Closed loop - wide swings

FlashPro questions & answers specific to the 2006-2009 (US, UK, Asia) S2000
Post Reply
jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Closed loop - wide swings

Post by jun1or »

I searched before posting.

Background:

06 s2k drive by wire, returnless fuel setup, PTE5862 turbo on sidewinder, e85 fuel, KB Boost-a-Pump.

Tuner never installed oem wideband, so I spent a month in open loop, with zero issues believe it or not. Very good part throttle tune. But I obviously know this is far from ideal, so I added a bung and my OEM primary o2, enabled closed loop, and, wowsers the s.trims swing hard up and down (at times). I ended up changing the s.trim min/max limits to -8 / 12 and l.trim min/max limits to -6 / 6, respectively, as to tame the swings in s.trim. (l.trims always looked good, nothing too crazy)

Questions:
1.) Any chance we could get "closed loop rate of change" tables like in s300 for FlashPro? Would that help even?
2.) Any suggestions for taming this issue? It's so bad at times that I can feel the car get a smidge jerky from the wide AFR swings in light load (see attached datalog, timestamps 7:15 - 8:15 for an example of exactly what I'm referring to. (also picture below)
3.) By limiting s.trim min/max values, does that mean l.trim values will change faster since I'm hitting the cap of s.trim, or does it not work like that!? I'm curious how l.trim values are calculated, retained, reset, etc. also
4.) Could this be fuel related? e85 here.

Any suggestions would be wonderful.

Mike

Image
Attachments
S2000 v29 smoothened.fpcal
(9.34 KiB) Downloaded 221 times
08-29-13 closed loop wide swings datalog (720-820 timestamps).fpdl
See timestamp 7:15 to 8:15 for example.
(66.3 KiB) Downloaded 225 times

User avatar
Hondata
Site Admin
Posts: 9278
Joined: Thu Aug 15, 2002 12:13 pm
Location: Torrance, CA
Contact:

Re: Closed loop - wide swings

Post by Hondata »

The short answer is that the system, as a whole, as a slow response. It could be the o2 itself, o2 placement, an exhaust air leak or something else in the closed loop feedback system. Using E85 might be a factor too, but the o2 itself would be the #1 likely cause.

Make sure the injector dead times are correct. To do this drive in open loop with a very slowly reducing throttle until the injectors cut out. If the AF goes lean before cut out, the dead times are too low.

Also ensure that the fuel tables are well tuned in open loop.
Hondata

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Closed loop - wide swings

Post by jun1or »

Hondata wrote:The short answer is that the system, as a whole, as a slow response. It could be the o2 itself, o2 placement, an exhaust air leak or something else in the closed loop feedback system. Using E85 might be a factor too, but the o2 itself would be the #1 likely cause.

Make sure the injector dead times are correct. To do this drive in open loop with a very slowly reducing throttle until the injectors cut out. If the AF goes lean before cut out, the dead times are too low.

Also ensure that the fuel tables are well tuned in open loop.
You may be correct with the injector dead times. It's on the dyno right now getting slight adjustments. I'll post back results as I discover them. Thank you for the suggestions.

Mike

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Closed loop - wide swings 2104 update

Post by jun1or »

Tuner did not adjust injector opening values back in 2013 when I last replied. I've been trying to track this thing down since 2013!
  • 1.) I checked the exhaust for air leaks a couple weeks ago, seemed fine.
    2.) I replaced the primary o2 sensor with a new OEM honda sensor, no change.
    3.) The other day I induced large electrical load and the swings got large again, so I figured 12v dead time value too low or high.
    4.) Video of above - http://youtu.be/rb08UKTIbJo
    5.) I tried to increase injector opening by 50us-100us increments, but didn't run well, high stft's etc, so I backed it out (should have tuned it out and kept testing, I forgot that you need to retune after changing injector opening times)
So.... Since last post I have converted to a return style fuel setup. Tonight I upped my base fuel pressure from 43.5psi to 60psi and increased the injector opening time values (see below) and it seems to be better.

I ended up increasing opening times as follows: (2006 S2000 Turbo, ID1000's or maybe just bosch900's per p/n & google, hmmm)
  • Injector opening values with 43.5 psi base fuel pressure
    Volts 6 8 10 12 16
    Time (us) 3402 2082 1456 1212 812

    Injector opening values with 60 psi base fuel pressure (seems better at first drive)
    Volts 6 8 10 12 16
    Time (us) 3900 3550 2156 1598 986

I realize this is getting to become a long post, but I did read the following a million times: I do have a question. I may be doing something wrong, but I have a hard time inducing overrun condition. I seem to remember this being an abrupt thing in 1st gear at very low speeds with s300, not 96kph per topic 16235 linked above! Regardless, tonight I have datalogged 4 "overrun" condition (I believe), in an effort to test dead times. How does it look?

Thanks,

Mike
Attachments
S2000 (e85) v56 low cam tuning.fpcal
(9.43 KiB) Downloaded 144 times
06-06-14 fuel overrun injector dead time test.fpdl
(114.26 KiB) Downloaded 147 times

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Closed loop - wide swings

Post by jun1or »

After increasing the dead times (and retuning to near perfection as always), the vast trims swings and hesitation went away. I'll continue to monitor. Hope this helps others.

FYI to readers: set S.Trim and L.Trim min and max to all zeros as to disable closed loop, and carefully tune, then re-enable. I found that when dead times were off, the delays and swings were too vast that it made tuning difficult and time consuming.

Mike

bunger78
Posts: 9
Joined: Tue Feb 19, 2013 12:41 pm

Re: Closed loop - wide swings

Post by bunger78 »

jun1or wrote:FYI to readers: set S.Trim and L.Trim min and max to all zeros as to disable closed loop, and carefully tune, then re-enable. I found that when dead times were off, the delays and swings were too vast that it made tuning difficult and time consuming.
I do this exact same thing and it really helped speed up my tuning, even when injector latency is dead on.

jun1or
Posts: 137
Joined: Tue Apr 14, 2009 10:09 am

Re: Closed loop - wide swings

Post by jun1or »

bunger78 wrote:
jun1or wrote:FYI to readers: set S.Trim and L.Trim min and max to all zeros as to disable closed loop, and carefully tune, then re-enable. I found that when dead times were off, the delays and swings were too vast that it made tuning difficult and time consuming.
I do this exact same thing and it really helped speed up my tuning, even when injector latency is dead on.
Yep!

I know the trims are compensated for in settings, but it must have something to do with the delays.

It would be nice to have a checkbox to disable closed loop, like in sManager/s300.

Post Reply