway..

Post by **jtbo** » 22 Mar 2006 05:39 pm

antiekeradio wrote:some oscillation between the MS and the O2 sensor, then...

perhaps something to do with gas flow in the downpipe. where have you put the O2 sensor?

think WB O2 sensor would totally cure your problem,anyway..

My O2 sensor is in exhaust manifold, right where it is from factory in B200F, so should be very good place.

Wideband would be naturally so much better, but that looks to be very far away with my current budget.

Adjusting mixture manually and making lambda correction to be active only from around 2000rpm makes problem go away, but still need to tune low rpm bit rich, I'm not sure but maybe this kills my city mpg.

I tested now 4700ohm resistor in place of that 390ohm resistor and that did not any good, I think that if that made any difference then it would be to wrong direction.

Next step is shielded wires and after that I will rebuild my circuit, if it still does not work, then I believe that it is something much more complicated.

Website · Post by **antiekeradio** » 22 Mar 2006 08:46 pm

unfortunately, technical considerations are not compatible with financial ones, then...

same reason why my DAF still looks like a 'poffertjespan'
(have a laugh, google that word)

Greetings Wouter

Post by **jtbo** » 22 Mar 2006 09:22 pm

antiekeradio wrote:unfortunately, technical considerations are not compatible with financial ones, then...

same reason why my DAF still looks like a 'poffertjespan'
(have a laugh, google that word)

Greetings Wouter

I got few places in my car too that would need a bit of hammer and plastic padding at least, but I have not started to do anything to those as I know that it usually will get just worse when I try to do it myself.

Also this brown thing, how to stop that, grrr.
http://www.janiervast.com/kuvat/volvote ... G_3321.JPG

Post by **pettaw** » 23 Mar 2006 01:08 am

Jani I've got a wideband for my car and have therefore got a fairly close fuel map. Mine kangaroos at light throttle anywhere between 1300-1700rpm light throttle and I can't see a fuelling cause for it, so I'm guessing its ignition related.

I know my engine is slightly different to yours cos its the high compression one, but I can send you my fuel map if you want it. Might be a better starting point for your tuning than a generated map which was way lean in places and stupidly rich in others for me anyway.

Give me a shout back if you want it cos I've got to transfer the files over from my laptop to the home computer.

Post by **pettaw** » 23 Mar 2006 01:21 am

You're right about the idle BTW. Can get very smooth idle at 12.5-13.0 AFR, but at 5.5%CO would never pass an MOT. Just about idles OK at 14.7, but anything much less idles like crap. I try and set idle AFR to 14.5-14.7.

Your kangarooing is caused by AFAIK the MS hunting between timing bins, so I would try and set the timing bins roung that area to be more similar to each other so if/when it jumps around it doesn't make so much difference. Do the same with the fuelling. I'm gonna play around with the actual values the tables are at so there's less gap and interpolation in that area cos I believe that's a big cause too. Resolution round that area isn't good enough.

Trouble is finding the time....

Post by **jtbo** » 23 Mar 2006 11:27 am

Actually I have made ignition map so that it is same with 1200rpm and 1600rpm and that is not helping, setting timing to 10-30 BTDC on problem area does not help, only thing that helps is to put it enough rich, bit same as with idle.

Idle advance I have set now to 17 BTDC, seems to give smoothest idle.

Advance starts to jump because engine tries to stop when this kangaroo jumping starts, then engine is trying to run, trying to stop, very fast looping, this all is related to mixture going bit lean, 14.7:1 is on the edge and I believe even 15:1 would cause this kangaroo thing.

Yesterday I did test a bit without lambda correction on low rpm and it was better, now just need to start slowly to lean mixture on low rpm area, need to do that manually and same time need to drive a car

These are still under development, started from scratch at sunday. Igntion map is made from renix map and then modified, generally engine seem to like bigger advance better.

When I get around to make accel enrichments to work I think I will get away several of these leaning on throttle situations. Currently there is just something weird, every 30 seconds it floods engine if there is any accel enrichments chosen and there is not even single value in logs that would trigger enrichments. However I did read that there is bug in latest extra that could do something like that in certain situations, but setting enrichments to decay should cure it.

Naturally I would like to look how your VE table looks, I think it is always nice to compare, but I don't know if it is how useful as engines may be quite different.

From megasquirt manual or site I did read that only few rare engines will idle properly at stoich (14.7:1) usually idle need to be somewhat richer, for Volvo engines 2,5%CO is ok value, here mot will fail if it is over 3%, but that might not be problem as I may need to switch back to carb for mot or then I will change car to older which I find quite tempting because opening possibility for turbo route.

Post by **pettaw** » 23 Mar 2006 04:25 pm

I'll have to try that on mine.

Post by **pettaw** » 23 Mar 2006 04:57 pm

here's my msq file.
Msq file

and here's my VE bins separately.
VE bins

Just in case you can't read the msq file, I'm using MT v2.25b723 and Msns-e v029d

Post by **jtbo** » 23 Mar 2006 06:05 pm

pettaw wrote:here's my msq file.
Msq file

and here's my VE bins separately.
VE bins

Just in case you can't read the msq file, I'm using MT v2.25b723 and Msns-e v029d

Thx from files, I did noticed that on constants page you had PWM time treshold 25.4, I have been told that it should be 25.5, I don't know if that makes lot of difference, though.

