Look in the Idle RPM table. It should start with something like 1280 at the top coldest temp and progress lower from there. or just post a copy of your tune for one of the more knowledgeable members to look at - rename the file's .pvt to .txt so you can attach it here.

 

I STRONGLY suggest NOT zero'ing the CDE tables as it will directly affect fuel delivery and doing so would throw VE into a tailspin, it will also need to be completely retuned. The specific CDE tables in a given tune were used when the calibration was developed and they directly affect the fuel equation, there are also other factors tied into this table as well. I'll explain further...

CDE is not difficult to understand, and for those familiar with computer code it is actually pretty easy. The function of the CDE tables is to allow the ECM to better predict the EGR effect the motor is seeing at light load, this effect is a result of reversion and can/will vary based on variables such as cam overlap, throttle body, and exhaust system design. Since we cant accurately measure what that is going on within the motor in this respect the CDE table is used to help levelize the VE table, think of CDE as a Gain that allows you to add or reduce fuel delivery, it is independent from VE and part of the fuel equation. Many have likely wondered why VE tables go higher than 100 when it is impossible to achieve a volumetric efficiency of more than 100% (without forced induction), the VE in our ECM is simply a numerical representation from code. For computer guys, you probably noticed that the max VE value is 127.5, (half of 256/hexadecimal). The CDE tables allow us to somewhat smooth/levelize the VE's (below 60kpa). In areas that are near/at the 127.5 limit adjusting the CDE values down moves the window of adjustment down so to speak (remember my reference to Gain). The CDE value directly affects fuel delivery, raising it adds fuel, lowering the values remove fuel. As far as smoothing the VE's it can help in some applications, others not as much, and in some examples VE is going to end up with peaks, mountains and valley's, it all depends on what the motor wants. Where we generally use CDE is in the 3000-4500 range on motors with semi aggressive cams, this really helps get the VE in place at light/mid loads without having to go crazy raising the displacement constant.

 

I tried a few times to rename it to .txt, but no luck. It shows 0 bytes in file and won't upload. Don't know what I'm doing wrong.

 

try zipping the file and then uploading it- that will work too

 

this is all very good stuff- obviously different tuners have different uses for this- but of course that is first based on having a firm understanding of it. This and related advanced tuning topics (in relation to tuning with PV) are what I wish would be discussed more often.

So I 100% get what you are saying (I'm a code guy) - but there has to be some use for tweaking CDE with aftermarket pipes, aggressive tune and AC. I get that the stock table values were constructed for a stock bike- makes total sense- But in looking at every single one of my DJ-provided stage1 tune files, including the latest TT stage1 tune file sent to me, the CDE table values are 100% the same as my stock ECM tune as pulled from the ECM. With my open pipes I would guess there is a completely different reversion situation occurring than with stock baffled pipes.

In doing my own slight experimentation, lowering CDE a bit and then running a TT AT session, results in those affected VEs adjusting lower in numerical value. Conversely, raising CDE slightly causes the affected VE's to raise in numerical value. I've only done slight adjustments (5% - 10%) just to see the affect.

My reason for doing it is my logged AFRs don't seem to track my lambda table values accurately, they are leaner in some areas (mostly 60kpa and lower). the decel areas read vastly leaner than the lambda values, as much as 2-3 points

So I was hoping that by adjusting CDE, I could help get the actual AFR to match the table AFR better in certain areas. (0-30TP in my VE tables = 0-60KPa)

 

The reason the CDE tables are the same in your maps is because the cals were both developed with the same cams (stock), and with many combinations we dont touch it much or at all, this all depends on what the motor wants, on the flip side sometime CDE gets modified extensively. Not to say you cannot improve the tune by adjusting CDE, however I would first get AF established by closely monitoring closed loop itself both CLI & AFF. You also have to know when and when not to validate data in your logs at light loads and transients as on off throttle & decel have quite a bit different activity going on. Functions such as decel fuel cutoff, reversion, & wall wetting will greatly affect live AFR. If you post a 20+ minute log with the default channels plus WBO2 Front & Rear (not in AT mode) we can take a look at your tune, your combination shoud be simple to get tuned in.

 

Thanks Jamie- Again all very good info to have posted here- Especially given that I plan to do a cam soon (probably a 204 or 57h), along with 50mm TB and 4.9 injectors. So yeh I got it, there is not one formula that fits all- regardless of tuning device the end result will come down to the experience of the tuner interpreting the log data.

I have a 30 min ride log from a few minutes ago but it is with TT-AT enabled. I'll get some regular riding in this weekend. Roy @ DJ sent me a modified tune file I've been using that leaves the injectors on at decel- which immediately cured by recurring shift pop / cherry bomb. so thumbs up on that one!

 

unrelated question- is there any trick to getting the TT-AT spark learning feature to work? I've tried everything and it does not seem to function. instructions and guidance on use are sparse.
I ran some tunes with purposefully inflated timing that I knew would cause some 3 to 4 knock events and no hits were ever detected. in the TT-AT ride logs, it clearly shows the knock events right where I expected them. For AT settings I have min count set to 1, min spark learn set to .50 and max to 10. Do I need to manually clear the adaptive knock retard table to zero first? i did not try that.

 

To be honest I dont generally use the spark correction feature and have not tested it for quite some time, I write all my spark tables manually using a combination of datalogging and specific dyno procedures I came up with over the years. Next week I will take a look at it.

 

Thanks Jamie- My angle on this is to come up with a way for the average PV user to tune their spark tables as close to possible to MBT. We have no tools at our disposal to do so other than the long path of learning to understand log data and then manually change / re-test / change / re-test etc (arduous process).

Thinking, if the PV spark learning feature can trim excessive spark from learned cells, then we should be able to bump areas of the spark tables higher than needed, run a spark tuning session, and let TT-AT trim as needed. where it is trimmed should be just before points of detonation and (hopefully) close to MBT. of course keeping a safe max retard value of 10 so as to not cause any actual harm in the process.

otherwise I'm not sure what the intention of the PV's spark learning feature actually is? The ecm itself already has a learning process but that is more for safety reasons than performance optimization.