The Gunson exhaust CO analyser - any experiences or thoughts?

Started by raueda1, 15 July 2018, 02:15 PM

raueda1

With the money I saved by not needed a fuel distributor rebuild I'm thinking about one of these exhaust CO analysers:
https://www.ebay.com/itm/GUNSON-DIGITAL-GASTESTER-VEHICLE-EXHAUST-GAS-ANALYSER/222258556547?epid=1922975513&hash=item33bfa46e83%3Ag%3AM-QAAOxyuCJRfDhD%3Asc%3AUSPSPriority%2184124%21US%21-1&LH_PrefLoc=3&_sacat=0&_nkw=gunson+gastester&_from=R40&rt=nc

The purpose, obviously, is to get mixture set to spec.  Reviews and comments seem to say the unit is more-or-less OK for that purpose, particularly on no-cat, pre-emission controlled cars (mine is early, uncontrolled emissions, euro 6.9).  Anybody have experiences or heard rumors about it? 

As an aside, my car now seems to run a bit hotter after installing new WUR.  I interpret this to mean that it's now running leaner, hence hotter.  It would be nice to dial in the mixture properly and not just by guesswork.

And as a further aside, maybe somebody can confirm or correct my understanding of the mixture setting:  The screw that sets the mixture (as reflected by CO level) does so by changing the geometry between the air sensor plate and the control barrel in the fuel distributor.   That mixture, right or wrong, could be considered the "base mixture" and is modified by vacuum enrichment, all the cold start hardware, geometry of the intake plenum, etc etc. Therefore, for example, if you're running a bit lean at idle you'll be running a bit lean the rest of the time too.   Or, better put, if it isn't right at idle it can't be right anyplace else either.  This is inherent to the whole design of the K-jet CIS system.   Regardless, it seems like getting the mixture and idle speed just right is a circular process cause mixture effects idle speed.  If the system is whacked out I'd expect to set mixture, reset idle, reset mixture etc.  Just trying to understand what I'm doing. Thanks and cheers,
-Dave
Now:  1976 6.9 Euro, 2015 GL550
Before that:  1966 230S, 1964 220SE coupe, 1977 Carrera 3.0

Randys01

Yes, I've been admiring this gas analyser myself for some time. I don't see how you could go wrong at that price.

Your synopsis of the K Jet is pretty right but, I am not a huge subscriber to the popular theory that  within the allowable variation in the idle value CO, it has a measurable influence on the whole of range mixture ratio. Yes theoretically the relationship between the air flap plate and the fuel plunger in the FD is a constant and therefore value X should prevail across the entire spectrum. However the so called perfect stoichometric ratio of 14.7:1 is rarely achieved in the rev range of any engine whether it be injected or carbie. Arguably modern EFI can get closest but back to the 70's and K jetronic.

Most engines tend to be jigged to run rich at power [turbo charging is a classic ] and leanish where any pollution out put is measured! If you look into the history of engine pollution management, over the past 55 years. ..  yes it goes back that far!...you will see that a lot of energy has been put into idle emission values.
California must have been one of the pioneers in this matter.  Cars as early as 1962 -for that market- had to adopt positive crankcase ventilation and California has been red hot on idle emission because 90 % of the time cars are idling!

Accordingly manufactures put a lot of effort into mitigating idle emissions. Which brings us to the mid 70's MB spec of 1 to 2.5% CO at 580-620 rpm. * this value might have slight variances depending on end market destination. The point is,  that is a huge range from a small base at very low revs, measured over a paltry  swept volume. ie this is not a precise science.

As the revs rise in response to an opening throttle, mixtures and therefore burnt by- product vary like crazy. The miserable value of 1 to 2% -measured as part of a small volume of air fuel burnt at idle- is lost as 4 or 5 or 6 or 7 litres starts to suck and burn up to 13 cubic feet of air per second. Each cylinder is probably contributing slightly different values to the debate because the air flow is not the same even if the fuel doled out is the same value. [CIS ]
All in all the whole thing is just an average of an average and we shouldn't be getting too excited about nuances of 1 %.
By the way......one of the most celebrated cases of cylinder mixture variation...and thus requiring each throat of 3 double barrel side draft webers to be separately jetted -was the fabled 6 cyl Chrysler Charger '72 E49 . It'sa great case study.

