LTFT on eManage Blue

[shnazzle]

Ooh,  might have missed that, was it in the spreadsheet file?

You must have!
It's under Fuel Status  s:) s:)  Values 2 (closed loop) and 4 (open loop)

[lamcote]

Oops, how did I miss that?

[lamcote]

I'll have another look.

[lamcote]

I've done some analysis to try to identify any trends, I'll email the sheet to you. It shows:
All rows highlighted in yellow are open loop data
All cells highlighted green in column X are where the MAF reading just changed by +/-10 or more
All cells highlighted green in column AD are where the TPS just changed by +/-4 or more

Do you notice any trend? I can't stare at it any longer for the moment!

One other thought, these logs seem to be logging data once per second. At 3,000rpm there are 100 combustion events per second, so we are getting around 1-2% of the total fueling data it, may be tricky to track down fueling trends with such a small sample of data?

[shnazzle]

Hmm... I've just checked and changed the sampling rate. It was set to 1sec so I set it to 0.1sec. That should be better.

We'll get there!?

I'll have a look later today at your work

[lamcote]

Do you know why there are 3 different fuel trims per bank?

[shnazzle]

I do not. I've always wondered this myself.

[lamcote]

My only guess is that there is a LTFT, a "proper" STFT and the third is logging the constant +/- AFR swing introduced to "drive" the O2 sensors. One of them does seem to go from +ve to -ve on a very regular basis.

[shnazzle]

Visually it looks like the following scenarios:
Low RPM + Open loop (status 4) = LTFT down
Low Rpm + Closed loop (status 2) = LTFT up
High RPM + Open loop = LTFT UP
High RPM + CLosed loop = LTFT down
WOT + Open loop / closed loop = random...all over the place. Can't see a correlation (yet).

Not surprisingly, more throttle = more timing advance. Not sure that's a massive revelation  s:) s:)  hehehe.
What is a revelation is why timing goes down in the middle of a climb in RPM?

How does this work?
Does the stock ECU have a theoretical timing map that it reads, and then applies a timing advance/retard on top of that?

[SteveJ]

Visually it looks like the following scenarios:
Low RPM + Open loop (status 4) = LTFT down
Low Rpm + Closed loop (status 2) = LTFT up
High RPM + Open loop = LTFT UP
High RPM + CLosed loop = LTFT down
WOT + Open loop / closed loop = random...all over the place. Can't see a correlation (yet).

Not surprisingly, more throttle = more timing advance. Not sure that's a massive revelation  s:) s:)  hehehe.
What is a revelation is why timing goes down in the middle of a climb in RPM?

Either knock detected or to adapt to the change in cam timing.

How does this work?
Does the stock ECU have a theoretical timing map that it reads, and then applies a timing advance/retard on top of that?

No it does it by magic (or smoke and mirrors)   s:lol: :lol: s:lol:  

It uses a timing map in exactly the same way as the piggyback does with offsets added / removed in response to knock being detected.

[SteveJ]

One other thought, these logs seem to be logging data once per second. At 3,000rpm there are 100 combustion events per second, so we are getting around 1-2% of the total fueling data it, may be tricky to track down fueling trends with such a small sample of data?

The Bosch sensors (and at a guess most other mfr's) have a response time of ~100ms (or a max of 10 events per second) so you wont be getting any more data than that anyway. They are only used to detect trends in the exhaust gas, not the result of every exhaust stroke.

[shnazzle]

Going to have to dig further into the potential that it's pulling back timing based on knock values.
That's obviously not favourable.

Does it mean this or is there another explanation?

Can the airflow adjustments also tamper with the timing directly? Or only indirectly via causing knock?

[SteveJ]

It's a multi-dimensional map which gives a timing value based upon multiple inputs (air temp, air flow, water temp, throttle position, engine rpm and possibly O2/fuelling as well). This value is then used along with an overall knock-detected offset adjustment.

Use of the knock sensor is essential for Toyota as they don't know what grade of fuel is in use (can range from ~91 RON in Japan/US to ~100 RON in Europe). It's the reason that it takes quite a while to get the benefits from Shell Optimax etc - the ECU will pull timing advance very quickly in response to knock but then add it back VERY slowly when it doesn't see any. This is one of the contributing factors resulting in better performance after an ECU reset.

How ECU's work in the real-world is very much a black art - hence why good tuners make mucho dinero!

The Lotus ECU in Izzy (my Elise) is base-tuned to 97RON (with a BIG sticker in the fuel filler alerting you to this) and it really hates having anything other than super-fresh fuel - I've had problems with fuel from garages that aren't very busy and have stale Super-Unleaded in their tanks.

