VE tables vs. engine characteristics
Turbo Mini Forums

last year, test engine ran fine on 680cc injectors and I could easily get AFRs above 13 at idle and nice smooth running.

last few days, "proper engine" would hardly run with 1000cc injectors and now I've put the 680cc ones back in it isn't much better, best I can get is AFRs around 11-12 if I drop the "req'd fuel" value really low.

Because of the "T" word, I've reloaded everything and the CPU now reports what it should and apart from a few, not important at the moment, glitches, MegaTunix gives me access to much more data than TunerStudio did even before.......

Nothing seems wrong now in the settings compared to last year but there are some fundamental differences with the actual engine.

All the settings in MS2 take the differences into account (ie, 1360cc instaed of 1275 so "req'd fuel" reset accordingly) but it seems my fuelling differences are only wrong at idle/low RPM - once above 2k RPM is seems to all come back in.

So do I need to skew the VE tables at low RPM, low duty ???

The differences are,

Initial test - 1275 MG Metro engine (N/A), fuel injection and twin widebands. Data from widebands was fine, AFRs around 14.

Second test - the "proper engine" 1360 but with a well skimmed head to compensate for the 14cc pistons and run with the turbo attached but on a carb to check out the dual wideband sample system. Again, data from widebands was fine and 14 easily achievable.

Present - "proper engine" but now with its's 8.1:1 CR ported head and everything else.

It seems the only fundamental difference is that my CR is now down at ~8.1:1

The change was intended, don't get me wrong, but is that change to CR responsible for the AFRs going all wrong at idle ????

I can get them right from 2k RPM upwards, just not at 1k RPM.

The only way around it seems to be to skew the VE tables at low RPM/duty but I'm hesitant to do it unless I'm sure it won't affect higher RPMs/Loads.

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

I think that you have answered your own question.

The fact that you have reduced the compression ratio from 10?:1 down to 8:1 means that the cylinders will still hold a lot more burnt air/fuel at the start of the compression stroke. Hence they need a lot less fuel.

I would expect you to have to reduce the VE by 20% to compensate.

I expect that it's OK from 2k upwards due to other factors cancelling out the affect of the lower CR.

Edited by Paul S on 20th Aug, 2010.

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."

OK Paul,

I think you have confirmed I'm thinking along the right lines.

What's the CR on your 998 ???

And can you post up a copy of your VE table (I assume you only have one as you're using the hybrid single pulse mode) so I can do a quick and dirty comparison to my two...

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

I'm running around 9:1 on the 998Ti.

VE Table:

Edited by Paul S on 20th Aug, 2010.

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."

when my engine was a low compression slug i needded idle afr's around 13:1 to get a decent idle. i could, however, get the afr's higher!

try turning your opening times down lower, particularly on the smaller (faster?!) injectors.

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

should add, i use the injector opening time as a fudge to improve ide afr / transients without really affecting WOT conditions, has a very pronounced effect. perhaps you have an issue with some correction on the injector opening time? like battery voltage?

turbo 16v k-series 11.9@118.9 :)

Denis O'Brien.

Denis,

I've tried reducing the injector openning times - I agree it's just a mathematical fudge to reduce the physical electrical pulse so should have more effect at low RPMs but it doesn't work for me....

Battery voltage may be an issue, the battery is totally knackered so I'm running it with a charger on all the time but that's the same as last year and I'm getting a steady 12-13V showing in the realtime display....

I'll buy a new one this weekend to eliminate that but the only fundamental difference to a year ago is the CR....

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

Well this is weird......

I can get it running reasonably well at 12-13 AFR, nowhere near as good as I want...

BUT,

I've now got the third wideband working in the exhaust after the turbo and it actually reads the average of the two before the turbo,

So that's good.

BUT, whilst playing with settings within MegaSquirt, if it does a burn there will be a short stutter.... expected so I normally leave the idle RPM high so it doesn't stall but recovers... but three days in a row now, I've done a burn that is a little bit too long and the engine stalls, then it simply will not re-start.

Even today with a new battery.

I've checked all the cranking settings to see if there is any significant difference between hot and cold but can't see anything...

If the last two days experience is repeated it will start fine tommorow once it's cold again.

I've got a few glitches when running, momentary jumps on the gauges which I think is down to the earthing - it's a bit temporary at the moment - or the fact that I need different settings on the VR pots when installed on the engine compared to the JimStim. But none of these seem related to the hot start problem...

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

Do you only get the glitches on the gauges of go you also have the engine stumbling? If it's only the gauges, it could be comm issue only depending if you're using MT, MTx or TS. The longer burn might also be a comm issue.

Have you tried simply shutting the engine and restarting it when the engine is warm? Does it do the same thing as when it stalls due to a burn? If it does then it's more likely a config or tune issue but if it can restart when you simply shut it down then it's more related to something with the PC (comm or tuning software).

I probably stated things you've already figured out but I just wanted to make sure.

Jean

http://www.jbperf.com/

Thanks Jean,

I't difficult at the moment to say whether the glitches on the gauges are cause or effect - because I'm far too rich at the moment it's not idling particularly smoothly so although a jump on a gauge seems to coincide with a "jump" in the engine note it's not yet clear if the gauge is reading simply what is happening or something in the comms is the cause.

ie, pulse width gauges suddenly dip and engine falters - was it a rich misfire causing the engine to instantaneously lose speed so the pulse width drops and the gauge is reporting it correctly, or is some glitch causing the pulsewidth to drop suddenly so the engine falters as a result.