Finally, in any case, to meet an arbitrary idle emission standard may not actually produce the optimal idle. You may find your engine idles best at a value outside the cited range. The 6.9 was never known for a smooth idle...how much of this is attributable to a suite of varying emission requirements over the years in different markets?
I say get the idle to be as smooth as possible. ..then for academic interest measure the emission. ! In a 40 year old engine you might be a bit surprised at the deviation.




raueda1

Thanks so much for that detailed discussion!  I guess I'll go for the analyser, why not? 

I know what you mean about the Chrysler Charger example.  For several years I drove a BMW 1800ti (yeah, google that!) with 2x 2 barrel Solex sidedraft carbs per the picture.  It was a really fun car to drive but carbs went out of sync very fast cause linkage was so complicated and who knows why else.  Getting equal airflow in each barrel was a bear.  I eventually got very good at using an airflow meter to keep it running well but I can't imagine trying to dial in specific emissions with a setup like that.
-Dave
Now:  1976 6.9 Euro, 2015 GL550
Before that:  1966 230S, 1964 220SE coupe, 1977 Carrera 3.0

s class

I have a great deal of experience in this area, and have used my Gunson CO meter on probably around 100 cars now. 

The CO setting on the K jet cars is extremely sensitive, and 1/4 turn equates to about 3% change in idle mixture.  I can confirm that changes in the idle mixture affect the mixture across the band. 

1. THe first step in setting up the car is to get the system and control pressures really correct. 

2. Get the cold start control pressure correct (or at least matched to the car's behavoir).  Typically the AAV's on these old cars are 'lazy', and so don't provide as much additional cold start air as originally designed.  Consequently, I find that WUR cold start pressure should be a little higher (ie leaner) than original spec .  ORiginal spec was typically 1.1 to 1.5 bar at 20 deg C.  I tweak them to match the car's AAV, and usually end up around 1.5 to 1.7 bar at 20 deg C.

3. Ensure that the CSV and TTS are working correctly, and not continually fuelling.

4. Ensure that the air mass meter flap disc is centered and the correct height

5. Ensure that throttle operation is correct - ranging from fully closed to fully open.

6. WIth the car properly warm, set the CO mixture for initial baseline operation - turn CCW (leaner) until the car starts stumbling (accompanied by additional sucking of air past the air meter flap) - this is too lean - turn back CW 1/4 turn. 

7. Set the idle speed to 800rpm using the idle air screw

8. Get ignition timing right - 0 to 5deg BTDC at idle with vacuum, 38-41 deg at 3000rpm with vacuum.  THis step is essential.  If you can't get these specs, then the distributor needs a rebuild, or you have vacuum leaks somewhere, or the vacuum plumbing is incorrect.

9. REpeat 7, the idle speed adjustment

10. repeat 8, ignition timing check.

11. Now, to tackle the CO.  The base adjustment above will probably have you around 3%.  Adjustments need to be made with the closing plug removed obviously, and this is an additional air leak ie false air.  Block the closing plug orifice with your finger.  If you hear no change in idle speed, then the setting is rich.  BAck off the adjustment in a CCW direction 1/32 of a turn at a time, until blocking and unblocking the orifice produces just the slightest, barely discernable change in idle speed - that will be about 1.5% CO.  If there is an obvious change in idle speed, you are lean, around 1%, the tipping point at which there is no change in idle speed is about 2.0%.

12.  Important ,... if you cannot achieve this result, you quite probably have ignition issues - coil, leads etc.

13. At this stage I usually recheck idle speed, ignition timing and then CO as above in 11, as they are all interrelated.

