Low gear deceleration issue

[martin7937]

ECU P/N 23710-70J00, European P10 SR20DE highport, using 32kb bin files, type 3 board, 370cc injectors + P11/N15 MAF.

I can't fix an issue while decelerating on gear, especially on 1st and 2nd (less noticeable but still there on 3rd and 4th). The issue itself could be described as a massive engine jerking.
The video does not clearly show the problem but if you pay more attention, you will see/hear what I am talking about.

Here is an explanation of what you will see on the log files:
1/ engine at idle
2/ then 1st gear and acceleration up to 2500/3000 RPM
3/ then 2nd gear and acceleration up to 2500/3000 RPM
4/ and then I let the engine decelerating on 2nd gear
5/ at 1500 RPM something happens, a table (unknown to me, I can't trace it) kicks on and here is where the problems start
6/ between 1500 and 1000 RPM is where the massive engine jerking appears

It is most noticeable while decelerating but, if it happens that I press the gas pedal while between 1000 and 1500 rpm, a massive acceleration engine jerking also appears, however for a much shorter time.

Below is the bin, together with 2x logs and a video.
The video is not high quality and it is quite quiet. Below 1200 rpm you might notice a knocking sound - ignore it, it is the busted lower engine mount (result of the problem I have).

Tried everything I could think of with no result. Please let me know if you have any suggestions. I feel like there is a missing table which triggers this odd behavior but I can't seem to be able finding it.

https://www.youtube.com/watch?v=c6u15Ed8OP8

[Matt]

That points in the log (need time values) are you seeing the jerking occurring?

what are the AFRs like?

I'm seeing lean on deceleration during the fuel cut area (File > Configuration > User mode: intermediate to view fuel cut maps)

Is this causing the problem?

[martin7937]

Looking at the log file ending 33, focus on time from 00:00:20:327 to 00:00:30:202
You can see the TPS value during that time being 0.46 which is my idle TPS position. This is a 2nd gear deceleration.

The ECU reacts as follows:
1/ at first there is an injection time of 0.75 leading to 20:1 AFR (should I see any injection time while decelerating at all?!?)
2/ somewhere around 1500 rpm an unknown to me table is being triggered leading to AFR 14:1 (I can't see any table turning green showing the interference)
3/ the massive jerking occurs right there, from 00:00:26:005 and it remains until I press the clutch pedal
4/ as a result of pressing the acceleration pedal again (00:00:30:202 --> see the TPS changing from 0.46) the jerking appears but this time in acceleration. However, the issue while accelerating is less noticeable but still there.

It feels like I am changing from one table to another and there is no smooth transition between them.

I am always using the intermediate user mode. Tweaking with the fuel cut tables didn't lead to any result at all.

[Bernardd]

What is your timing during the decell issue? Try adjusting it to see what happens.
Do you have a lightened flywheel?

[martin7937]

What is your timing during the decell issue? Try adjusting it to see what happens.
Do you have a lightened flywheel?

The timing during the decell was really wrong. In the log file you can see it decreasing from about 15 degrees down to 4. Thanks!
I got it fixed and below is how it looks now. What I did was to set all values in the Idle/Deceleration Gear Timing vs RPM table to match (as close as possible) the values in the primary timing table. It now runs a bit smoother but there is a side effect - the RPM jump a lot when I press the clutch and if I am under 2000 RPM. Anyway, this is easy to fix.

Is it normal that I see injection time values during deceleration? And besides the TTPmin, is there another table dealing with the injection time when TPS is ON?

Yes, I have the fidanza flywheel. I know it could cause some issues due to the lower weight. But it should not be that noticeable.

Another question - is it possible to tune the AFR during deceleration?
Tweaking with the Fuel Cut/Fuel Recovery values (I lowered them) made things better. But I think I am running too rich(14:1) when the fuel recovery table kicks in. Probably when this table kicks in, the ECU starts reading from the main fuel table, is this how it works?

[Matt]

There is no enrichment on decel. It uses the main fuel map for all fueling. Enirch Idle tables are only used for warmup

[martin7937]

How about my other question - is it normal that I see injection time values during deceleration? And besides the TTPmin, is there another table dealing with the injection time when TPS is ON?

[Matt]

Yes it is normal to see injection time on decel at certain points where there is minimal injection and then it restores

TP min is minimum load and TP max is maximum load values. These are load limits to stablise MAF voltage fluctuations

[martin7937]

Nothing helps, the problem is still there. No matter what I do, no matter which table I play with, the problem is there.
1/ A new first load column was created in the fuel table with TP value matching the load during deceleration. No matter how I change the values in this column (no matter the AFR), the problem is there.
2/ Tried to play with the TTPmin - no result.
3/ All tables related to fuel cut and fuel recovery were tested in different combinations - no result. I even tried to lower the values in the fuel cut tables down to 1000 rpm. But once I hit 1000 rpm, the problem occurs again and it is even worse.
4/ Tried different timing values, both in the main timing table and the low/high gear timing tables, still nothing.
5/ Changing the Latency and K makes no difference at all.

