For the first time today a had a trigger error. The i could induce it a second time and made a log (fortunately laptop was with me).
After that a put another gear it and rev it more slowly up top and got no errors whatsoever.
what could be causing that? Is it a minor issue and it was made that present because of the settings i have on the secondary trigger area?
log
http://www.vems.hu/files/Billman/v3.3_n001441-2010.11.08-17.21.51.zip
See http://www.vemssupport.com/forum/index.php/topic,1469.0.html I raised it and it seems it's a firmware bug
so you mean i should only set it to show errors in the sec trigger setting "Event of bad trigger" Just show and leave the rest at 0 to 359?
That's what I've been told, yes
Set "event at bad trigger position" to "just show (default)" = which is the default setting.
I don't think this setting is the problem, but you never know. You never said if this caused the engine to stumble before you let your foot of the gas, or after.
I would fix some of the below first :
Unless you're actively using the knock sensor, which is highly untested with recent firmware, I would set "knockchan" to 240 as is recommended to turn it off.
I think you are using way too much ignition advance on low cruising load, but you may not spend much time below 80 kPa with an aggressive driving style.. :)
You're running the engine far too lean in my opinion, this could make the engine mis-fire and cause trigger errors. Lambda 0.90 at full load is for N/A engines only and they usually run richer than that. Set to 0.82-0.83 for anything above 30-40 kPa boost (160 kPa row and up for you).
Most importantly, you have set the laws of physics out of play by setting the MAT/TPS enrichment table to 100% all over. Drag'n'drop the default table from one of the config files in the "VemsTune\defaultfirmwareconfigs" directory.
Your tune will change a lot depending on ambient temperature if you don't do this!
Make sure input functions you're not using are disabled, two are not and you're probably not aware. In this case they were not dangerous.
You're not using the correctly calculated req fuel if your injectors are truly 630 cc/min, but they probably are since you have higher values in the VE table than what you'd normally see. You should have 4.5 ms, not 3.1 ms. This affects the scaling of some of the enrichment functions.
Since you've enabled EGO correction for all loads you can use it for tuning with "VE tune by statistics" in the Tools menu. This makes it easier for you when you have set the MAT/TPS enrichment table to the correct values. Just drive around for a while and let it suggest VE changes. Apply those and manually smooth out the table and you should be good to go.
I would not recommend any EGO correction at full load in the long haul. If you need correction at those loads you usually have a serious problem somewhere and you should not trust the lambda sensor for this job. On an Audi like this I would disable it above 120-130 kPa and 30+ % TPS, so it never correct when you "get on it".
Some other things could probably be said, but these are the most obvious ones from a short 5 minute review.
Ignition looks about right to me the engine is a 2.5 stroker the same as mine
Quote from: mattias on November 08, 2010, 05:27:28 PM
You never said if this caused the engine to stumble before you let your foot of the gas, or after.
It happened before i let the throttle off. (note i don't have an inverter in the car now but i have an adaptor made but not use it till now)
Quote from: mattias on November 08, 2010, 05:27:28 PM
Unless you're actively using the knock sensor, which is highly untested with recent firmware, I would set "knockchan" to 240 as is recommended to turn it off.
I just monitor them to see if they pick up something. Right now as i am in the process to increase boost i am listening directly to the one knock sensor and ignition at least at these areas seems good.
Quote from: mattias on November 08, 2010, 05:27:28 PM
Most importantly, you have set the laws of physics out of play by setting the MAT/TPS enrichment table to 100% all over. Drag'n'drop the default table from one of the config files in the "VemsTune\defaultfirmwareconfigs" directory.
Your tune will change a lot depending on ambient temperature if you don't do this!
I am going to do it but since the EGO is for the time beeing on all the time i have no issues.
Quote from: mattias on November 08, 2010, 05:27:28 PM
You're not using the correctly calculated req fuel if your injectors are truly 630 cc/min, but they probably are since you have higher values in the VE table than what you'd normally see. You should have 4.5 ms, not 3.1 ms. This affects the scaling of some of the enrichment functions.
