Gunson testing on 83 SC

[ThomO]

I bought a Gunson gas tester and wanted to set my CO level. I had some vacuum leak issues which I believe I have resolved. The car runs good, start good cold or hot. The only issue is driving at highway speeds at a constant throttle I get a hesitation or cut out type of feeling.

I am thinking with everything else that was jacked with on my car the mixture has been played with to compensate for vacuum leaks.

I did a bunch of searching and reading (forum, bentley, hayes, gunson manual) and I think I have the right process.

My car is an 83 SC US model with cat and O2.


I am going to

physically remove the O2 from the cat
warm up the gunson on a spare battery
drive the car till it is warm
insert the probe in the O2 bung
measure and adjust as required


Does that sound correct

I do no have a dwell meter so not going to do that, and I am not going to buy a wide band.


Please help with my sanity check.

Thanks

Thom O

[tirwin]

Thom,

I have never used the Gunson but some of the exhaust analyzers I have seen have the probe on a spring thing that centers the probe in the middle of the exhaust pipe so no need to remove the O2 sensor.

I'm just curious what your voltage reading is on the O2 when your AFR and CO is set properly. I just tested mine last night and it was reading .86V at idle right after a 30 min drive. I need to get a proper dwell meter to check for sure but I think I'm still running rich after Tony helped me with the cold control pressure on my WUR.

Since you're doing this, would you mind doing a little test? When you do your exhaust gas measurement, disconnect the O2 sensor and plug in the + of a voltmeter to the female end of the O2 plug and the - to ground.

[Peter Zimmermann]

Thom; you have the right idea. But, you should have an exhaust sample port that is on the exhaust just forward of the cat body. It points toward the inner sidewall of the LR tire, and has either a cap nut or a small bolt that plugs it's end. Open the port (no need to remove the sensor), find a hose that fits snugly onto it, and over the CO machine's "sniffer," then take a sample.

1. You must disconnect the oxygen sensor's wire in the engine compartment
2. Get the engine just hot enough to open the front cooler (hot to touch)
3. Make adjustments as quickly as possible

BTW; It's possible that the hesitation that you feel is your O2 sensor, just disconnect its wire and drive the car; I'll bet it's smoother.

[ThomO]

I will take the measurement tirwin.

Pete I am currently running with the O2 sensor disconnected. It does not seem to matter if it is plugged or unplugged.

The hesitation is pretty substantial it can move me a little (ex nose tackle). It is really taking away from the joy of driving the car.

I have an aftermarket eastern cat not the original factory cat so no test port.

The car was a beater that has been a messed over pound puppy (just the way I like em)

I have learned a great deal on cis my previous experience is L-jetronic on 928's and carbs on Detroit steel.

I have gone thru everything else control pressures, vacuum leaks, yada yada yada.

I know that the CO value is off since I have played with it to get it to run. Now it is time to set it correctly hence the Gunson.

advice always excepted

thanks for the responses

[NeedSpace]

Thom any luck with this, I am thinking of buying a Gunson myself and am curious how things go. If you could take pics or video that would be great.

[ThomO]

So here is an update after the weekend.

I went t a muffler shop and had them put a second O2 bung before my cat to use as a test port.

I hooked up the tester to a spare battery and went for a warm up cruise.

came back dialed in the tester

inserted the probe

and started to adjudt the CIS which changed the idle, adjust the idle which changed the CO.

It is a back and forth process untill you get it dialed with the correct idle and CO reading.

The car runs great feels like it has more power and the surging is pretty much gone

overall very happy

[tirwin]

Thom,

Were you able to check the voltage on the O2 sensor? Just curious. There is a Porsche tune-up DVD that claims you can use the O2 sensor voltage as a proxy for AFR with "proper" being @ .7V. That is of course assuming a good O2 sensor and no other issues.

[Bob Kontak]

Originally Posted by tirwin

Were you able to check the voltage on the O2 sensor? Just curious. There is a Porsche tune-up DVD that claims you can use the O2 sensor voltage as a proxy for AFR with "proper" being @ .7V. That is of course assuming a good O2 sensor and no other issues.

I have done this - Interesting on the 0.7V as normal. I just "figured" half way would be the sweet spot. Around 0.450 volts.

[ThomO]

No I forgot to check it

I will check it and get back to you

[tirwin]

Bob,

This forum has taught me to become suspicious of Porsche folklore, so I thought it would be an interesting experiment. My kids enjoy playing a game called "Fact or Crap". Maybe we should start a "Porsche Mythbusters" where we verify or debunk these things.