[stupink]

kinda half read stuff because I just got in and want to go bed but found myself on laptop on here :p  but think comments relevant, ignore if not..   det happens easiest in the mid rev range (from experience I find 3,500rpm about the usual place to get misfires etc). so that's probably why ignitions going funky around there. just preventative measures..     I think perhaps you're interpreting the figures wrong too?? if not then again ignore, but sounds slightly like you're reading -5 on ignition as "read base figure, don't run that run it -5.."  but its not like the fuel trims, they're not ajustments typically displayed, they're actual figures of whats being run.. unless you're looking at something other than what i'm thinking.. the ultimate for example can show requested and adjusted ignition figures.

[shnazzle]

Next on my list of PIDs to find... Knock detection.
I know it's there as I saw that Steve's funky handheld device showed knock.

I had no knock at all at idle. But if it's knocking at full throttle then I want to tone down my timing advances

[SteveJ]

He's reading from the OBD2 so it's the absolute figure that the stock ECU is trying to deliver. I can't recall how the Blue & Ultimate do their timing adjustments - if it's the same as the Dastek then the TDC sensor is intercepted and modified to trick the stock ECU so it doesn't know that the engine is actually a few degrees away from where it thinks it is based on the piggyback map.

[shnazzle]

kinda half read stuff because I just got in and want to go bed but found myself on laptop on here :p  but think comments relevant, ignore if not..   det happens easiest in the mid rev range (from experience I find 3,500rpm about the usual place to get misfires etc). so that's probably why ignitions going funky around there. just preventative measures..     I think perhaps you're interpreting the figures wrong too?? if not then again ignore, but sounds slightly like you're reading -5 on ignition as "read base figure, don't run that run it -5.."  but its not like the fuel trims, they're not ajustments typically displayed, they're actual figures of whats being run.. unless you're looking at something other than what i'm thinking.. the ultimate for example can show requested and adjusted ignition figures.

I'm reading is at -5 is the sum total of the timing in the table(s), the EMB and the stock timing adjustment.

Unfortunately the Blue only shows the adjustment.
I'd have to subtract the adjustment from the obd2 total to get the stock ECU value. But only if the logs are 100% sync.. Which they're  not

[SteveJ]

kinda half read stuff because I just got in and want to go bed but found myself on laptop on here :p  but think comments relevant, ignore if not..   det happens easiest in the mid rev range (from experience I find 3,500rpm about the usual place to get misfires etc). so that's probably why ignitions going funky around there. just preventative measures..     I think perhaps you're interpreting the figures wrong too?? if not then again ignore, but sounds slightly like you're reading -5 on ignition as "read base figure, don't run that run it -5.."  but its not like the fuel trims, they're not ajustments typically displayed, they're actual figures of whats being run.. unless you're looking at something other than what i'm thinking.. the ultimate for example can show requested and adjusted ignition figures.

I'm reading is at -5 is the sum total of the timing in the table(s), the EMB and the stock timing adjustment.

Unfortunately the Blue only shows the adjustment.
I'd have to subtract the adjustment from the obd2 total to get the stock ECU value. But only if the logs are 100% sync.. Which they're  not

What you need is the Profec e-01 (actually a boost controller) - it show the real-time absolute timing values from the Blue (and Ultimate) (along with TPS, airflow, temp etc) and can log it all to an SD card.

[shnazzle]

I can see now how people end up with a profec, wideband, ultimate, map (not me obviously)  etc...

And the next step to a link, apexi, etc is a small step away.

I've added a misfire PID to my logs now. Might be interesting to see if i get and and if so, how many.

Obviously I'd rather have none

[SteveJ]

A quick look at the Toyota data-stream analysis done by a few people seems to indicate the knock timing adjustments are returned in PID 21B2 (http://www.mthrfknwin.com/diagnostic-port-obd2-detail-manufacturer-specific-pids/) but it's not a straight-forward value (see the URL for more detail).

[shnazzle]

Oh that's different from the one I found (01F3). I'll have a read of that

[shnazzle]

Ah I think 21B2 is already being looked at. That's the timing value in the logs in think.

01F3 is the number of misfires

[shnazzle]

My new Torque Pro live screen.

Then monitoring everything in the background to file at 0.1ms intervals.

Obd2 has been flakey last week so I might have to try another adapter. I want a USB one actually

[SteveJ]

Ah I think 21B2 is already being looked at. That's the timing value in the logs in think.

01F3 is the number of misfires

21B2 is the timing correction based on knock - if you're seeing that changing then knock is being detected (which is quite normal but shouldn't result in excessive changes).

Basic timing advance is a standard OBDII PID - list here https://en.m.wikipedia.org/wiki/OBD-II_PIDs

[SteveJ]

Then monitoring everything in the background to file at 0.1ms intervals.

Obd2 has been flakey last week so I might have to try another adapter. I want a USB one actually but can't find one

Don't blame the adapter too readily - the ECU used in the '2 is limited in the amount of data it will serve up on the OBDII connector before you start getting timing errors. Sampling more than a couple of PIDs at 10Hz is going to overwhelm it. (P.S. I think you mean 0.1 seconds not milliseconds as that would equate to 10,000 samples/second which is way outside what OBD is capable of  s;) s;)  )