MicroSquirt.com read only forum (Jan.15,2018)

Hey,

I'm finally at breaking point on the Ford 2-wire PWM idle valve. I've played, tweaked, tuned, searched all over the forum and Google, but nowhere and notime have I figured or found a way to tune the thing.
So perhaps anyone who's experience - or even the programmers who've wrote the code to control it can put some info down in a collective place. Perhaps we can get this pinned if successful or edited and entered into the manual?

The idle valve I'm using is from a UK spec Ford Escort Zetec. Most Fords using the EEC-IV here in the UK have the same idle valve, though they are often re-cased to suit the particular car.

My problem:
Using the stock Ford ECU (EEC-IV), the engine will settle on a rock solid idle of about 950-1000rpm.

When using the MS (MS1, latest MSnS-Extra code, modified with TIP120, all configured for PWM idle), no matter what numbers are punched in, it ALWAYS occillates. (Occillation defined as revs increase beyond Fast idle, then below Slow idle). You can only stop occillation by applying throttle above TPS threshold - by foot I can get it to sit at 1000rpm, very slowly bringing my foot off can get it to idle at 700-800rpm. The slightest hiccup by the engine causes occillation to return!

On both ECU's, unplugging the idle valve, the car will idle at about 600rpm(!)

Judging by the engine running without it plugged in (electricly), in theory, the valve must be closed when unpowered.

I have a dead valve in front of me showing me different! As-it-stands, the valve is about 20-33% open. Of the 2 holes on the valve, I cant say which is inlet or outlet - so I can't tell if the valve is held closed by vacuum or naturally allows a small amount of air to pass.


Another post (http://www.msefi.com/viewtopic.php?p=11 ... dle#116248), a member, Keithg, suggests:
[quote=Keithg]Change deadband range to 40 from 200.

Change idle valve freq to 100 from 45. Crank DC to 50 and min freq to 10 and closed freq to 0. Try idle activation at 300 instead of 600. [/quote]

But this made little to no difference for mine (US ones may be different to UK ones?).

So, any ideas? Any UK people with successful closed-loop settings you wouldn't mind sharing?

Ta much!

See this thread.
http://www.msefi.com/viewtopic.php?t=17238&highlight=
I'm using a valve off a 98 4.6L Mustang. It is normally partially open until vacuum is applied, then it closes.
I did not spend much time trying to run closed loop. I'm running a 351W and varying loads (A/C or Lights) don't have much affect on my rpm's. I found that adjusting my throttle stop up to give me 650 rpm idle and running warm-up only works best for me.

Thanks RNash.

Looking at that thread, I see only warmup is successful.

Problem with mine is that I dont have much scope to adjust throttle.

The throttle body is non-adjustable (though I may drill & tap an adjuster screw in for now).

I was ideally looking for Closed-loop control, as it means I don't have to fiddle and tweak for when high current devices kick in or out.

Actually, I have been using closed loop for quite a while now. I wrote part of the code for MSnS-E. It is not perfect, but it can work. It can be a bear to tune. I can help, but you will need to post an MSQ and possibly a datalog along with any description of what it is doing. Pogoingcan be caused by 2 things, mainly, 1) too coarse of adjustment (increase the 'freq' number to get finer adjustment) 2) too aggressive of adjustment curve (page 2 settings)
.

KeithG

Posted by email.

Hey Keith.

I'll try to get something tonight or Friday.

I've bought an o-scope, so I'll also scope the valve when controlled by the stock ecu (then try understand what it means ).
So if were still at a loss, that might prove useful?

Great!

I thought of one more reason it could poo. If the settling time is not set long enough, it will not reach a steady idle and it will just dashpot continually.

KeithG

Hey Keith, I've attached the readings from the scope if they make sense.

The frequency appears constant at all times, so I guess the amplitude changes to adjust the valve opening.
Is that how the MS code works?

The Megasquirt should be adjusting the Pulsewidth to vary the idle. narrow pulsewidths are lower average DC to the valve, and the valve should be more closed, wide pulsewidths should be higher Average DC to the valve to open it more. Hence the reason they say Pulsewidth Modulation (PWM).

When I scope out my valve I get a nice squarewave at 50% duty cycle from ground to 12V. Your scope pics show a waveform riding a DC level.

Are you using one of the EEC-IV adaptor boards for this with the TIP120 on the board, or have you the Megasquirt mounted TIP120 and jumpered out the PWMIN to PWMOUT on the adaptor board?

As Scott says, the ECU (MS or EEC) should be varying the pulsewidth not the amplitude. Something looks fishy, As he said, it should be a nice square wave with either ECU.

KeithG

The waveform above is true EEC-IV output. The picture is from a diagnostic manual, but when I scoped the EEC-IV, the waveform was identicle to that image.
I watched the 'scope at idle (engine warmed up to running temp), when the fan kicked in, the idle settled, but the frequency stayed the same, hence why I said the EEC-IV must vary amplitude - effectively making the valve 'stronger' against the pressure against it as opposed to the valve position. Though, I can't say I saw evidence of the amplitude theory.

I scoped in DC and AC, and the waveform was the same.
I havent looked at the EEC-IV circuit, so I don't know if theres a capacitor in there to cause the rounding of the waveform?

Do your EEC-IV's output a square wave? Because, as you can see - these put out a saw wave.
Playing with a generic sqare wave, a capacitor to ground of 0.002nF can make the waveform I get

I'll stick the scope on the MS later, but I'll say it'll give a square wave that your expecting.

As for the Adapter boards, the MS that I tested the other week was jumpered internally and had the TIP120 on the adapter.
My own MS when I last used that used the TIP120 internally.
Both had the same 'occillation' problem with the revs.