Siamese Code Trial - Take Three
Turbo Mini Forums

Carrying on from Jean's thread about the new code.

Log of AFRs etc at part load attached.

We will do some more next wekend, weather permitting. Hopefully get full throttle sorted.

Attachments:

44.48kb

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

And another one:

Attachments:

44.54kb

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Do you have any acceleration enrichment set? And if so, does it get triggered? That could throw things off if there are false events (or even real ones).

By the way, I haven't done anything about it so transients behaviour is undetermined at this point.

Jean

http://www.jbperf.com/

No acceleration enrichment set at the moment. Just trying to baseline with a simple set of rules.

It does lean off if you stamp on the throttle, so we need to do something.

The reson that I'm not showing any full throttle logs is that we have a rouge value in the timing table at 90kPA MAP and 4000 rpm what should have been 150 degrees was only 15 degrees. So at anything above 3000 rpm and 80kPA MAP went very lean at the outer cylinders. Unfortuntaly did not find out until everything was packed away and we were running the MS on the stim to check settings.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Also, the reason that the AFRs are a bit on the lean side is that we had to reduce the fueling to the inner cylinders by 30-40% to get them close !!!

We need to increase the Req_Fuel to compensate.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

30-40%!!! That's huge. Could it be that the timing of the outer cylinders is off and you're actually fueling the inner cylinders from the outer cylinders pulse?

What happens if you increase or decrease timing? Again, if you put the timing of the inner cylinders 180 degrees out (instead of same timing), it will make it easier to set the correct timing for the outer cylinders. With them only 180 degrees apart, it could be possible that the end of the inner cylinder pulse feeds the outer cylinder and vice-versa which will make it tough to establish the correct timing.

Jean

http://www.jbperf.com/

We tried shifting the timing from around 90 to 180 degrees BTDC for the mid pulse. The best settings at full thrittle are in the region of 140-160 degrees BTDC.

I have some calculations that back up figures of this magnitude.

I want to stick with firing into the open valve for now. Shifting the timing of the inner cylinders back 180 degrees puts us back to where we were with the semi-sequential setup.

We need to re-calibrate the sensors and do some more runs. Probably adjust the timing now that the AFRs are closer. If we can get the inners to run lean, we can add some more fuel back in.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

That sounds good. At this point, you certainly have more experience than I do so I'm pretty much just a back seat driver (and probably as useful) :)

Have you posted the calculations to come up with the approximate timing? If not, that would be interesting to see.

Jean

http://www.jbperf.com/

The good thing is that the code has sufficient flexibilty to make things work.

Every change of timing or VE shows on the AFR meters.

I'll update my injection timing spreadsheet and post for comment later.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

The attached spreadsheet shows a method of calculating the injection timing points.

It is based on Marcel's work detailed on his website.

http://www.starchak.ca/efi/speed.htm

It assumes an average port velocity through a single stroke of 180 degrees, hence is an approximation.

Enjoy.

EDIT: Calcs removed after errors found. I'll post the revised calcs later.

Edited by Paul S on 1st Mar, 2009.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Are those the values with the new ve rates? In the 1.5-3k range we are using injection timing rates close to the spreadsheet.

Yes, I recalculated it this morning using the actual VE values for the outer cylinders.

Interestingly, if i reduce the VE, as we have done for the inner cylinders, it means that you need more injection timing advance to get the fuel in before overlapping with the outer cylinders.

We should consider invoking the second injection timing table and advancing it, say 30-40 degrees to see if that helps.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Paul,

I have been wondering how well the engine starts with the code. Do you have to crank for a while?

Jean

http://www.jbperf.com/

Jean,

Probably just a couple of crank revs more than with the original code, but no real noticeable difference.

It synchs straight away.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Still struggling to get the inner cylinders to run at the same AFR as the outers at wide open throttle.

Adjusting the injection timing point to the optimum brings them closer than with the semi-sequential code.

If we reduce the VE for the inner cylinders, both the inner and outer cylinders run leaner.

I think that it is due to wall wetting which may be exagerated by the spray type injectors.

It's fine at low load, but as soon as the pulse width goes above a certain value, the inners run rich.

I'm now going to fit some pencil stream injectors and run more tests.

Edited by Paul S on 28th Feb, 2009.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

What duty are you running at?
Can you up (or lower) the fuel pressure to see if the wetting occurs at the same RPM , or if it is more duty-clcle (pulses merging) related?

Significant difference on the new injectors.

They are slightly smaller at 375 cc/min.

The optimum mid-pulse timing advance at idle is 40 degrees less.

Haven't been able to do much testing under load as Sturgeo has gone out. Hopefully we will get some time tomorrow to play with the setup.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Edited by Paul S on 28th Feb, 2009.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

What value are you using for the opening time? If this is not correct, the actual fuel timing is somewhat affected but more importantly, the fueling is not affected as you would think when updating the VE table(s).

If it's too long even with a very small VE value you will still get a significant amount of fuel being injected. With the very small pulse widths you're using, if you vary the VE value a certain percentage, the actual fueling will not vary by the same percentage because a significant part of the actual fuel being injected is in the opening time which is constant.

So it might be good to start lowering it if you have not done anything other than use the default 1ms.

Jean

http://www.jbperf.com/

Edited by Paul S on 28th Feb, 2009.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

That sounds good. It should mean that the error, if any, is not significant.

Jean

http://www.jbperf.com/

Well, Ive spent another weekend on this and still no nearer.

We start by trying to use injection timing alone to get the AFRs as close as possible. Using the same injection timing for inner and outer cylinders. Also same VE tables for inner and outer cylinders.

Then we trim the VE tables for the inner cylinders to try to lean them out.

Then we adjust the injection timing for the inner and outer cylinders.

But, above 3000 rpm we cannot get them any nearer than they were with the original semi-sequential code.

There is something going on here that we do not understand.

Yours disillusioned.........

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

Can you send me your msq? I'd like to have a look on the bench to see if there is something with the code that's not as it should be.

Jean

http://www.jbperf.com/

Jean,

Email on it way to you.

I must say that the car is running very well.

We were doing 90mph on the flat. On my private test track, of course.

That's a full weight 998, 9:1 CR with a 2.95 diff at 4500rpm with 25 stone on board.

Takes 60nanas to do 90 which equates to around 75 curley wurleys which would equate to about 75nanas at 5250rpm, assuming a flat torque curve.

Edited by Paul S on 1st Mar, 2009.

Saul Bellow - "A great deal of intelligence can be invested in ignorance when the need for illusion is deep."
Stephen Hawking - "The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."

msq received. I'll have a look and report back to you.

It would be very interesting to see what the AFR distribution is on a well tuned high power turbo 5-port engine. It is possible that you're already in a better situation with your current tune.

Of course, if you (we?) can find a way to correct the distribution perfectly, it would be even better. And I don't see how that can't be achieved somehow but it may involve a lot more work.

Jean

http://www.jbperf.com/