My injectors flow 630 at 3 bar but i use a 3.8 bar fual pressure regulator.
I would like to thank both of you and if you find any other issues lease let me know.
Quote from: Jamo on November 08, 2010, 05:48:39 PM
Ignition looks about right to me the engine is a 2.5 stroker the same as mine
Is it the lean cruise lambda of 1.05 that makes 38 degrees possible?
It's certainly interesting because I managed better than Motronic did on an Audi S4 using only 28-30 degrees timing, lambda 1.02.
It was not stock, 630 cc/min injectors, tubular exhaust manifold, larger turbo, larger intercooler, etc..
Did you confirmed tdc with timing light as many seen different figures although not so far away.
Yes, it was confirmed.
Quote from: mattias on November 08, 2010, 10:53:36 PM
Yes, it was confirmed.
Can you share this map? Till 0.8 bar i hear no knock by listening directly to the sensor.
I don't think you understand, the engine doesn't knock at these low loads. You can use too much advance and never know it, you will only loose some torque and get higher fuel consumption as a result of pushing the throttle a bit more to compensate.
aaa, i see. I thought you meant it knocked.
You tuned on a dyno?
Can you share this map?
@back to topic
What else can cause this sec trigger error?
Engine starts every time with no issues or backfires, cam sensor is new.
Quote from: mattias on November 08, 2010, 09:28:15 PM
Is it the lean cruise lambda of 1.05 that makes 38 degrees possible?
It's certainly interesting because I managed better than Motronic did on an Audi S4 using only 28-30 degrees timing, lambda 1.02.
It was not stock, 630 cc/min injectors, tubular exhaust manifold, larger turbo, larger intercooler, etc..
I had interesting case when owner of one VW 2L 16V got extremely low consumption on ~2000km trip after he incorrectly orient trigger wheel by ~20deg after engine repair. He got something like 47...55deg at 1.05..1.10 lambda at cruising. Now I put around 45deg with normal 25...30 WOT advance. And owner is very happy. CR 10.5:1 same shape of combustion chamber as 20VT. So I am confused.
This is why I asked if anyone knew if the lean mixture causes a much slower burn, requiring much more advance - and it probably does so more than I thought. I will try going for leaner mixture and more advance next time.
Wasn't there a thread about this around here?
Not much slower burning when comparing to our usual rich mixtures. But density are lower, flame initial forming takes longer (initial pressure rising after spark). And advance numbers can be very high. Main reason of low advance values on OE is NOx I think.
We also have issues to set high advance numbers for cruising. Especially on bigger engines. Cruising speeds of bigger ones starts even as low as 1500Rpm - too close to idling speed. As result advance table becomes too sloppy. And it can lead to bad jerking in transition area.
Solution here can be some kind of damper for advance what prevent from fast advance changing. Some ECUs have this feature.
The case of the VW engine has something working in it's favor, the high C/R allows a good homogenised mixture that helps a lean and slow burn to take place, sometimes I wonder how much a degree or two can make .. there might be little difference between 30 and 35 degrees advance in some situations. In any case the older Audi 20 valve turbo engine has less C/R to handle the boost but like most 4 valve/cyl engines it's superior to older 2 valve/cyl engines (unless you're comparing it to old high C/R racing engines).
I am too looking for a feature where there is a separate idle advance table, or even just a single advance number perhaps affected by coolant temperature. This is very common in other systems and would allow a better driving experience and fuel consumption with engines that utilise the low rpm for driving. I'll make a wish. :)
Quote from: mattias on November 09, 2010, 02:44:21 PM
I am too looking for a feature where there is a separate idle advance table, or even just a single advance number perhaps affected by coolant temperature. This is very common in other systems and would allow a better driving experience and fuel consumption with engines that utilise the low rpm for driving. I'll make a wish. :)
Key question is how ECU will migrates from main table to idle advance and back.