14. Final CO test with Gunson as described below.


[color=blue]'76 6.9 Euro[/color], [color=red]'78 6.9 AMG[/color], '80 280SE, [color=brown]'74 350SE[/color], [color=black]'82 500SEL euro full hydro, '83 500SEL euro full hydro [/color], '81 500SL

s class

OK, the first thing to decide when using the Gunson is what actual CO figure you are targetting.  Spec for the K jet 116 and 107 cars was usually 0.5 to 1.5%, but that is at sea level.  K jet self compensates for altitude, but is limited in its scope to about 1200m.  I'm in Johannesburg at about 1700m, and it is generally accepted practice to aim for 1.5 to 2.0% here for two reasons :
a) to deal with being outside of the altitude range that Kjet can cope with
b) these engines are old now with stretched timing chains, carboned up ports etc etc.

The Gunson works by comparing the exhaust gas to ambient air.  Ambient air has a typical background CO level of 2.0%.  The Gunson unit is extremely sensitive.  It takes about 30 mins powered up before it achieves stable readings.  If you try before that you will be frustrated.  After about 15 mins of being powered up, adjust the calibration knob to 2.0% (ie the background air).  Its easy too see when its stabilised, as if you wait another 5 mins, it should still be at 1.9 to 2.0%.  If it has dropped below that, it means that the initial warmup phase was not completed before you calibrated for background air.

The gunson is intended to be powered off the car's battery.  This does NOT work, because the battery voltage fluctuates as the engine is revved., buggering up the calibration of the Gunson.  Much, much better is to power the gunson off another car battery, loose on the bench.  We go a step further, and run it off a dedicated, stable 12V supply unit. 

Also, bear in mind that when you stick the Gunson sniffer up the tail pipe, it slowly gets polluted with exhaust gasses and so the test duration is limited to a few minutes.  If you carry out a test on a properly warmed up car, properly set up a 1.5%, you would see the following :
a) initial reading 2.0% (due to calibration in background air)
b) within 15 secs, reading will climb to about 2.2 or 2.3% as the measurement cell acclimatises to the hot exhaust gasses
c) after about 30 secs, the reading will slowly drop, in steps of 0.1% until it reaches 1.5% (or whatever the car is set at)
d) after about a minute, the readings will climb slowly as the Gunson's sensor cell gets polluted.

The minimum measured ie in c) above, is your reading.

My procedure is as follows :
1) power up the Gunson for about 15 mins
2) drive car hard preferably on highway for about 15 mins to burn chambers and plugs clean
3) perform test as per a) to d) above
4) adjust CO screw as required
5) go for another 15 min drive, during which time the Gunson cell will burn clean and it will settle back to 2.0% calibrated setting
6) perform test as per a) to d) above
7) repeat this until you get your setting.

My experience is that if you fail to do proper highway speed driving, the plugs will be a bit foulled, and one will get a false high reading.  The difference it makes too the result is about 0.5%.  In other words, if you set the CO level by only idling the car and driving around the block a few times, and you set it at 1.7%, then after a proper highway burn, if you retest, you will find it is actually about 1.2%. 

Setting the CO below about 1.5% (notwithstanding my comments above about altitude) will result in a noticeable 'seat of the pants dyno' drop in performance.  Going down from 1.5 to 1.2% will cost you about 10% in perceived pickup.  Conversely, going richer, above 2.0% will have a similar loss of pickup.  Going to 3% will feel noticeably 'stodgy' and lazy in comparison to 1.5% to 1.7%. 

A final comment, since the Gunson works by comparing the exhaust gas to the ambient air (its reference), it follows that you cannot do this in a closed garage or workshop where the background air is contaminated with exhaust gases.  We always do the test outside the workshop in open air. 


