It's also possible to get 18:1 AFR reading if only 3 of 4 cylinders are getting fuel. Cylinder number 4's idle passage was blocked by fuel varnish? during a very cold damp night time drive.

It would drive on the starter circuit, then rubbish whenever I put the "choke" in. Dunno if 3 cylinders were then really rich and one was not sparking, or the starter circuit supplied enough to all cylinders drive upto 40-ish mph. Hard to tell in the dark, as my vacuum gauge is not backlit. 

 

I think ice on the carb body got sucked in the idle circuit inlet hole on back of carb. I initially thought that hole was for starter, but no that takes air from holey front of jet cover. The ice melted and dislodged ancient varnish and then blocked. That's my theory. 

 

Edited 5 February, 20188 yr by dvderlm

Found some nice guideline figures for idle, cruise, high load and acceleration. Camshaft overlap makes a difference.

Including Bosch measured max power at 12.5 to 14:1, max economy at 16.2 to 17.6:1 and 12 to 13:1for peak torque.

 

Idle can be stoich (especially with a cat to soak up the NOx) or as rich as 12.5 for wild cam, or 13 to 13.5 in better behaved.

So my MOT measured 13.9 with correct advance appears nearly green in comparison 

 

http://www.autospeed.com/cms/article.html?&A=1595

 

 

In my recent trials in very cold air at 17.2:1 with no foot on any pedal down slight slope at ~1100rpm=28mph in 5th the engine still pulls just. Any leaner - pushing throttle gently so pump jet action and it is limp.

Edited 3 March, 20188 yr by dvderlm

The Sorg conversion with twin weber 40 DCOE 136/137 carbs for HH engine has Tuv emissions spec of 3.3% CO at 1030rpm.

That's 13.3:1 AFR.

https://www.motor-talk.de/bilder/suche-ein-gutachten-fuer-40er-webervergaser-fuer-meinen-polo-86c-1-3l-g75386024/sorg-6-i208605156.html

This year's MOT - easy emissions pass.

 

CO 1.589 = 13.9:1 AFR

and 1088ppm HC.

 

Idle speed was a little high, but I'll blame the mended accelerator pedal mount screwing up the cable tension.

(sadly failed on fuel-tank filler neck leak)

how did thier lambda reading compare to your gauge? 

About 0.7 richer than my gauge which wanders to settle about 14.6 when hot (AEM wideband with Bosch Lsu 4.x sensor), but I was n't present when reading was done.

Dunno if bonnet was open or engine really hot. Speed was a bit over wanted but analogue tacho is not accurate versus digital timing light.

It was about 4 degrees outside temp  when I arrived at garage, with all its bay doors wide open. 

 

I could look at the logs and try to work it out, but I don't know how many times car was started or moved. They got it on the brake rollers for first time in ages as on winter wheels with new winter tyres.

Edited 22 November, 20187 yr by dvderlm

And passed again (been SORNed while I welded in a new passenger side floor).

Failed first time with 9683ppm HC which was sooted spark plug in cylinder 4.

Took some BP Ultimate 97 petrol and 4 vacuum gauges then idle air bleed open on cylinder 2 to weaken its vacuum to match the others within 1.5psi or 3inHG.

Cylinder 4 feeds the brake servo so it is always a bit lower.

CO 2.76 = 13.5:1 AFR

and 1180 ppm HC.

Same 44 in 7850.1 idle jet.

Dergo air filters with RamAir foam and short trumpets. WR5DS silver plugs

Pretty sure that the logging AEM gauge thinks the AFR was 14.2 to 14.9 during the test period.

So at least 0.7 leaner than the SunPro 4 gas analyzer.

I suppose it is the unburnt fuel my wideband lambda sensor does not detect, just sees more oxygen.

Edited 29 October, 2025Oct 29 by dvderlm

Question, can you get an AFR meter that will go down the Polo GT manifold sampling point? Either a permanent gauge or just a temporary probe, either will do; it will help me with setting up my Mk2 PB AFM.

Dunno, sorry. Not got a GT exhaust manifold.

I think all lambda sensors have the same thread of M18x1.5.

But you need a wideband sensor and gauge to do tuning under load. A narrowband sensor will only tell you if its stoich, which is okay for idle setup, but a Gunson "Colortune" see through sparkplug is cheaper & better for that.

I think the Dergo filters are keeping more petrol vapour laden air to be sucked back in through the small holes on back of carb that feed idle stacks with air, than with open trumpets in foam socks and those holes unfiltered.

(while the jet covers must feed the starter jet circuit with air because they have a gasket).

So perhaps smaller idle jets needed.

But Sorg 1008 plenum design also feeds those holes for Webers and the Sorg Gutachten uses 55f21 idles ...

So 44 idle jet is stingy by comparison.

More progression holes in my Dellos than Weber though.

Tim Engel says count the mixture screw turns which should be about 3.5 to 4 from bottom(?)

for DHLA on Lotus (midpoint of highest vacuum).

More and is too lean. Less and is too rich.

I can use a colortune plug to check that it's near stoich idle too, and also see what removing the jet cover and gasket does.

Edited 31 October, 2025Oct 31 by dvderlm

And the Nova Sport with the Irmscher UK plenum that is spitting image of the Sorg used 59f21 idles in road spec. But only 100 main.. must run on the idles to about 4500rpm.

No, hold on ... I'm not sure it cover the rear small holes. Just 4 round tubes underneath not 6. ??

https://www.novasportregister.com/history/the-sport-pack/airbox-carbs/

Found a pic confirming the small rear holes are not in the plenum but free air with a Sorg airbox. https://www.clubpolo.co.uk/topic/123349-sorg-motorsport-twin-40-airbox/

On 31/10/2025 at 20:52, dvderlm said:

Dunno, sorry. Not got a GT exhaust manifold.

I think all lambda sensors have the same thread of M18x1.5.

But you need a wideband sensor and gauge to do tuning under load. A narrowband sensor will only tell you if its stoich, which is okay for idle setup, but a Gunson "Colortune" see through sparkplug is cheaper & better for that.

I have a colourtune so i'll be using that initially. It occurred to me that something more permanent would probably help quite a lot, as a colourtune is only safe up to 2000rpm.

vacuum gauge is way underrated in my opinion.

1/10th the price of a wideband meter.

And 1/3rd the price of a new colortune, crikey they have gone up in price - must be all the restored minis.

Can set idle and static ignition timing.

Instantly see when 2nd stage opens on progressive carb.

Can even film it and the road with a dashcam and drive up and down hills then watch the recording back. Repeat with different carb jets, dynamic ignition timing etc. If records sound you can add seat of pants feedback too between way points like lampposts for timing.

From inlet manifold below throttle plate or from all 4 inlet runners tee pieced through a brass snubber (or plastic fish tank one way valve).

Edited 3 November, 2025Nov 3 by dvderlm

On 29/10/2025 at 19:18, dvderlm said:

So at least 0.7 leaner than the SunPro 4 gas analyzer.

A friend suggests the 4 gas analyzer is calibrated for E10 so reads about 0.4 AFR richer when E5 fuel is burnt. This plus the unburnt fuel from unheated inlet manifold and air filter in cool garage would make a difference. Wideband gauges are accurate to about +-0.1 AFR.

And the BP ultimate 97 has no Ethanol if from Fawley refinery or depot.

So maybe 4 gas analyzer using Brettschneider equation would read 0.88 richer if gas analyzer calibrated for E10.

I had a mix, so maybe E2.5 ?

Edited 9 December, 2025Dec 9 by dvderlm