A Google search produces a few images for narrowband O2 sensor graphs. The graphs seem to indicate that the curve on a narrowband is non-linear and there is quite a large voltage range around stoichiometric (.2-.8V). If this is universally true for all narrowband O2 sensors, then I think this idea should go in the "crap" category unless other data can be found that corroborates it. Now maybe if someone actually took a sample of enough cars that were set up properly and tested the O2 voltage on all them, you would find the vast majority of them showed .7V. And that's why I was asking Thom for his help.

I guess my point is that if it could be proven to be true that .7V is the "right" value on a properly set up car, it could be useful as a quick and easy test for diagnosing that the AFR is off on CIS K-lambda cars ('81-'83) -- something to go in the CIS for Dummies thread as a check.

Since we have seen a lot of people erroneously start messing with their AFR to fix CIS issues, could this also be an easy way to get them back to a setting in the right ballpark if they don't have access to a gas analyzer or a dwell meter?

A very high or low O2 sensor reading would seem to indicate an overly rich or lean condition, but I don't think you could infer the AFR is set properly. The other problem here is that perfect stoichiometric ratio is outside the best power band, so it very much depends on how your car is tuned.

I'm about to replace my O2 sensor with a wideband that has a narrowband output so I'm collecting some data so I can do a before and after comparison. Hoping to solicit input from others to do this check so we can officially corroborate or bust the myth.

[ClickClickBoom]

Thom,
Glad the Gunsons worked for you! Now that you are in the ballpark, try adjusting the mixture seat of the pants, little richer, little leaner find out where your car is happiest. I tuned my CIS with the Gunsons and then via the seat of the pants method. Too lean and the teltale pinging will surface, too rich and the mosquitos willl leave town. i adjusted my mixture in the early winter and early summer! The 3mm mixture screw was moved only tiny amounts to find the sweet spot, lots of drive, twist and drive. The mixture screw moved only by tiny pressure amounts. Just remember to shut off the car between adjustments. Sweet thing is if you get too far off, the Gunsons will get you back to home point.
Enjoy!
Eric

[Joe Bob]

Have you put a gauge on the fuel pressure?

[dicklague]

Originally Posted by tirwin

Bob,

This forum has taught me to become suspicious of Porsche folklore, so I thought it would be an interesting experiment. My kids enjoy playing a game called "Fact or Crap". Maybe we should start a "Porsche Mythbusters" where we verify or debunk these things.

A Google search produces a few images for narrowband O2 sensor graphs. The graphs seem to indicate that the curve on a narrowband is non-linear and there is quite a large voltage range around stoichiometric (.2-.8V). If this is universally true for all narrowband O2 sensors, then I think this idea should go in the "crap" category unless other data can be found that corroborates it. Now maybe if someone actually took a sample of enough cars that were set up properly and tested the O2 voltage on all them, you would find the vast majority of them showed .7V. And that's why I was asking Thom for his help.

I guess my point is that if it could be proven to be true that .7V is the "right" value on a properly set up car, it could be useful as a quick and easy test for diagnosing that the AFR is off on CIS K-lambda cars ('81-'83) -- something to go in the CIS for Dummies thread as a check.

Since we have seen a lot of people erroneously start messing with their AFR to fix CIS issues, could this also be an easy way to get them back to a setting in the right ballpark if they don't have access to a gas analyzer or a dwell meter?

A very high or low O2 sensor reading would seem to indicate an overly rich or lean condition, but I don't think you could infer the AFR is set properly. The other problem here is that perfect stoichiometric ratio is outside the best power band, so it very much depends on how your car is tuned.

I'm about to replace my O2 sensor with a wideband that has a narrowband output so I'm collecting some data so I can do a before and after comparison. Hoping to solicit input from others to do this check so we can officially corroborate or bust the myth.

I would not use a narrow band to measure anything in a Porsche.
Narrow band sensors are like switches. TOO LEAN ------ TOO RICH. I would not trust them to set my car.

All this talk of CO and Gunson's has me confused. Why not use an AFR instrument that uses a modern WIDE band sensor and really know what the AFR is before the cat? AFR and CO are direct conversions and you can find charts by Googling.

I have to admit I am smitten by the Daytona-Sensors Wego IV instrument. It gathers RPM, AFR, MAP and can save 2 hours of data that you can easily download with a USB cable.

I have one permanently installed in my 1973 MFI 911 and it has been a miracle in getting the car to run right. You can review 2 hours of running and see what your AFR is throughout the range and not just at idle as in the Gunson.

In summary, why use ancient Gunson CO for adjustment when you can use much newer tech and faster and more accurate AFR wideband instruments?

The Wego IV is maybe $75 more money than importing a Gunson and the difference in speed and accuracy and being able to collect the data is not even on the same planet.

By the way our host Pelican sells the Wego IV.