Mods: Please leave this in Racing & Comp. This is directly applicable to those tuning AP1s for STR class. Moving this to another forum will lead to everyone just bitching about how much the VAFC sucks, advice to buy a 'real' standalone, and complaints about how the rules suck and should be changed for their wants.

I finally got my wideband O2 sensor along with the VAFC2 and harness installed. I went for a short drive and logged the AFR and various other parameters with no changes at all in the VAFC2. Then I dropped VTEC engagement to 4000 rpm and did a couple of pulls. Data was collected with a Race-Technology DL1.

Black = throttle
Red = RPM
Orange = AFR
Blue = Longitudinal acceleration

2nd gear pull:


1st gear pull:


Looks too lean around 4k, so I've since changed the crossover to 5k rpm to be safer. As I understand it you can't add fuel with a VAFC2, so that means that dropping VTEC below 5k is a little too risky with only a VAFC2. Any thoughts on that or anything else about these graphs? Any thoughts for parameters I should log? There is a pin for knock on the ECU, is that just going to confuse me or will it be helpful?

Note the extra fuel from 5-6k and the corresponding dip in long. accel. My first step will be to reduce fuel in the 5-6k range and also closer to redline. As I understand it I should target a 12:1 or 12.5:1 AFR. This is my first time doing any engine tuning so I'm going to approach this cautiously.

For reference - powertrain mods to this car:
- K&N FIPK intake
- Berk 63mm high-flow cat (with STR-legal secondary O2 sensor bung and second O2 bung for wideband welded upstream from the cat)
- Innovate LC-1 wideband O2 sensor/controller
- Stock exhaust for these tests, will need to check with shorty exhaust later

 

You can add and subtract fueling by %'s with the VAFC. Theres about 15 defineable RPM points to correct fuel and it interpolates between the points.

It looks like there is power to be gained lowerring VTEC but I haven't tried it because of the potential issues and I don't understand VTEC unmatch.
I never messed with the partial throttle maps either just 80% throttle.

On a 100% stock 03 with the cat on I made about an extra 4HP up top leaning it out at the top RPM's.
With a testpipe it made about 6-7HP from VTEC. My tuner targetted 13.3:1 AFR measured at the exhaust.

 

13.3:1 AFR is generally assumed to be the Lean-Best-Torque AFR, but things are starting to get a little warm there. I don’t know that I would track a car running 13.3 AFR on pump gas, but autocross ‘should’ be fine. I don’t think you’d loose much torque running 12.5:1 but you’ll drop the temps a bit.

I haven’t seen it on the internet, but talking to Nick it sounds like there are some electronics issues on the S2000 if you try to add fuel. He mentioned something about how the VAFC circuit tries to increase the MAP value could (or did?) burn out the sensor.

The VAFC-NEO is on my car but I have only dropped the V-tech point to 5400rpm. Hopefully in the next week or two it’ll find its way to the dyno to get tuned. I’ll post up the results and corrections when it gets done.

On of my big concerns is managing the ‘learned state’ of the ECU. It is my understanding that the ECU tries to learn new VE tables based on closed loop operation. I ‘think’ it extrapolates those tables to full load operation, but I have no idea if it also tries to apply them to high rpm (I doubt you get a lot of closed loop 8k rpm info for the ECU). Since our ECU needs to be in a ready state to be STR legal, I can’t just pull the backup fuse and reset the ECU every day. So, make sure only to modify the fueling in the open loop range so the ECU doesn’t ‘correct’ the VE table for you, and make sure when you do the tuning you’ve gotten plenty of time on the ECU since the last time it’s been reset.

 

Oh! Since you have the wideband on the car, you can help do some experimentation if you’re feeling scientific and all.

-Reset the ECU, set the v-tech to a pretty low rpm and do a data log with very little driving around before the log (i.e. warm up the car before resetting)
-Drive around the block a few times between 4k and 6k rpm at light load (closed-loop, maybe 1st gear)
-Repeat the accel log and see if it’s running richer in the new v-tech range

 

The problem for me and why I didn't mess with VTEC is the partial throttle corrections.

Say you drop VTEC to 5000RPM's and add fuel to compensate for the thirsty high cam.
On the WOT map thats ok but you only need to be under about 80% throttle to be on the partial throttle corrections.
If you didn't tune the partial throttle correct you are going to be well lean here.

I think the "VTEC unmatch" setting has something to do with correcting this.

 

At what throttle position does vtec not engage? I was under the impression that there was a number of criteria for the CSM switch including throttle position.

 

My understanding from a bit of research is that the VAFC and its relatives only work when the car is in open loop (around WOT). No partial throttle tuning...but not such a bit deal on the track.

 

I don't know what the stock logic is, but for the VAFC you can set a MAP and throttle position to engage vtech in addition to the rpm. I left the MAP and throttle qualifiers off as I don't think I want the vtech actuating if I get off and on the throttle at high rpm.

 

Sure, I'll try this. No guarantees when but I'll give it a shot this weekend. My search skills are weak this morning, which fuse do I pull to reset the ECU?

As for engaging VTEC: The VAFC lights up red when the second cam is engaged. In my little test it definitely changed to red at 4000 rpm regardless of throttle position. That makes me wonder as the ECU is likely in closed-loop mode when above 4k at light throttle. I assume it will be trying to correct for whatever the VAFC is doing at that time, that's bad - I think...

I was thinking about resetting all VAFC parameters to have no corrections and log a bunch of pulls at varying throttle positions. That should give a map to show when the ECU goes into open-loop, indicated by AFR going from 14.7:1 to something dramatically lower. Is this a useful test? I'm hoping that it'll give me a chart that correlates rpm and throttle position to open vs. closed loop.

This seems like stuff that must have been done before by the various VAFC tuners so they could know which regions of the rpm/throttle map to avoid, but I can't find this info anywhere. Maybe the tuners are keeping their secrets for themselves? Or maybe they're just guessing?

Then I've got to figure out what the cryptic explanations in the manual for each of the parameters in the VAFC really mean. Does anyone have the "Operation Transition Diagram" as noted in the VAFC manual on the parts list on page 8? Is that useful to help figure out what the unit is doing? My used controller didn't come with that.

Nick - Care to share your experience with adding fuel? Anything I've found here on S2ki just said "don't do it" without any reasoning given. It's tough to know if that's an old wives tale that gets repeated as gospel or if there's a legitimate reason.

 

There isn't a problem with raising the MAP voltage perse - the problem is on the other end - the ECU doesn't know what to do with the higher voltage inputs or isn't rated for the higher voltage (I'm not sure which). The whole idea is that you are pushing the ECU into a different pressure column of the VE table.

So if you are already at max pressure that the ECU is capable of (or programmed for) by sending even more voltage at the ECU - you are theoretically running "off the map" and in my car it resulted in misfires and eventually limp mode. There was no permanent damage done as the car was fine when I removed everything. Jason Wong is currently have the same trouble with his 05.


With the VAFC2/Neo you can set a MAP pressure window just like you can in Flashpro. For those that want to play around with it:

My FP Window is from 4500-5500 (so no matter the pressure - the cam engages at 5500)

4500 = .950bar
5000 = .920bar
5000 = .900bar
5500 = .541bar

I don't really know if thats optimal but at partial throttle the car doesn't hit the second cam until 5500.