I see that you have quite bit larger reqfuel, I have only 15.2, maybe that is because I have no vacuum hose to fuel pressure regulator so my fuel pressure is constant 2,5bar. I felt it is easier to adjust fuel map when there is less changing variables, ie. fuel pressure.

It is nice how there is different ways to approach similar setup

Need to sort out that aux air valve next so I don't need to wait over 10 minutes to get normal idle, that is making also problems to low throttle behaviour.

Post by **pettaw** » 23 Mar 2006 07:28 pm

Mmm, interesting to see your VE bins, cos your idle VE is a lot lower than mine. If I set mine that low the engine would cut out. I've used the stock wiring so all 4 injectors are wired into the the same output, so maybe that accounts for it.

It would be interesting to see your file too.

If the air line to your fuel pressure regulator is disconnected, then effectively you're fuel pressure will be all over the place, so I would reconnect it. It alters in real life, but actually will be more constant in relation to manifold vacuum.

Post by **jtbo** » 23 Mar 2006 08:10 pm

pettaw wrote:Mmm, interesting to see your VE bins, cos your idle VE is a lot lower than mine. If I set mine that low the engine would cut out. I've used the stock wiring so all 4 injectors are wired into the the same output, so maybe that accounts for it.

It would be interesting to see your file too.

If the air line to your fuel pressure regulator is disconnected, then effectively you're fuel pressure will be all over the place, so I would reconnect it. It alters in real life, but actually will be more constant in relation to manifold vacuum.

I don't quite follow when you say that "fuel pressure will be all over the place" as it is now constant 100kpa (ok, 92-98 in this area) so then fuel pressure is not reduced by regulator valve and is constant 2,5bar, which I noticed with fuel pressure gauge before it got broken.
From my understanding fuel pressure regulator is valve that has spring and vacuum is greater than spring strength and so fuel pressure is reduced when there is enough vacuum (low map value)

That is why I think my low map VE bins are naturally lower as I have 2,5bar fuel pressure instead of 1,5 that is with 20kpa if I remember right.
Fuel pump is making around 5bar pressure at max and fuel pressure regulator keeps that in 2,5bar and because pressure is affected only by injector opening it should be constant as fuel pressure regulator spring takes care of adjusting pressure to 2,5bar when injectors are under higher duty cycle.

I had lot of trouble with leaning out before I did disconnect fuel pressure regulator vacuum hose, I needed to run high reqfuel and ve-map had over 100 values, seemed like injectors would been too small, all this was cured by disconnecting that hose.

I think that I'm right with this, but naturally I can be wrong as I have not very much information about pressure valves. But certainly my car runs better and is easier to tune now.

This is my msq file, it is for extra 29j version:
http://www.janiervast.com/gallery/d/159 ... 221759_msq

Many settings are not tuned yet, acceleration enrichments are not in use, cold start is set to something that starts the car somehow, etc.

My idle is actually bit of rich side, I can only guess, but I think it is 13 to 13.5.

I have only 2 injectors in one bank, 2 in other, injector driver fets are not warming at all even I don't have any heatsink in them. I think that there is only cpu that warms a bit in ms board, but even that is just bit warmer than metal case, that I have still not been able to finish

Post by **pettaw** » 23 Mar 2006 10:41 pm

Ahh OK, well, the fuel pressure regulator Edit: needs to lower the pressure so it maintains a constant drop across the manifolds so the injector patterns remain the same.

That would explain how you're managing to run much lower VE settings than me and still have a very rich mixture.

I haven't had a chance to look at your file yet. My accel enrichments are all fairly well tuned. I'm using TPSdot and also the cold start and afterstart are also not far off.

I'm also having to run over 100%VE in some places and even those at the top of the scale are slightly leanish, but because I run closed loop with my wideband, luckily the EGO correction sorts that out. I would try my msq file, modified to work with your setup and reconnect the vacuum hose and see what happens.

Post by **jtbo** » 24 Mar 2006 12:38 am

I try to look tomorrow little of my aux air valve maybe after that I could test your map.

Also I think that it is possible to calculate correction for fuel pressure as I think it is fairly linear and then I just need to calculate change in each map setting and how much it changes to VE.

I try to get my fuel map adjusted so that I don't need much of EGO correction, also I will eventually setup VE table so that it won't lean from 14.7:1 only can make it richer, that is recommentation for Momo engine management, maybe good for ms too

Post by **pettaw** » 24 Mar 2006 01:29 am

That's a strange idea, adjust the VE so that at full negative correction EGO can only just get to 14.7. Might be worth an experiment.

Trouble is the car drives great, except this hunting at 1300-1700 RPM light throttle.

Try my fuel map, but make sure you reconnect that hose to the fuel pressure regulator and reset the REQ_FUEL to my settings too. You should find the accel bins won't be too far off either.