[color=blue]'76 6.9 Euro[/color], [color=red]'78 6.9 AMG[/color], '80 280SE, [color=brown]'74 350SE[/color], [color=black]'82 500SEL euro full hydro, '83 500SEL euro full hydro [/color], '81 500SL

raueda1

Quote from: s class on 18 July 2018, 04:33 AM
OK, the first thing to decide when using the Gunson is what actual CO figure you are targetting.  Spec for the K jet 116 and 107 cars was usually 0.5 to 1.5%, but that is at sea level.  K jet self compensates for altitude, but is limited in its scope to about 1200m.  I'm in Johannesburg at about 1700m, and it is generally accepted practice to aim for 1.5 to 2.0% here for two reasons :
a) to deal with being outside of the altitude range that Kjet can cope with
b) these engines are old now with stretched timing chains, carboned up ports etc etc.

The Gunson works by comparing the exhaust gas to ambient air.  Ambient air has a typical background CO level of 2.0%.  The Gunson unit is extremely sensitive.  It takes about 30 mins powered up before it achieves stable readings.  If you try before that you will be frustrated.  After about 15 mins of being powered up, adjust the calibration knob to 2.0% (ie the background air).  Its easy too see when its stabilised, as if you wait another 5 mins, it should still be at 1.9 to 2.0%.  If it has dropped below that, it means that the initial warmup phase was not completed before you calibrated for background air.

The gunson is intended to be powered off the car's battery.  This does NOT work, because the battery voltage fluctuates as the engine is revved., buggering up the calibration of the Gunson.  Much, much better is to power the gunson off another car battery, loose on the bench.  We go a step further, and run it off a dedicated, stable 12V supply unit. 

Also, bear in mind that when you stick the Gunson sniffer up the tail pipe, it slowly gets polluted with exhaust gasses and so the test duration is limited to a few minutes.  If you carry out a test on a properly warmed up car, properly set up a 1.5%, you would see the following :
a) initial reading 2.0% (due to calibration in background air)
b) within 15 secs, reading will climb to about 2.2 or 2.3% as the measurement cell acclimatises to the hot exhaust gasses
c) after about 30 secs, the reading will slowly drop, in steps of 0.1% until it reaches 1.5% (or whatever the car is set at)
d) after about a minute, the readings will climb slowly as the Gunson's sensor cell gets polluted.

The minimum measured ie in c) above, is your reading.

My procedure is as follows :
1) power up the Gunson for about 15 mins
2) drive car hard preferably on highway for about 15 mins to burn chambers and plugs clean
3) perform test as per a) to d) above
4) adjust CO screw as required
5) go for another 15 min drive, during which time the Gunson cell will burn clean and it will settle back to 2.0% calibrated setting
6) perform test as per a) to d) above
7) repeat this until you get your setting.

My experience is that if you fail to do proper highway speed driving, the plugs will be a bit foulled, and one will get a false high reading.  The difference it makes too the result is about 0.5%.  In other words, if you set the CO level by only idling the car and driving around the block a few times, and you set it at 1.7%, then after a proper highway burn, if you retest, you will find it is actually about 1.2%. 

Setting the CO below about 1.5% (notwithstanding my comments above about altitude) will result in a noticeable 'seat of the pants dyno' drop in performance.  Going down from 1.5 to 1.2% will cost you about 10% in perceived pickup.  Conversely, going richer, above 2.0% will have a similar loss of pickup.  Going to 3% will feel noticeably 'stodgy' and lazy in comparison to 1.5% to 1.7%. 

A final comment, since the Gunson works by comparing the exhaust gas to the ambient air (its reference), it follows that you cannot do this in a closed garage or workshop where the background air is contaminated with exhaust gases.  We always do the test outside the workshop in open air.
This is great, super helpful, thanks for getting this info out there.  I do have a question though that I can't quite make sense of.  I read the Gunson manual where they state that the CO level in air is 2%.  :o :o :o This is just flat out incorrect, if it were true we'd be dead.  Ambient atmospheric CO levels are typically in the ppm range.  However, on another forum I read that what they really mean is ambient CO2 levels -- that the device is actually measuring CO2 and calculating a number for CO using ambient CO2 as a proxy for CO.  But this troubles me too, since ambient CO2 levels are only about 0.04%.  Finally, there's CO2 in the exhaust too, lots of it, so how does the device distinguish between that and CO? 

