News:

New Users: Send [email protected] an email with your account name. I've turned off auto-approve to reduce the 100 new spam accounts a day. Thanks, -Jason

Main Menu

Idle PWM - rattling sound?

Started by GintsK, October 24, 2007, 11:24:27 PM

Previous topic - Next topic

GintsK

Hello.

Can anybody give some idea about idle valve PWMing: generally it works, but IAC give out rattling sound. It is irritating at engine off/ignition on. With scope it looks like PWM is very irregular but give average duty cycle as desired. Changing of frequency does not help much. Only change rattling a little. What here can help? Big capacitor or something? I have this problem on two cars with simple 2-pin IACs of different constructions: Audi 7A engine and Mazda 323F '89 with 1.5L engine (turbo now ;))

Gints

[email protected]

Have you tried an external flyback diode?

GintsK

External? I drive it from INJFET and my flyback 30V diode is inside VEMS - it's for Mazda. In Audi case I use Power Flyback board - so relative low voltage flyback.  May be I try to high frequencies: put 1...10 in Megatune? And separate firmware branches is used too: for Audi 1.1.15 or 18, for mazda 1.0.73. Both have rattling.

Does someone suffer from this or this is problem for me only?

Gints

[email protected]

I've found that the IACV clatters on a few of the installs that I've done, its one of those 'features' that goes away if you turn the ignition off.  One of the spare 1N4004 diodes would be worth trying.

hilly

What firmware are you using ?

On v1.0.53 for sure most of the PWM frequency options in MegaTune run the stepper at a low frequency (10hz ish). The duty cycle control isn't very constant and this causes the stepper to jump about. However if you pick a PWM value which is a multiple of 64 (0, 64, 128, 192) then the output runs at 1.2KHz and the stepper movement is smooth. Give it a try.

Hilly

GintsK


hilly

Gints

I got caught out by the idle PWM as well, my Bosch valve was rattling about like a mad thing.
I expected the iac speed to look something like the chart you posted, but I measured it with an oscilloscope and it was all over the place.

This the the value in MegaTune


I started at 1 and kept incrementing the value whilst measuring the frequency and found no patteren until I hit 64 and it then jumped up to 1.2KHz and the valve worked like a charm.

Hilly

[email protected]

If you started at zero you'd find that the result would have been the same - I've showed the development team these findings and was shown:  IACFREQ_PWM ((4 + (config.iac_speed & 0x3F)) << 7) so 64 is same as 0

GintsK

Yes. I just try it with scope on stimulator. Looks like ~680Hz (?). And yes much more stable like any other setting. Still fluctuate, but discountinuties is much shorter - ~1ms. with others 5ms and longer pauses is a norm. It causes valve clicking.

Does so high frequency can't give some heat problems?

So next week will try 0 on Mazda. 

hilly

There should be no increased heat problems as the power to the stepper is still the same, PWM remember.
The flyback diode will have to work harder due to the increased number of edges it is suppressing, but no worries there either.

Hilly

[email protected]

I've just been given the following instructions: To stop the IACV clicking:
Set iac_cold_start_pos & iac_warm_start_pos to 0.

Tcal

#11
That does stop it too, but iÃ,´m using speed 0. IAC valve works much smoother with that.
Is there a reason to why not use it in 0 speed??

[email protected]

In some cases 0 speed will be too fast for certain solenoids to react, if you're getting good results then stick with 0.

z0tya

We have a same rattling problem, but with BY399 3A fast diode from shop.
Is it good for this? Or need change to 1N4004?
Anyway the iac actuator speed not mean frequency in Hz but the table values above. So maybe 1-15 valid?
Can I try the 64 value?

Quote from: [email protected] on October 25, 2007, 09:49:11 PM
I've found that the IACV clatters on a few of the installs that I've done, its one of those 'features' that goes away if you turn the ignition off.  One of the spare 1N4004 diodes would be worth trying.

GintsK

Usualy only acceptable value is 0. signal isn't true pwming, but some kind of pseudo pwm - irregular and with interrupts.

Any diode should do the job.

May be some capacitor can help to mask rattling?

Gints