3rdCarMX5

iTrader: (1)

So I have had this problem for a while now, it isn't a huge deal. A 1000rpm stable idle isn't a bad thing. I think I know what I need to do, but let's see what tuning through choice voting gets us.

When the idle valve is closed the rpm doesn't drop below 975-1000rpm. I have verified this with the idle valve test.

I can set my idle target to a value less than this but it doesn't reach it. I can set it to a value higher than 1000 and it will find it quickly.

The stock ecu will idle at the stock rpm. It is a little lumpy though.

Do I need to adjust the air bypass screw on the throttle body? Will this affect running the stock ecu?

I just can't imagine how the stock ecu can get down to around 750-800rpm and the MS3x can't.

Braineack

iTrader: (62)

what's the lowest min idle duty cycle? do you have a diode installed.

3rdCarMX5

iTrader: (1)

The min idle duty cycle is 18, max is up around 80 I think, and I have the diode installed.

Braineack

iTrader: (62)

at around 19% that valve shoudl be closed if you're using the expander circuit with flyback diode.

sixshooter

iTrader: (12)

Double check for a small vacuum leak somewhere.

hornetball

iTrader: (4)

Does the '99 TB have a bypass screw like the older TBs?

3rdCarMX5

iTrader: (1)

It shouldn't be a vacuum leak because it has always been there, even with the stock ecu, going back and forth.

The TB has a screw that looks like a bypass. I couldn't find much info, though.

hornetball

iTrader: (4)

In that case, the screw is probably out of adjustment. Turn CW to reduce RPM and CCW to increase RPM at a fixed idle duty cycle (i.e., warmup-only mode).

Braineack

iTrader: (62)

oh wait he's a 99, i think i read 91.

IIRC those have a range of like 4-30% or something silly. but you said you used idle test mode?

he doesn't need to touch the damn idle speed screw if his stock ecu can achieve lower numbers. he needs to tune correctly, either the min/max duty cycles, the valve freq. or something else.

3rdCarMX5

iTrader: (1)

I agree. 18% is the value where no change occurs as it goes lower.

How does frequency affect the idle speed?

I have tried all kinds of changes in my idle settings with no change.

hornetball

iTrader: (4)

Didn't he just say "It shouldn't be a vacuum leak because it has always been there, even with the stock ecu"?

My interpretation is that he had high idle with the stock ECU, which means his idle screw might be out of whack or he might have the aforementioned vacuum leak.

But . . .

3rdCarMX5

iTrader: (1)

Sorry, idle is stock rpm with stock ecu- which is why I didn't want to touch the throttle.

hornetball

iTrader: (4)

Roger that. I wouldn't touch it either then.

You might check your timing in the idle region. Idle RPM will follow ignition advance (more advance = more RPM). Retarding the timing should bring it right down.

hornetball

iTrader: (4)

For reference, I run 12° in the idle region. I bump that up to 20° when the AC is engaged.

bmxfuel007

iTrader: (21)

what does your closed loop idle initial values (duty) table look like? been messing with mine lately as I've had my idle hang up high randomly until I adjusted some of that table. I'm not sure if this would pertain to your scenario though

maybe post a screen shot of your closed loop idle settings, table, etc. ?

Braineack

iTrader: (62)

there;s no reason to run that much timing with the a/c running.

i run 10° at 750RPM with the a/c engaged.





please dont post datalogs or even your msq, those will not be helpful.

3rdCarMX5

iTrader: (1)

My idle timing is 15*. I have tried 10* when I was chasing other problems, but I will see what difference it makes and do it again.

My tuning laptop doesn't have an internet connection, and I have been busy with work.

3rdCarMX5

iTrader: (1)

Thread resurrection!

So I am ready to troubleshoot this again. I now have the tools and more knowledge to track it down.

Summary:
I can idle at ~1200rpm. It won't go lower when duty cycle is changed in TS. It will go higher.
When I unplug the idle valve the car dies. IACV is working, and with the stock ECU it operates as expected.

I think it must be something wrong with my MS3 wiring (I used 99 ? 05 MS3 | Frank's Westfield MX5 99 ? 05 MS3 | and made a thread on here about building it, It works other than this idle issue and the knock sensor isn't calibrated properly). I have an oscilloscope now, does anyone have an example of what the signal should look like before I go and hook it up? When I made this thing it was pretty new to have an MS3x and now a lot of people do, so hopefully there is more information now.

I will try to get my MSQ posted soon.

Braineack

iTrader: (62)

Did you do the flyback mod? Unplugging idle vavle should increase idle

3rdCarMX5

iTrader: (1)

Yeah, flyback diode is in place.

https://www.miataturbo.net/megasquir...e2/#post878449

Here is the post with the detailed info, from when I built it. I may have made a mistake that will be self-evident.