The current firmware revision is pre-1.1 beta 19. PID values and settings may change as the firmware changes.

What are y'all running for VVT settings?

I have noted that Revereant scoped the stock ECU output at 285hz.
I am currently running at 253hz. I started at 153hz, as JasonC was running ~100hz on his AEM, but my solenoid buzzed quite loudly.

Adjust injector timing: Off - I'm thinking I should turn that setting on because I am changing when the valves are opening and closing.
Minimum angle: 277.5
Maximum angle: 323.5
More Duty=advance
Use Hold Duty: Off

My PID settings are all 50. I have played with them for an hour and gotten nowhere. I want to see other people's settings so I can ballpark them and go from there. The default 50,50,50 causes such spasmodical oscillations that my 92 OPG bounces like it would with detonation.

Targets:

This is what I interpolated from the basemap of the VVTtuner. Y8s have given me some input for changes I need to make to it. No advance over 5000 rpm. Taper from full advance at 3500 to none at 5000. No advance below 3000 and 70kpa.

My current setup is stock VVT motor with VICS Intake manifold. I/H/test pipe with stock catback. I have a 1N4001 diode, banded end towards 12V, soldered in right at the VVT solenoid.

 

First, my settings:

-- VVT Settings --
Frequency: 306 Hz
On/Off or PID: PID
Adjust inj. timing based on: Cam 1
Commanded vs. actual: Actual

Control Interval: Synced to cam
Range check measured angles: Yes


--PID Parameters--
Intake
More duty means: Advance
Use hold duty: Off
Porportional Gain: 50
Integral Gain: 17
Differential Gain: 130

-- Cam Settings --
CAM1
Input: Main cam
Output: Nitrous2
Minimum: 275.0
Maximum: 319.0
int.exh.: Intake

Map:

 

I know Lars benefited from adjusting the rpm points a bit for his vvt settings. I hope he posts it; we used yours as a base.

 

253 is not an option?? Note that if you run Idle on Idle and VVT on VVT, you have to run the *same* frequency for both. I moved VVT to Nitrous 2 as shown above to avoid this. I run idle now closer to 500 and vvt closer to 300.

Turn injector timing ON. You want the end of the squirt to be in the same relative spot compare to intake valve opening.

just like tuning any other PID, start with D at zero and P at default and I at zero. slowly increase I until you are able to maintain the target accurately once things settle down. Don't worry about transients until you get steady state accuracy. Once you get I, you can raise or lower P to get response improvements. set it either to not overshoot at all and you're done or set it to get a TEENIE TINY bit of overshoot and then raise D to eliminate the overshoot. If you raise P I or D to the point of oscillation, back off and set it to the highest number below oscillation.

My experience is that VVT is almost completely RPM dependent. I guess it's related to what Jason C said in the intake thread about volume flow versus mass flow. Anyway... I saw no significant changes in performance by advancing or retarding with respect to boost. There may be some benefit, but I didn't measure any in my tests. Not that it matters for you N/A.

Like you saw me say before, keep things fully retarded at idle. Notice I have mine set to actually below zero to force the cam to full retard.

Also I mentioned not running any/much advance below 70kPa. We probably don't have the means of testing the effects very well of advancing the cam during cruise, but it can lead to reversion since there is significant overlap at the highest advance numbers. Even at full retard, the intake valves are just starting to open right as the exhaust valves close.

If you want to try to get more MPG, you can do some back to back freeway tuning and advance in 5 degree increments and see if the fuel usage rate increases or decreases. it may be hard to measure though. telltale positive signs would be (for same cruise RPM), higher TPS and lower PW while still maintaining speed (rpm)

good. it'll prevent frying the flyback circuit.

 

Y8s, why the negative "force-to-zero" numbers at idle? Doesn't that create a delay and then maybe overshoot in response when the duty shoots up as you come off idle? Not that a delay would be very noticable here, but it's just a thought. Why is that better than say 0.2* at idle or something?

