Re : '89 200TQ Decceleration cut-off valve, OXS sensor

[e6941tb@ix.netcom.com]

I would like to thank Dave Lawson and Scott Mockry for quick and
detailed replies to my questions.
I have not started implementing the advice yet. Weather permitting, I'll
address the issue this coming weekend.


> Re: idle switch ok?: On my previous 5000T and on my 89 200TQ, I   
> found that the idle switch had some cracked internal solder joints which made the 
> switch work ok one minute and then go open circuit or have high resistance 
> the next minute.  I found this to be the case  on another list members 87 5000TQ 
> as well.  If you have never replaced this switch or have never taken it apart then it
> very well could be the source of the problem. Taking the switch apart is a bit of a 
> pain as the plastic cover is solvent welded on.

Yeah, I've done it a year ago and the procedure was a subject of one of
my first posts with this news group.

> The schematic also shows a coolant temp sensor that connects to pin "T" on the idle/decel
> control unit and that it should have a resistance of aprox 130 ohms at 176 deg F. You
> may want to verify this temp sensor is working ok.

I will do that.

> The signal driving the  frequency valve (OXS test connector) 
> that controls the mixture is around  70 Hz and is a 
> Pulse Width Modulated (PWM) signal (resembles a square wave) It
> is the Negative Duty Cycle that you need to be concerned with. In other
> words it is the time when the signal is low (pulled down to ground) when
> the Frequency valve is turned on. I am not that familar with the WaveTek
> DMM's but in some cases you can select whether you  are measuring the 
> negative duty cycle or the positive duty cycle. Sometimes those obscure 
> details are buried in the owners manual.

Aha! That explains it! The Wavetek measures Duty C-le as a ratio of a
"hill"/period. If it's a negative one, then it's a "valley"/period. I
get 60%, Bently sez 40%, so we're in the same boat here. That's a
relief.
 
> When you were viewing this frequency valve waveform using the oscilloscope, 
> the sharp spikes you were seeing on the square wave were caused by 
> the inductive voltage "kick" that occurs when the solenoid is turned
> off and is considered normal.

Audi could've invested $0.20 in a reversed biased 1N4002 across the
solenoid coil on this $40k+ car. Slubs.

> With the O2 sensor connected, 
> engine and sensor warmed up etc. you should be able to view the
> change in the Pulse width as the ECU responds to the voltage swing
> from the O2 sensor. O2 sensor voltage goes high (~0.85 V, rich) and then
> the OXS frequency valve signal pulse width should be reduced, i.e. Duty cycle is
> reduced, then O2 sensor voltage swings low (~0.15V, lean) and then
> the duty cycle is increased slightly and so it goes cycling back and forth to
> tweak the mixture slightly.

> On my cyberdyne a/f meter, when rpms are above 1200 and throttle
> is closed, the o2 sensor reports no fuel, ie no bars are lit on the gauge.
> 

I built my own version, that also listens to the OXS sensor and shows
the results on a colour LED bargraph (I did not like the Cyberdine's
enclosure from Radio Shlock and did not like the Hal Meter's zig-zaging
light travel). The light bar swings exactly like you are describing.

Incidently, before the Deccel Valve had stopped working, I did indeed
see the light disappear completely, when cruising downhill in a low gear
with my both feet clear of the pedals. No fuel was being delievered to
the motor.

Right now the light bar happily swings left to right even on the
downhill. I am essentially feeding fuel to the the engine while using it
to slow down the car. Rather dumb. Need to address it immediately. Right
now the trip computer shows 12mi/gal. Ouch! And it ain't no Watkins Glen
yet.

Stay tuned. 

Igor Kessel
'89 200TQ