CIS duty cycle too low

[blpetry]

Hello all,

Working on a 1982 US SC.

From a cold start with O2 sensor disconnected, duty cycle is showing 18% (9 deg X 2 on V8 scale). From cold, when hot, always. Right around 18%, with O2 disconnected (or connected for that matter).

Grounding the O2 lead causes the duty cycle to rise ever so slightly, but definitely moves - maybe 2%.

Applying 1.5v to the O2 lead causes the duty cycle to go down ever so slightly, again, definitely moves - maybe 2%.

Car runs but the mixture setting was way too lean with that low of a dwell on the FV. Had to push the plate up to get enough gas to start. Warmed up and raised the mixture until I got .5v from the O2 sensor and a good idle (with reference to previous setting so I can return when this is fixed), and car idles, runs and drives pretty darn good, but I suspect it is not getting the benefit of full throttle enrichment either and is probably still pretty lean in mid range

I checked the lambda relay and it is "working" (otherwise there would be no dwell at all).

I'm kind of at a loss as to where to go from here. Replace the relay? Try another relay? Bad brain box? Bad FV? Bad connection? Bad ground?

This isn't my first rodeo with CIS as I've owned an '84 VW GTI for years and have used the exact same procedures quite a few times before, and am completely familiar with the "dance" that the duty cycle goes through on warm up through to closed loop operation.

Any suggestions would be super-appreciated!

Regards,
Brian Petry
Springboro, OH

[matt demaria]

I would use a better instrument for measuring this. I know that the whole length of the pulse to the frequency valve is about 14ms long. Fully warm, o2 disconnected, the duty cycle is at 50%. This pattern is a "peak & hold" pattern. Your tool might not be picking this up. A good scope will display this.
Open loop voltage at the o2 wire to the computer (with o2 disconnected) is about .5 volts. If you measured .5v at the o2 (disconn) when you set the co screw, and reconnected the o2, the voltage should have been swinging back and forth (1 hz).

[blpetry]

Cool thanks for the reply Matt. When you say better instrument do you mean another type of dwell meter or an oscilloscope or something? I've got one of those too, but haven't ever used it for anything automotive. If you have a model of dwell meter to recommend I'd be interested, I've tried my sunpro unit from the flaps and an ancient battery powered unit from sears with the same results. Hadn't thought to measure the o2 voltage when hooked up but will try. Thanks!

[matt demaria]

Use a scope. (I use a vantage pro for this). Adjust time base. Adjust y axis to 50 volt scale. Synch off of the large turnoff spike.
I use a breakout box for this, which I assume you do not have. If you tap into the frequency valve at the engine side be sure to use shielded leads.

[blpetry]

I haven't hooked the car up to the scope yet, but you are definitely correct about normal dwell meters not working on these systems. The issue with my car was that the frequency valve was not operating at all. Prior to connecting the car up to the scope, I went through and cleaned the grounds to the computer and eliminated the relay from the equation with a jumper from pin 30 to pins 87a and 87. The frequency valve did not pulse. Turned out the connection at the front of the motor by the LH shock tower was knackered. A little contact cleaner and the FV was pulsing. Returning the mixture screw to where it was before this ordeal put the car in normal running order.

I do, however, want to confirm the mixture using the scope, just because. Can I just hook the leads up to test port and ground, set up the scope like you described, and adjust the mixture screw to where the square wave is 50% up and 50% down (with the O2 connected and warm)? I'm not an EE so sorry if that is a stupid question... I do play one in my garage though. Or should I just be happy my car is running good and go for a drive?

Thanks again, this forum is awesome!