Miata Turbo Forum - Boost cars, acquire cats.

Miata Turbo Forum - Boost cars, acquire cats. (https://www.miataturbo.net/)
-   AEM (https://www.miataturbo.net/aem-59/)
-   -   the thread where Savington asks a lot of really, really dumb questions. (https://www.miataturbo.net/aem-59/thread-where-savington-asks-lot-really-really-dumb-questions-42173/)

Savington 03-12-2010 06:53 PM

1 Attachment(s)

Originally Posted by JasonC SBB (Post 537231)
Post a screenshot of the cam #1 template.

Did you look at the parameter screen to see what duty cycle it's actually outputting?

And, if your "cam start" number is correct, it should idle at ~23* cam degrees if you're outputting zero duty cycle.

Attachment 199480

I can snap one with the car running if you want it. I did check the actual duty cycle on the parameter screen, it was changing as I adjusted the duty vs. target table. I fiddled a bit with the target table to see if that had any effect, didn't seem to.

JasonC SBB 03-12-2010 07:28 PM

Re-read my posts 86 and 87.
Your numbers will be RETARD not advance.

0 signifies max advance.

Savington 03-12-2010 08:29 PM


Originally Posted by JasonC SBB (Post 537248)
Re-read my posts 86 and 87.
Your numbers will be RETARD not advance.

0 signifies max advance.

I know. I was trying to advance it at idle. Still was unable to do anything with it.

JasonC SBB 03-12-2010 08:39 PM

OK then check that it's reading correctly first.
At some higher RPM ...
Does it read 23* retard with 0 duty cycle?
With 100% duty cycle, does it then go to 0-1* retard?

The easy way to directly output a given duty cycle is with the VVC Out max and min options. e.g. set min to 99 or 100% if you want max duty, and set max to 0 or 1 if you want 0. You may have to change both e.g. set min to 98 and max to 99, then set min to 0 and max to 1.

If it's not reading correctly, adjust "cam start".

EDIT:

Goddamit I reviewed my post #86 again.

Manual says
Cam Range = (360 Degrees/# of Fuel Teeth per rev)/0.8

That's PER REV (2 teeth), not per cycle (4 teeth).
That would give 225* for "cam range" (not 112.5) if you wanted to read crank degrees.
Because the fucking software won't accept >100*, I suggested using 56.25 and think in terms of cam degrees. Well that won't work either because it would need 112.5 to do that which is >100.

Here's an idea. Instead of reading off cam advance in cam degrees, how about 0 to 10? (yes a bit lame). Think of it as 10ths of max retard/advance. To do that the "cam range" multiplier needs to be 47.87.

You then need to adjust "cam start" so that max retard reads out close to 10, or max advance reads close to 0.

Savington 03-14-2010 12:28 AM

The best things are discovered by accident. All the crap I did with the settings at idle did nothing, but as soon as I backed it out of the garage tonight to go autocrossing tomorrow, it fell flat on its face above 2500rpm. Sure enough, the cam angle is changing by 12+ degrees. I turned it off since it obviously needs to be tuned, but we're there.

Savington 03-18-2010 10:38 PM

Had a chance to play with this a bit tonight.

First thing: Higher number in the target table is more advance. I can hold the car at 2500rpm, and if I increase the target in that cell the car will eventually die (too much advance). Same with duty cycle, 100%DC is no advance. If I set it at 60, the car idles fine. Set it at 40 and the car will immediately die, just like it did when I put +12v on the solenoid at the beginning.

With the error table set up as Jason had me set it up (positive slope), the car obviously dies as soon as there is even the slightest bit of transfer in the spool valve, since the assumption was that more duty = more advance. I need to poke through the manual again, but we may be able to flip all of that around by just inverting the signal.

Numbers aren't quite important yet, but if possible I want to use a multiple of the actual advance being applied, vs a 0-10 scale, just so it's easy to discuss settings and such with Hydra guys.

y8s 03-18-2010 10:48 PM

and the adaptronic guys. not that I've implemented it there yet. but it speaks in degrees. it's just a matter of getting the offset value correct.

JasonC SBB 03-18-2010 11:28 PM

Take a step back.

What are the readouts of "Cam #1 ADV" and "Cam #1 Teeth" if you disconnect the solenoid?
If you apply 12V?

If "VVC#1 invert" box is unchecked, 0 duty means 0V and 100% duty means 12V goes to solenoid.

Savington 03-19-2010 01:13 AM


Originally Posted by JasonC SBB (Post 540772)
Take a step back.

What are the readouts of "Cam #1 ADV" and "Cam #1 Teeth" if you disconnect the solenoid?
If you apply 12V?

If "VVC#1 invert" box is unchecked, 0 duty means 0V and 100% duty means 12V goes to solenoid.

I just checked it. I set the cam range to 47.27 as you said to get a 0-10 scale, with 10 being full retard and 0 being full advance. I then set the table to ~7 to try to get a small amount of advance out of it.

Bringing the duty cycle down to 10 did nothing, but then at about 6% it would shut the car off from an idle. I revved it to 3000rpm and it ran rough, and then I quickly inverted the error table and it stabilized at ~7 "degrees". Cam #1 angle is reading approximately 1.81, full retard is 1.92.

JasonC SBB 03-19-2010 01:19 AM

So what you need to do is invert the error table's slope, and set the integral gain to -1.

I don't understand why 10% duty cycle did nothing then 6% shut the car down.

Savington 03-19-2010 03:05 AM


Originally Posted by JasonC SBB (Post 540823)
So what you need to do is invert the error table's slope, and set the integral gain to -1.

I don't understand why 10% duty cycle did nothing then 6% shut the car down.

I forgot about integral gain, I will flip that tomorrow and report back on the results. I will also look at EXACTLY what duty cycle engages the solenoid.

We may want to expand on the range, larger than just 0-10 - it reduces resolution in the error table by quite a bit.

Savington 03-19-2010 11:18 PM


Originally Posted by Savington (Post 540839)
I will also look at EXACTLY what duty cycle engages the solenoid.

30%. 30.08%, to be exact. With the error table set at 0%, 30.4% will hold full retard, 30.08% holds full retard, and as soon as I go down to 29.6% it begins to slowly advance the cam. I can then go back to 30.08 and hold whatever advance I want (within a degree, about as accurate as the Cam#1ADV ever gets).

Savington 03-19-2010 11:33 PM

2 Attachment(s)
Attachment 199226



Using these numbers:

Attachment 199227

Savington 03-20-2010 03:27 PM


Originally Posted by Savington (Post 540741)
Had a chance to play with this a bit tonight.

First thing: Higher number in the target table is more advance. I can hold the car at 2500rpm, and if I increase the target in that cell the car will eventually die (too much advance). Same with duty cycle, 100%DC is no advance. If I set it at 60, the car idles fine. Set it at 40 and the car will immediately die, just like it did when I put +12v on the solenoid at the beginning.

This is half wrong. Higher numbers are more retard, but so is higher duty cycle. Not sure why the car was dying with higher targets, but lower target numbers are definitely more advance.

JasonC SBB 03-22-2010 10:39 PM

The largest number you can put into cam range is 100. If you put in 100 the cam readout will go from 0 to 20.9. If you put in 95.74 it will read from 0 to 20. The resolution of the target map will be about 2.2* (crank degrees) which is OK.

If you install a crank trigger wheel with at least 6-1 teeth then you can make it readout 0-47*.

It looks like the output of inj#9 is inverted - check the box "vvc#1 invert" so that 0 duty displayed is actually 0 output.


All times are GMT -4. The time now is 10:41 PM.


© 2024 MH Sub I, LLC dba Internet Brands