power sag at 3200 rpm

[3p141592654]

7MGTE with CT26 running at 11psi. Stock fuel system, 75mm exhaust, cartech intercooler.

I feel (and datalog) a sag in power that start at 3200 rpm and last until 3600 rpm. It is subtle, not a collapse, but just doesn't make as much power. After 3600 rpm power stays solid to redline.

It happens about 0.5 second after the ecu switches open loop, and it happens long after the fuel pump switches to high speed, so I don't think it is due to either of those things.

Has anyone else noticed this. It is defintely more noticeable as the boost is increased. Datalog is attached.

 

[suprarx7nut]

I'm on a 50 trim ct, 3" exhaust turbo back, stock ic, 11 psi. I don't notice any power sag.

Your datlog is interesting. It would be neat (and maybe meaningful) to plot various parameters vs load (load as the x axis). Using excel you can easily get the correlation value too, if its not visually obvious.

You hit the sag right when timing hits the lowest point. Knock also seems to hit a high point.

 

[toyotanos]

Couple things stand out to me, although they may be nothing more than the interface.
1- TPS reading 24%. If the ECU is only seeing 24% throttle, it may not be as aggressive with timing as it could be, not sure how the stock ECU treats tps in relation to timing and fuel though.
2- 18* of timing seems a bit much, even for stock + a lil extra boost. If your timing is even a little off, that could send it straight into knock territory. More boost=more knock.

It might be worth while running a wideband to see what the AF ratios are doing during the sag period.

 

[hvyman]

If this is under full boost I'm thinking the fuel pump switch over. If full boost and second gear it's rather quick to go from 3200-3600.

Timing seems fine to me. Even at 11 psi. Stock turbo doesn't push out all the much air and if that was na or cruising timing would be at about 30 or higher.

Unless timing was higher and the Ecu detected knock and backed it off and that's where you felt the sluggishness.

 

[3p141592654]

Thanks for the suggestions. WOT is 60% in this datalog, so 24% is about half WOT. From the seat-of-the-pants it is making pretty good power. There is definitely substantial timing reduction due to detected knock, but no knock is audible to me. I agree a wideband would be useful here. I need to get that installed if I am going to take this any further. The O2 sensor stays rich the whole time, so A/F is staying at least 14.7:1 or richer.

At 3200 rpm I am at 11psi (using an EBC) even at 50% WOT as logged here. The FP switches to high speed about one second earlier, so I see your point that maybe the actual rail pressure has not responded in time to the increased demands. You can see that once it goes open loop injector duration goes from 8ms to 12ms, so a pretty big jump in fuel demand. Half a second later it sags, so maybe it is a pressure problem.

I'll change the FP high switchover to be earlier, and see if it helps. Its good to know that they don't all do this. Means I have a solvable problem somewhere.

Interesting idea to plot vs Load. I'll give that a try.