I'll have to start logging the data properly to be sure but I really want to get the AFRs better first, the excessivly rich mixture will not be doing my brand new bores much good.

Comms is next on my list - Graham T had lots of problems and ended up dropping the frame rate but I've avoided it so far as there were no issues when testing on the garage floor and no issues on the JimStim. Software is a mixture of MTx and TS. (The old MT simply doesn't work on 3.0.3r, well not for me anyway, I have to go back to 3.0.3e for it to work). So both are obviously a lot different to when I had it running well last year.

It's difficult to say for certain if the hot restart (or lack of it) is related to the burn and/or glitches or just inherant (ie, tune) because the few times I've tried, results have been variable, sometimes instant, sometimes after ages cranking with the throttle wide open - difficult with a knackered battery.....

I do have a design "deficiency" which has only just become obvious now it's all in the car which certainly isn't helping..... Unlike Paul's design with a stainless inlet manifold that sweeps high above his stainless exhaust maniold, my aluminium inlet manifold sits right down inside the standard Metro Turbo cast iron exhaust manifold with 1mm clearance.....so the inlet manifold runs very hot and as soon as the engine stops hot, I get masses of heat soak into the MAT sensor and its reading climbs very fast. I guess this has a fair impact on the fuelling equation. So I think if I can restart quickly it starts but as the stall is probably something I just altered I try to alter it back first which takes time so then I have to crank with throttle wide open to cool the sensor, then the battery dies......

Annoying because the sensor I'm using is from a common turbocharged engine with an aluminium inlet and physically looks to have the bulb well seperated by plastic/epoxy from the brass body but, with hindsight, the engine it belongs on is x-flow with an alloy head so the inlet manifold doesn't ever get hot...... I'll have to dismantle my manifold and redrill it to accept a nice wide nylon sleeve in that area to eliminate that from the list of possible reasons.

Today's job is to go buy a brand new battery - the one I used yesterday was "new" in that I borrowed it from another car but it wasn't much better in reallity.

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

perhaps stick the sensor in the pipe from the ic rod ?

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

hiff 44 !!!

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, as an aside, RPM is part of the equation (which is deriven from the universal gas equation pv+mRT), it's just "hidden" in the MegaSquirt version as part of "VE" (VE values allways rise with RPM, see your own table above).

However, your point about MAP instantaneously rising could certainly override the RPM drop...

That certainly suggests the downwards glitch in pulse width is cause rather than effect.

I have a new battery (outrageous price), just put in on charge to top it up and done some more testing on the JimStim....

Pulse width does everything I'd expect at varying MAP/MAT above 2k RPM but as soon as I wind below 2k RPM it rises a lot (which would explain my rich mixtures below 2k) and is affected transiently by TPS when it shouldn't be....

So I think you are onto something with AE being the root cause but, if so, something else got screwed when I lost all the settings which I haven't yet spotted.

TPS is calibrated (several times) as are the MAP and CTS, all three checked several times and all three show nice and smooth on the MTx real time display - thinking about it more, it seems to be only the calculated values that "glitch" (and RPM which jumps because of the glitches) whilst the measured values were steady...... I'll double check once the battery is fully charged.

And,
Robert, moving the sensor would be a bodge and it isn't that difficult to enlarge the hole and sleeve it,
Fabrice, I'd never screw an HIF44 into the sensor hole no matter how big I tried to make it :)

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

Try and log or observe Gammae. If that is steady when the stumbles ocurr, then it will be AE.

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."

Will do - it will be tommorow now, I'll leave the battery on a trickle charge overnight.

Meantime I'll save an msq as it now is then go searching all the settings on the JimStim.

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

Oh right,
Then maybe a twin's 44!

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 ???

Medusa + injection = too much torque for the dyno ..https://youtu.be/qg5o0_tJxYM

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

I'm sure you have this set correctly, but do you have multiply map set to "multiply"?

Mine changed to "don't multiply" after one of the software updates. I'm sure it was not a change I made... After I changed it back to Multiply, I was able to get good tick over AFR's and better start up, hot or cold, which I could not prior to that.

’77 Clubman build thread
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=618189

Siamese 5 port EFI testing
http://www.turbominis.co.uk/forums/index.php?p=vt&tid=611675

Just as a test to make sure that it is not a calculation issue.......

Whilst on the JimStim, put the REQ_FUEL value up by a factor of ten.

If the pulse widths settles down below 2k, then it's a code issue.

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."

Ahhhhhrrrrggghhhh

Graham, it was set to multiply, Paul had already suggested I check that....

I've just checked again and it's back to "don't".......

That is reading it through MegaTunix, not the "T" word that my psychoanalyst has told me to avoid.....

I'm seem to have been well and truely screwed by the TunerStudio "update".

Paul, I'll do your test in a few minutes once I've saved all the files with "MAP multiply" setting suddenly being wrong again.

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

UNBELIEVABLE....

MegaTunix and MegaTune (only works with an early ini file loaded) both have the ability to read from the cpu and both say "Multiply MAP" is OFF.

TunerStudio says it is on but can't read from the cpu to confirm it but even if I "burn" "on" in TunerStudio, both MT and MTx say it is "off".

I need my resident psyhoanalyst to restrain me......

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