18psi

iTrader: (76)

Ok messed with it some more. I think the D actually goes from higher to lower like P, because using a high value acutally didn't dampen the overshoot/undershoot for me like I thought it would. Using a low value didn't either.

I dunno.. Maybe my MS is possessed lol .

Here's where I'm at as of today:



Its probably not ideal, but so far seems to work great for me. No bucking at all today, and doesn't seem to jigsaw as violently as it did before. It does still jump a few times on initial throttle tip in, then settles right in and holds targets somewhat smoothly.

I of course will be testing this more in the next few days, because when the car is cold is where the real test lies, and I only get one of those a day here. Once its warmed up its pretty much warm all day.

Bryce

iTrader: (24)

That's odd. I'd expect to manually tune a PID control loop the same way anywhere, whether its controlling VVT, or the field excitation on a 1500KW turbodiesel genset.

Vlad, lets see that log file!


I spent a little time tuning this yesterday. Then I came back to the same question I always ask myself when tuning: How good is good enough?

Actually hitting the target is an obvious goal, but how quickly do I need to hit the target? Within 500, 100, 50, 10ms? Now once I'm there, do I need to stay within 1 degree, 2 degrees, or 5 degrees of the target?

It would be great to have a set of definite response numbers to aim for so I can achieve those, then say my VVT tuning is complete and move on to the next thing. I understand that with the large amount of variables and conditions affecting the response of a given tune can cause it to vary greatly (oil temp, viscosity, rpm, etc) so maybe this is just wishful thinking.


Anyone have a log that would be considered a textbook example of an ideal response?

18psi

iTrader: (76)

This is not the most current, but you get the idea. Still goes a bit nuts between 2200-2800 right after throttle tip in (which is why I took out even more from the intake table down there), then "settles in" and seems to hold targets fairly smoothly within 1* of them.

Whats weird is that sometimes it will be DEAD SMOOTH, like butter, and other times it will be jumpy like that. Both times car is completely warmed up. I'm about to try it again today with car dead cold and see what happens.

*edit: I hear you on the "good enough" part - the rest of my tune seems to be great, and I'm really OCD about tuning the car this time. Everything has to be flawless. Well it pretty much is now, except for this one last part.

I have no idea how I managed to completely ignore what VVT was doing before and just now realized how wacky it is when you compare target vs actual. Oh well, guess I just thought it was actually doing what its supposed to do lol

18psi

iTrader: (76)

here's another from this morning. It still bucked a tiny bit when dead cold, so I completely took out any and all advance below 3k.

by that time it was already hot, so here's another hot log...

...seems to be pretty solid now when hot.

noname4me

I recently did an oil change and put Mobile 1 0w-40, instead of Rotella T6 5w-40 which I was using when I tuned VVT last spring (MS3X). Sure enough I can see in the logs that the VVT actual angle is oscillating even at idle, which I haven't had since tuning the PID settings 9+ months ago.

Have you found the VVT PID loop settings to be this fragile that oil viscosity changes require tweaking the values? Or will better baseline VVT PID settings enable a MS3 to control the VVT angle reliability even with oil viscosity and/or temperature changes?

emilio700

iTrader: (3)

I had not noticed it but it's not surprising. We change engines, rebuild, swap hardware and return so often around here it would be difficult to notice an A/B change from oil alone. I know that twice, we had two cars with just about identical setups that required different base duty, min/max and PID settings. I didn't analyze what was different, I just tuned it and moved on.

I would think that there should be a setting(s) that allows for a reasonable change in oil viscosity. I do all my tuning hot on the dyno. It's a bit of a luxury tuning a race car as things like cold start, warmup cycle and off idle tip in don't have to be spot on perfect. It's possible that if I watched target vs actual VVT when cold on some of the dialed race engines we have here, I'd see more error than I like.

18psi

iTrader: (76)

You almost read my mind: I was thinking exactly the same thing yesterday - "maybe this Rotella I'm using is too much for the stock oil valve to handle causing the oscillations"....After all its way more thin when cold.

