multiple WB's and trim + EGO question

[Yves]

Have to come back to this as I'm having some difficulties.

I would have attached an msq but I get the message the quota has been reached...

Anyway.

My sensors are hard wired as follows :

ADC 1 = WB cyl 2 (via Can)
ADC 2 = WB cyl 1 (via can)
ADC3 = WB cyl 3 (via can)
ADC4 = WB cyl 4 (via can)
ADC5 = WB cyl 5 (via can)
ADC6 = WB cyl 6 (via can)
ADC7 = WB cyl 7 = JS4 (ADC 7 main board)
ADC8 = WB cyl 8 = normal EGO via main board.

firing order is 1-8-4-3-6-5-7-2 and
A = 1
B = 8
C = 4
D = 3
E = 6
F = 5
G = 7
H = 2



I was trying to tune the cyls trim and when touching the number 8 trim it gave a reading on the number one EGO correction gauge...

I don't see what I'm doing wrong here.

[Yves]

Based on what I'm seeing the egocorrection channels are numbered according to the listing in the ego tab.

[jsmcortina]

Based on what I'm seeing the egocorrection channels are numbered according to the listing in the ego tab.

That sounds correct.

James

[racingmini_mtl]

Based on what I'm seeing the egocorrection channels are numbered according to the listing in the ego tab.

That sounds correct.

James

I thought the trims were supposed to be with respect to the firing order as defined in the engine and sequential settings. Otherwise what is this firing order used for? Ignition? The menu does say that the trims are for _cylinders_ 1-4, 5-8...

Jean

[jsmcortina]

Trim tables are per cylinder number yes. I think / thought OP was referring to the datalog channels?

James

[Yves]

Must say all this is rather confusing.

On the one hand you seem to have egocorrection being applied to the cyl assigned to the egochannel (left hand column) and on the other hand, when you put egocorrection gauges on the dash, they will follow the egochannel numbering (left hand column). This means in my case the number 8 cyl is causing the egocorrection gauge number one to react...
Meanwhile, i left the egochannel under the gauge properties the same, but changed the name on the gauge to the correct cylinder so at least I can see how much deviation the trim is causing/not causing to which cyl.

Now, the egocorrection should correct the injector attached to it. However, I hope that the trimautotune feature follows the same thing. I assume so, but...
Futhermore I seem to notice that the trimautotune seems to react differently when EGO is on and when it is not on. Not sure how it is supposed to trim the fuel per cylinder, but if it's not based on the egocorrection it could be chasing it's own tail. For example I've done 2 consecutive trimautotunes, where in between each I've reset the tables to 0 again. Once was with egocorrection on and once without. The difference in trim on some cyls was mayor for some cyls. I mean in the orther of 20% or so. I have set itb balance with a gauge, so unless something significantly has changed (which I presume is not the case since both my maf's read about the same) I do no think a couple of cylinders would read 20% out....

OR if it is correct what I'm seeing I've opened a can of worms here with individual WB's

[Marek]

There is no reason why individual cylinders may not be far apart, particularly when airflow/rpm is low.

There is a slight oversimplification in the fuelling algorithm, in that it assumes that straight line fuelling along the lines of y=mx+c, where you can enter a different "c", but all injectors are assumed to have the same slope "m". Once you move from a four cylinder engine to an eight, or to a v12, it is often the case that not all of the injectors flow anywhere near the same amounts as eachother and this could much more easily be dealt with this way, allowing the trim table to do its job of actually trimming the cylinder airflow, not both of these parameters together. It gets even more interesting for dual fuel users, especially when I can see that the second set of injectors are not quite linear. See http://www.msextra.com/forums/viewtopic ... 25&t=57691
Essentially, the more inputs you have, the more likely they won't be well matched to each other.

Regarding EGO trims, perhaps the most sensible suggestion is to put different coloured electrical tape around the leads of the widebands and assign the same colour traces when looking at the relevant graphs in MLV.

It may be helpful for the firmware/TS interface to allow text labels to be applied or have an "EGO wiring order" or remapping in the same way that we have a "firing order" so that EGO1-8 (as gauges) can be associated with SeqPW1-8 and Trim1-8 more easily. I note that the IOX allows the ADC channels to be remapped this way.

kind regards
Marek

[Yves]

I just came back from a short drive with autotrim on.

Most cyls are close together but I have a couple that have a wide spread. This is more particularly the first cyl in one airbox and the last cyl in the next. And the issue is also more prevalent at low kPa (decel).
Going to reset the cyl balance next and see how far they are off.

I'm also seeing some difference in my older Innovate 4.2's and the 14point7 4.9's.

Anyways, it seems to balance out the engine operation a lot more and tame the negative effects of an aggressive cam.

[Yves]

Ok, today I found out my exhaust header gaskets started leaking big time. This was also apparent by the AFR fluctuating...

My bad, should have checked basic things first.

[LT401Vette]

Must say all this is rather confusing.