Apparently it's still a useful tool but there's something about the principles that I'm missing.  The numbers are just screwy.  But it still seems like a good thing to have, especially if it can actually improve seat-of-the-pants performance.
-Dave
Now:  1976 6.9 Euro, 2015 GL550
Before that:  1966 230S, 1964 220SE coupe, 1977 Carrera 3.0

Type17

I've used a Gunson's CO meter on my cars since '94 and, after s class's comprehensive instructions above, I have nothing further to add, other than it's a great bit of kit to have to hand for any pre-cat petrol car.
'76 350SE in Silver-Green

adamb

I got one last year. It's good as an amateur tool but doesn't beat garage analysers. The main limitation is the absence of HC measurement.

Pop Alexandra

Quote from: adamb on 19 July 2018, 05:25 PMI got one last year. It's good as an amateur tool but doesn't beat garage analysers. The main limitation is the absence of HC measurement.
I have the same feeling. Good to get an overall idea, but I'd go for other manufacturers in case you need accurate measurements.

ramiro

I just baught a AFR gauge with a wideband sensor and welded a bung to the exhaust, when i want to adjust the car i just temporaly add the gauge to the car , also that way i am able to check the mixture at load and that way found that my overhauled distributor was to lean and had to be overhauled again.

Feather535

Reviving this old thread.  I just acquired one of these instruments, but after unpacking and assembling it and going through the set-up procedure, it won't calibrate correctly.  After waiting 10 minutes with the unit connected to 12v power, the display showed negative numbers (like -2.0).  Turning the adjustment knob all the way clockwise and waiting another 10 minutes only brought it to 0.7%, nowhere near the 2.0% the instructions call for.  Has anyone else encountered this issue?
--------
1977 280SE
1983 300D (sold)
1988 BMW 535is
1999 E320T (sold)
2009 E350T (wife's)

adamb

Quote from: Feather535 on 03 August 2023, 11:56 AMReviving this old thread.  I just acquired one of these instruments, but after unpacking and assembling it and going through the set-up procedure, it won't calibrate correctly.  After waiting 10 minutes with the unit connected to 12v power, the display showed negative numbers (like -2.0).  Turning the adjustment knob all the way clockwise and waiting another 10 minutes only brought it to 0.7%, nowhere near the 2.0% the instructions call for.  Has anyone else encountered this issue?
No not had that problem. Turning the dial works correctly on mine.

BMA

Would like to add something according to the question, what is a good Co value for the 6.9 engine!?
Just a statement in advance: I am not an expert!
I will probably start soon an extra post for that topic as soon as I get more information and measured datas. My car runs very good and after getting a refurbished warm up regulator (if this is the right word, German word: " Warmlaufregler") my expert told me to let measure the Co value at idle. According to the Mercedes speficatins it was exactly 1,5 % at idle around 750 rpm. He said, that is too low. The Mercedes specs are also too low just for a 6.9 M 100 engine. He is running for years with his car at 3.5%! He told me they measured that with Co 3.5% you get nearly Lambda 1! So I made a date with my local Mercedes dealership and with an obvious experienced guy we increased the value. He confirmed that the Co should be more than 1,5% but thinks 3,5 should be too much. Before we did something he connected his service device to my engine measuring the ignition angle. Everything ok. We ended with Co 3,1% at 750 rpm and Lambda 0,92. He told me I can come back if  it is too high and fuel consumption would increase. I am just testing the result and can currently just feel the car accelerates faster, is getting 230 km/h faster, runs very smooth and turns on very good in every condition. It seems even my fuel consumption went down!....Will let you know more as soon as I have more and better valid datas! Regards from Germany, Björn
happy having finally a 6.9 # 7271 in my garage, ...so nice to cruise!