Siamese Code Trial - Take Three
Turbo Mini Forums

I'm inclined to think that the problem is wall wetting.

It would only need 5-10% of the fuel injected to stay in the port until the next opening of the inner cylinder inlet valve to be added to the next injection to cause the AFR distribution we are seeing.

Reducing the injection pulse to the inner cylinder affects the AFR on the inner and outer cylinders, hence, there is some mixing of the pulses even though they are 180 degrees apart.

For a short period, both pulses sit in the inlet manifold. The second/outer cylinder pulse has started firing well before the inner cylinder valve has started to open.

A single pulse at the inlet valve overlap would not suffer from this problem.

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, it could just be my basic train of thought and it might not even be worth thinking about but...

i imagine when you inner cylinder afr's this is being represented by a wideband on the center exhaust primary?

is it possbile that because of the extra gas flowing through this part of the exhaust from the two cylinders that it is serving, rather that the one that the two outers are serving, that this is doing something to your readings?

im sure youve thought about this but i havent seen it mentioned? as i say your the experts, im just trying to think of possibilities.

sorry if this is pointless, im not used to all this sun.

EDIT: just thought that if you have two widebands then the outers wideband is obviously going to be connected after the two outer primaries join. therfore my question/"idea" is jeffing useless! i'll keep trying

Edited by James_H on 1st Mar, 2009.

Paul,

As I mentioned in my email, if you want to try a single pulse, you can simply use the dual mode setup (which you are already using) and fill the second VE table with zeros. Then set the timing to place the pulse where you need it.

Please note that even with the second table full of zeros cranking will be done with 2 pulses based on table 1.

Jean

http://www.jbperf.com/

Jean,

Can you please advise whether the "Injection Timing 2" table applies to the first or secong pulse?

I've just drawn a timeline for the injection pulses and inlet valve opening for a single port and it has thrown up some surprising facts and some ideas as to what to try next.

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 timing table 2 is for the cylinders 2 and 3 (as is the VE table 2). At least that was the intent and what I remember validating. :)

Jean

http://www.jbperf.com/

Well, weve been out this morning trying a few things and we are getting closer.

The problem is that there are too many variables with 2 VE tables and 2 injection timing tables.

Large variations (10-20%) on the VE tables seem to have little impact, so I'm inclined to leave them at the apparent optimum setting of VE2 = 80% VE1 above 60% MAP.

The injection timing tables only require precise figures above 3000 rpm. 80 degrees at idle and 115 degrees at 100% MAP up to 3500rpm give good results.

We are now working on refining the injection timing above 3500rpm and 80-100% MAP.

We have refined the Injection Timing Table 1 by driving around at full throttle, ragging the poor little engine to death

Tomorrow we will have a go at Injection Timing Table 2. It seems to want to go in the opposite direction to table 1

Edited by Paul S on 7th 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."

Paul,

Did you try to run with a single pulse on the overlap? By the way, if you need to switch from one mode to the other you can do it by having the second VE table only partially filled with zeros and having it go from one mode to the other at a specific RPM/load. But that's even more variables.

I must say that as complex as the code was to write, it seems as complex, if not more, to actually get to use it fully and efficiently especially since you're the first to do it. And you're doing a great job. In some ways, the code is more sophisticated than the Rover MPi code so this is really breaking new ground as far as I know.

Jean

http://www.jbperf.com/

In an ideal world it would be posssible to just tune the inner and outer cylinders separately by tuning the respective VE and Timing tables. But, they interact significantly and changing the setting for the outer cylinder affects the inner.

The biggest problem is getting enough fuel into the outer cylinders without making the inners richer.

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."

Played with Injection Timing Table 2 this morning but then developed an intermittent missfire :(

Nothing obvious from Megatune, so decided it must be an electrical problem.

Fitted another coil and then it started to poor with rain, so no more this weekend.

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."

Just been studying the logs taken yesterday and the previous weekend.

We can be cruising at 3000 rpm and say 60% MAP with equal AFRs. We then stop and then accelerate hard up through the gears. Once over 3000rpm the AFRs start spreading with the inners going rich to around 11-12:1 and the outers at 13-14:1. We then drop back to a cruise at 3000 rpm and the AFRs are still apart whereas before we gunned it, under the same conditions, they were equal. Odd.

I've concluded that it's either the inner coil starting to fail giving us a poor spark at high speed and thus showing as rich running. Or the LM-1/LC-1 are playing up.

I've already changed the coil, but have not had a chance to test it. I've also got a TechEdge DIY Wideband on order, so I'll try that once it's built.

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,

Have you checked the logs to make sure that the pulse widths are indeed the same at 3000RPM before and after going up? If they are then it's fine otherwise there might be something funky with the code.

Jean

http://www.jbperf.com/

could this not be down to wall wetting?

if once you go above 3000rpm at full load, the injected fuel is sticking to the ports, then when you drop back down to cruise the fuel that is on the walls is still being sucked in?

do the readings stabalise after a period of cruise? if they do then could this be the case?

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."

ahh, i see why you think it could be a "hardware" problem then.

do you know what cylinder the intermittent misfire was developed on?

Schrödinger's cat - so which one am I ???

Paul,

It would be really good if you could take MS logs because that would show if there is anything weird with either the inputs or the outputs.

It's always possible that something gets messed up and if the inputs are wrong then it will output something wrong. Or it could simply be the outputs that are wrong for some reason. Also, it's best to do it in the car because it could be related to noise or other type of interference which you won't see on the bench with the stim.

Jean

http://www.jbperf.com/

A missing track and YOU had to find out!?! That's really bad: a major mess up without them acknowledging it first or, even worse, without them knowing it.

http://www.jbperf.com/

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."

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."

Schrödinger's cat - so which one am I ???

Schrödinger's cat - so which one am I ???

They should have found that out before the first board got out of their shop. And if they missed that, they should have had the errata the first time there was a post about it.

Like I said that's really bad and things like that tell a lot about a company.

Jean

http://www.jbperf.com/

I don't know how good that WB controller is but it certainly is cheap and small: http://14point7.com/Widebands/NAW_OEM/NAW_OEM.htm.

Jean

http://www.jbperf.com/

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."

Schrödinger's cat - so which one am I ???