 

What I don't get is why the switch points are so rapid. Why have VVT when you could have vtech, then? There's no analog going on here (well, there's a bit in your 12/19/whatever) but I'm surprised you don't see slow movements from one place to another. I guess when you switch past a resonance peak you need to jump to the other side to pick up the next, but in windows I would think you'd want to follow the pressure waves.

Maybe I'm on the crack.



Someday, I really will make a "efficiency" gauge which is basically fuel per mile. I've always got some hare brained scheme for how I'll get better mileage and I never have a good way to test it. I've thought about just watching duty cycle, rpm, all sorts of things, but really why complicated it? Effective pulse width / VSS should do.

 

I didn't spend that much time tuning. It could be worthwhile to do more exhaustive testing and find the slight varations that produce more power or mpg, but since I don't have a miata, someone else will have to do that.

Also, it's analog likely for reasons beyond pure power. It could also be an emissions thing. I know VVT is used in some cases to keep combusted material in the chamber similar to EGR.

 

Bryce, what's the base map for VVTuner look like? Do you know how it was developed?

 

From the tuning "guide":

It looks similar to the one I got loaded in mine, but if it have developed over time I dunno.

How it was created is another matter.
"The values preprogrammed to VVTuner were found to work well with a standard 2001-2005 Miata VVT motor after extensive research."

 

crap

 

It was tuned on our dyno using Kevin's C/Street Prepared Miata autocross car.

 

So no f/i vvt tuning/development?

Would you recommend any changes to the table if used in a f/i car?

I realize the real way is to hop onto a dyno and find out what works, but not all of us can just up and do that.

 

Matt C (or anyone) do you have PID/valve settings significantly different from Y8s?

 

My experience with MS3 is that the PID, base duty, offset and a few other things are engine specific. We'll start with a map from another similar (or identical) engine then tweak from there. Different oil viscosity, component inertia, solenoid resistance, butterflies in Brazil..

Copied from Y8S post with my comments/observations in blue

-- VVT Settings --
Frequency: 511 Hz we're using a different output than Y8S
On/Off or PID: PID
Adjust inj. timing based on: Cam 1
Commanded vs. actual: Actual

Control Interval: Synced to cam
Range check measured angles: Yes


--PID Parameters--
Intake
More duty means: Advance
Use hold duty: Off - I set this to off when done but the value in grayed out, supposedly disabled, window affects target. So set this to on, select value then set it to off. Takes a few tries to get a stable target vs actual at idle. No idea why this works this way but it does.
Proportional Gain: 50- varies
Integral Gain: 17 - varies
Differential Gain: 130 - varies. I have never needed more gain than about 30 to get good targets

-- Cam Settings --
CAM1
Input: Main cam
Output: Nitrous2
Minimum: 275.0 - varies
Maximum: 319.0 - varies. I always make sure the spread is 45°. Y8S is 44° here
int.exh.: Intake

 

If it's 44, then my car has a range of 44. Since those numbers are determined experimentally and based on physical limits of the cam gear guts, I used what I measured.

 

Recently, we've fitted a VVT motor with a Flyin' Miata trigger wheel and had James put together some experimental code to decode the FM trigger with a Miata cam sensor. This is giving us more accurate data on what the cam is doing and we are using this to improve the VVT control. However, the final results of this development are still not out yet.

 

Good job, math/engineering crew. Keep up the good work.

 

For which of your ECU's?

 

probably the ms3x firmware in general.

 

Glad to see that, I'm running 511 also. :-)

Which means... everyone's table should be offset by the min angle difference from their's to someone else's.

I did a super-quick, @ idle check and the numbers were bouncing so I only got 41 degrees of spread. I'll try again at higher rpm, I just wanted to see if the solenoid was working.

You're saying that the number is important, but you need it to be off? You find this number by holding some steady rpm then tweaking the test mode PWM up and down until you get it to stick somewhere in the middle?


I'd be curious to see one set you've used.

Like I said, I'll have to recheck mine. I got 269/310 - I really don't trust the 310 number.


I, too, am excited to see variation across boost levels, not sure the dyno access I'll get when I get it.