I am really running out of options here.

I will appreciate if anybody could at least try for a moment get into my problem. There are logs, the bin is there, vids, clear explanation on what is happening.

Well yes, this is not a RB engine, it is a darn N/A SR. But if there is a board being sold for such application, there should be available support for it as well...

[skylinegtrhr]

Did you do compresion test? Those 370 are I hope oem s14? Can you put different cas from car which run fine?

[martin7937]

Compression test shows equal readings on all four cylinders (~11.75 with JWT S4 cams in), with differences not exceeding 0.25.
The 370cc injectors are indeed OEM S14, tested, working great, removed from a half cut with less than 75k miles.
Last September I used for a week a distributor from another stock SR20 P10 which is still working great - no result.

[skylinegtrhr]

And when this start to happen? Did you change anything on setup? I presume that you did map car on dyno before?

[martin7937]

I had this problem ever since I started dealing with nistune. I first purchased a type 3 board installed in a 83J ECU, which was using 64kb bin. Totally useless ECU, especially having the wrongly working VE table - as far as I know - an issue still on the list to be fixed by Matt. So 3 or 4 years now.
Couple of months ago, I bought another ECU with a nistune board, this time the right one - using 32kb bin. I was hoping to finally sort all things out. And yes, finally I was able to tune AFR, timing, K and etc. My car runs extremely well right now but, when rpm are higher than 2000.
No changes to the setup for the past more then 2 years.
All recent mapping was street done, no dyno. I tried to dyno my first ECU but only wasted my money because of the faulty VE table.

[Matt]

I had this problem ever since I started dealing with nistune. I first purchased a type 3 board installed in a 83J ECU, which was using 64kb bin. Totally useless ECU, especially having the wrongly working VE table - as far as I know - an issue still on the list to be fixed by Matt. All recent mapping was street done, no dyno. I tried to dyno my first ECU but only wasted my money because of the faulty VE table. So 3 or 4 years now.

There is no issue with the VE map in the ECU itself. When you fit an ECU it comes with factory mappings loaded. This means it will work identically as the ECU did without the board fitted.

The problems identified at the moment with the VE map are not to do with its function but the TPS tracing is a few cells out and I'm looking into it at the moment. Previously it was thought that VE maps were TP based but after further investigation I found that not to be the case. Usually we have not even needed to modify the VE map say on my RB25DET and just changed the fuel map instead.

Couple of months ago, I bought another ECU with a nistune board, this time the right one - using 32kb bin. I was hoping to finally sort all things out. And yes, finally I was able to tune AFR, timing, K and etc. My car runs extremely well right now but, when rpm are higher than 2000.

Just looking over your posted tune file and it does not right at all compared to the factory files originally loaded in the ECU.
- Your fuel map is flat value of 10 across the board after 3500rpm
- Your timing map is also flat value of 10 across the board after 3500rpm
- Your load scales have been cut off also

I would suggest putting those maps back to factory and then pull back the load scales a bit back since they should be reduced about 30% from current values

Your K constant has changed from 31206 to 21000. This is resulting from changing the MAF from the P10 SR20DE to P11 SR20DE MAF (should be about the same flow figures) and from 259CC to 370CC injectors (which is the main change)

Because the K constant has been reduced by 32% this affects any TP referenced table or values. This is a side effect of changing K by too much. That also includes other things like cold start, closed loop operation etc. What I would suggest is a MAF upgrade like Z32 MAF which will bring K constant back higher

I'm looking into the TP affected tables in the meantime as well as adding an extra injection multiplier to the Nissan code for various models to avoid moving K constant around

[martin7937]

Your K constant has changed from 31206 to 21000. This is resulting from changing the MAF from the P10 SR20DE to P11 SR20DE MAF (should be about the same flow figures) and from 259CC to 370CC injectors (which is the main change)

What I would suggest is a MAF upgrade like Z32 MAF which will bring K constant back higher

Having the injectors resized from 259cc to 370cc, I end up with K value 21844. Here is how it changes after changing the MAF:
1/ no change for the P11 SR20DE MAF, no change for the Z31 VG30ET MAF
2/ K 33096 for J30 VG30DE MAF (should be E60)
3/ K 36406 for:
- the Z33 VQ35DE MAF, being the Nissan 350z MAF,
- A32 VQ30DE MAF (should be N60)
- Q45 VQ30DE MAF
4/ K 41040 for the Z32 VG30DETT MAF, which I believe is the N62 MAF

I am willing to give your suggestion a try.
However, having the above in mind, isn't it better to go after J30 (E60) instead of the Z32 (N62)?
The Z32 will bump +10000 the original K over the stock value.

Just looking over your posted tune file and it does not right at all compared to the factory files originally loaded in the ECU.
- Your fuel map is flat value of 10 across the board after 3500rpm
- Your timing map is also flat value of 10 across the board after 3500rpm
- Your load scales have been cut off also

^^^All of the above was changed on purpose by me so we can focus on the problem area only. Currently the very last right column shows TP 55, being 33% less than the original TP (80).