I guess I will test this theory when the built engine gets dropped in soon and I have to break it in on regular 10w-30 dino oil

TheV's

iTrader: (2)

Was there ever an update to support this 36-1 setup on MS3X ? Just going down this path myself. Thanks very much !

Reverant

iTrader: (10)

No, it's not supported.

If you use it, you will be stuck with batch fuel, wasted spark and no VT.

TheV's

iTrader: (2)

OK, so what do I need to do to get VVT and sequential Spark ? Can I use the 4 tooth crank wheel with the VVT 2+1 Cam sensor ?

Reverant

iTrader: (10)

You can use the stock 4 tooth crank wheel with the stock 2+1 cam wheel, yes. Or you can use FM's 36-2 wheel.

TheV's

iTrader: (2)

OK - probably a Dumb Question - since I already have the 36-1 wheel - can I just grind a tooth off it and make it a 36-2 ?

Reverant

iTrader: (10)

If grinding a tooth off makes it look exactly like FM's 36-2, sure.

TheV's

iTrader: (2)

Looks like the angle before TDC might be off, but I assume I can adjust this via the trigger Angel/offset right ?...I assume I'm setting this up this as Spark mode "Miata 36-2"

Ben

iTrader: (33)

No -- this is hard coded in the Miata 36-2 wheel decoder.

dynodragon

iTrader: (1)

When my car was stock (a long time ago) I actually got a CEL from the VVT system after using a lower viscosity oil. Replaced the oil, cleared the CEL, Issue resolved.

NiklasFalk

iTrader: (1)

I have a noob issue with MS3 and VVT (Rev MS3-Basic with 270/319 as min/max in the base map).

How do I make it work for my range (-13 to 34 relative degrees).
With the VVtuner I advanced the intake wheel one tooth to make the "zero" +3 degrees, but I suspect the my cams would behave better with the intake at about -2 (103 instead of 105 as grinded).

If I set the target to 0 or -2 the VVT does nothing (0 DC which means the cam is at -13) and in the log it's reported as 270(?).
When I had the cam one tooth advanced the numbers in the log looked reasonable (2-3 as min when regulating to 2-5).

Am I assuming correctly that i should change the min/max from 270/319 to 257/306?
And can I reduce the max even more as a safeguard to valve-piston contact (which happens at 34 degrees relative).
But will zero relative still mean the same zero (relative to range or crank)?

To ask a really direct question, what happens if the cam falls outside of the min/max?
Does the DC just go to zero, which would mean that the min value should be lower than the minimum possible (or it will never regulate back)?

Sorry if this have been covered in length before in this thread, but I did not start a new one at least...

NiklasFalk

iTrader: (1)

Hmm, did my homework and tested the VVT sweep to 257-302.
I set these numbers as min and max, but that means that 0 relative advance is the 257 (It looked like that), the max retarded, which for me means 13 degrees retarded (measured and reported in VVTuner at zero duty).

So if I aim for 102 degrees with my 105 grinded cam, should I aim for 10 degrees advanced (-3 from the zero)?

Not the most logical setup for the numbers, but it makes it equally confusing for all setups I guess.

Ben

iTrader: (33)

Your question doesn't make sense. You can't run less advance than 0° relative. This is your min position. Properly set your min and max, and then the VVT table becomes "relative".

NiklasFalk

iTrader: (1)

I see that now (after reading the setup homework), as a noob I just assumed zero meant zero, not whatever angle is min.

But if I set min to 2 degrees higher I can set -1 as target and it will regulate towards it, so it seems like it could be done (with "some" changes).

In all it just means one more conversion when the tuner asks what angle the cam is at, 105 or 102.
Eh, it's 257+13-3 I guess, which is 10 in the table, i.e 102 of course...
Since every engine is unique it would not hurt to have a "zero" value in the settings and get the true cam phase in the target, logs and display. It would help comparison at least.

But VVT in MS3x is an old horse, this must have been beaten on before. Or am I the only one who have min so far away from zero (13 degrees, not just a few)?