On the one hand you seem to have egocorrection being applied to the cyl assigned to the egochannel (left hand column) and on the other hand, when you put egocorrection gauges on the dash, they will follow the egochannel numbering (left hand column). This means in my case the number 8 cyl is causing the egocorrection gauge number one to react...
Meanwhile, i left the egochannel under the gauge properties the same, but changed the name on the gauge to the correct cylinder so at least I can see how much deviation the trim is causing/not causing to which cyl.

Now, the egocorrection should correct the injector attached to it. However, I hope that the trimautotune feature follows the same thing. I assume so, but...
Futhermore I seem to notice that the trimautotune seems to react differently when EGO is on and when it is not on. Not sure how it is supposed to trim the fuel per cylinder, but if it's not based on the egocorrection it could be chasing it's own tail. For example I've done 2 consecutive trimautotunes, where in between each I've reset the tables to 0 again. Once was with egocorrection on and once without. The difference in trim on some cyls was mayor for some cyls. I mean in the orther of 20% or so. I have set itb balance with a gauge, so unless something significantly has changed (which I presume is not the case since both my maf's read about the same) I do no think a couple of cylinders would read 20% out....

OR if it is correct what I'm seeing I've opened a can of worms here with individual WB's

I'm sorry for the delayed response, I was out of town.

In Trim Autotune, when you select the EGO Sensor, that will drive what afr and egoCorrection channel are used.

Example, if you set Cyl 1 to EGO2, then afr2 and egoCorr2 will be used.

[Yves]

Coming back to this :

Before my leave on holiday I found out that my exhaust header gasket were leaking. So I changed them, but hadn't had much time to test. Any exhaust leak in the header system has of course a great influence on the AFR reading but the effect is rather minor per WB so that it could easily be mistaken as a fueling error (either too rich or too lean) causing any autotrim feature to work wrong.

Next thing is that your fuel tables need to be perfectly tuned. Any error in that respect gets interpreted by as a deviation. So if you are already lean in a certain area in the main fuel table, you will get worse readings on the cyls that inhale more air, causing the system to generally overcorrect on that cylinder. Not sure if that explains it well.
Same thing happens when your overall to rich. With the extra remark that an overly rich situation could cause one or too cyls to start misfiring which is then seen by the WB as a very lean situation and the autotrim corrects accordingly. This only aggravates the situation.

Last thing : the logged fields were not foreseen for the use of multiple WB's. I usually use the autotune feature in MLV which seems to be more accurate and this imposes problems as MLV normally uses AFR1 and EGO CORR 1 for changing fueling. Luckily the new TS version allows for adding custom channels, so I needed to add an average AFR and average EGO.

[chips60bug]

Sorry to bring this one back from the dead, but I am in a similar situation with a 4 cyl car I just installed a MegaSpartan setup in. The car is turbocharged, so eventually back pressure data will be captured and the wideband readings will be adjusted to compensate for it. In the mean time, the sensors are in place and I am ready to start tuning this thing. My question is regarding having a 5th sensor in place, I have the post turbo sensor I've always used still in service, since I can't trust the readings from the MegaSpartan until the comp is coded in. Is there any way way to make TS use the 5th sensor for VEAL duties, and still adjust the trim tables correctly. Also, it would be nice to have EGOcorr active while the back pressure is low, in order to keep all 4 cylinders happy and in check. What needs to happen as far as settings to make this work? I've attached an msq showing how I have configured the widebands for the moment. I previously had the 5th one setup on EGO5, but I have now switched it to EGO1 and then have offset all of the AFR/EGO Mapping sensors down by 1. I also shifted them all on the Trim table analyze page as well. Is this the right procedure to allow me to tune the VE table by the 5th and post turbo sensor, while allowing me to trim each individual one?

EGO.jpg

I hope what I am saying makes sense

[racingmini_mtl]

That doesn't answer your question but you should use CAN EGO for the first 4 ports because the MS3 will not read the values correctly with this being misaligned with the CAN device. CAN EGO in port 2 is for the second port on the CAN device so you won't get the first sensor data and there won't be any data for EGO5.

Jean

[LT401Vette]

Currently, you don't need the 5th WB.. If you are set to 4 sensors, normal autotune will use the average of the 4 sensors while using the individual sensors for trim autotune.

If you add a 5th, then if will use the average of all 5 for normal autotune. Which is should be fine.
The same is true for egoCorr... However, with the 5th sensor installed, I'm not sure what your egoCorr5 will be reading. It is likely holding steady at 100, so that would just make the error look 20% lower and the direction would still be correct.

[chips60bug]

Good to know. My biggest issue is that I can't really trust the readings from the 4 sensors once the exhaust pressurizes. I still need the data from them though. How can I use number 5 WB for all tuning feedback and 1-4 for basic data gathering. I need to be able to show all 5 sensors on a data log. I have no problem with turning EGOcorr off for now.