I assume 20VT could be happy with even more than my 45 degries on VW due to lower CR.
Quote from: mattias on November 08, 2010, 09:28:15 PM
Quote from: Jamo on November 08, 2010, 05:48:39 PM
Ignition looks about right to me the engine is a 2.5 stroker the same as mine
Is it the lean cruise lambda of 1.05 that makes 38 degrees possible?
It's certainly interesting because I managed better than Motronic did on an Audi S4 using only 28-30 degrees timing, lambda 1.02.
It was not stock, 630 cc/min injectors, tubular exhaust manifold, larger turbo, larger intercooler, etc..
It's a diesel block / petrol head with a longer stroker 86.4 vs 95.5 and lower CR because of the custom pistons
Quote from: GintsK on November 10, 2010, 06:32:07 AMI assume 20VT could be happy with even more than my 45 degries on VW due to lower CR.
There are two sides to the coin, I'm not sure it will work out like that, with lower C/R you can end up with less squish area and less turbulence to create a homogenised air/fuel mixture that won't burn as well, creating less power.
would the 1.1.85 FW version cure the trigger error on the RS2? Or is it for other Audi engines?
[1.1.85] - not released yet
* auditrigger divby2 changed to divbyANY
o toothrel_missing=01 for divby2
o toothrel_missing=19 for divby26
o toothrel_missing=1a for divby27
o with auditrigger 11000 RPM could be reached earlier (in benchtests) using divby27. Not 25500 RPM as with lower toothcounts, but sufficient.
o note: auditrigger divby3 (which is used in hundreds of 135 tooth audi engines) maintains compatibility
o question: what about Nissan 360 impulse dividing?
There was a false secondary trigger warning, which should not cause any issues (only warn) with 1.1.80 and 1.1.81. I've updated to 1.1.84 on a S4 (AAN) and S2 (3B with 60-2 trigger from more recent VW/Audi engine), they don't warn now.
You could try 1.1.84 to avoid false warnings for secondary trigger, to see the real ones. You must use latest VemsTune to make upgrade simple, new SafetyMode settings are automatically set.
The changes for 1.1.81 is to support other dividers for triggers with many teeth, but different count than Audi 5 cyl (135), now supports 130 tooth on Porsche 4 cyl by using different dividers.
whenever a fw version upgrade is done should one always recheck the timing with a strobe light timing gun?
or should this be done when a trigger setting is changed?
Always check when you change firmware
$^&£
this means that i will have to end up buying a gun...
Quote from: mattias on November 18, 2010, 05:54:57 PM
There was a false secondary trigger warning, which should not cause any issues (only warn) with 1.1.80 and 1.1.81. I've updated to 1.1.84 on a S4 (AAN) and S2 (3B with 60-2 trigger from more recent VW/Audi engine), they don't warn now.
You could try 1.1.84 to avoid false warnings for secondary trigger, to see the real ones. You must use latest VemsTune to make upgrade simple, new SafetyMode settings are automatically set.
The changes for 1.1.81 is to support other dividers for triggers with many teeth, but different count than Audi 5 cyl (135), now supports 130 tooth on Porsche 4 cyl by using different dividers.
do you think this strange shutter i got the other day was because of the 1.1.81 error and the fact that i didnt had "Just show" on vems tune?
Yes, if you don't have "just show (default)" in the setting for what to do with a bad trigger event on the secondary trigger, then you'll actually experience a consequence, and the engine will not run right. The engine should run just fine if you just use that setting. An upgrade to 1.1.84 is a very easy one if you're already at 1.1.81, if you used much older firmware like 1.1.27 then I'd definately advice using a strobe light to verify correct ignition timing.
I believe AVPs timing was the same for 1.1.27 to 1.1.81 am i right?
very tiny difference
on 1.1.27 i used 56 and on 1.1.81 it has to be between 56.5 and 57