Let's tune what we've got the Keith Franck way...
The LM-2 kitYou can see the Bosch wideband 02 sensor in it's bung just after my y-pipe. When I am not running the LM-2 unit I can unscrew the sensor and put a blanking screw in. It can damage a sensor to run it without power. (hmm must redo the teflon tape on the gearbox lower bolt)
The TPS (Throttle Position Sensor) is mounted using bits that weber supply with the device. I need some small spacers to mount it a little straighter but it works fine as is. It is a, lightly sprung, rotary pot of about 5K (I think) so the three connections are 5V, GND and Output.
I mentioned before that I used the Speedy Cables RVC conversion circuitry inside the tach to provide a stable 5V for the TPS and a stable pulse for the RPM reading in the logger. The back of the tach also made a handy point to provide a common ground for the system. So of the three cables one goes straight to the O2 sensor to power it and read the output. One gets 12V to power the unit, which I pick up behind the tach. The third has the rpm pulse input and the tps input, again all found behind the tach.
I used an old suction phone holder to mount the LM-2 on the window and I can hang it in reverse for reading the settings whilst under the hood.
This is brilliant as you can instantly see changes in AFR (AIR to Fuel Ratio) and RPM whilst fiddling with the carbs. Plus the RPM is much more accurate than the Smiths gauge.
The float level.The Keith method is to have the fuel 2mm below the passage way for the auxillary venturi. This equates to 25mm below the deck that the jets screw into.
Best done with the engine running but I find it easier to do it just after it has stopped. Using a night light for a kindle (!) I get a reflection on the surface of the fuel. Lowering a small screwdriver into the hole I can easily see when the surface of the fuel is broken. The addition of a bullet connector sleeve onto the shaft means it records how deep this was like this:
Hmm 26mm, not bad, with keiths +-1mm tolerance and certainly not enough for me to dismantle the carbs to fiddle.
I should probably buy an optical device from Keith but I'm used to this method of mine now.
If you set your float any other way at any other level then you don't know what your missing!!
The BalanceLook at me - four hands! Get yourself one of these nifty STE devices they are brilliant.
This was done at about 1000rpm. It's not bad, all of them are about 4Kg/m3.
Eagle eye'd will notice that number 4 is a bit low.
It used to be really low until I did this:
Down the barrel of number 4, can you see it? Look closely at the lower half of the butterfly.
See the 1mm hole I drilled!
Now if I had more modern DCOE rather than the 18s you see here then the air bypass adjustments could achieve this instead.
As it stands next time I take the carbs off I will enlarge the hole a little.
The IdleI currently have 45f9 in all the barrels, this is as per spec for 40DCOE18 on a Twink. I have tried the weber alternatives around this value e.g. 45f8, 50f8, 50f9 and none are as well behaved through the transition circuit as the 45f9. The 45f8 were next best but will have to test them again to quantify this.
So ensuring that the idle never goes over 1000rpm I close the mixture screws in turn and mark them with a pen at 6 o'clock. The engine should stumble when any one screw is closed or you have an air leak! I open each up about 1/2 a turn then go back a bit until I find the max rpm (careful of that idle). I find that this max rpm is on a bit of a plateau rather than a peak so I go for the weak end of it on all 4 barrels. This results in an AFR of 16-17. The "best lean idle". All very well but this results in stupidly lean mixtures, above 20, when you are actually driving - especially on the over-run. So as per Keith's wisdom all the barrels are made a bit richer until the LM-2 reads 12.5.
That's it.
So here is the spread of AFR I get. What we are after is as much 12.5 as possible:
Hmm must be careful though. This is a 45min run across town: barely ever WOT, mostly under 3000rpm and lot of idle (traffic lights).
With a lot of idling in the data I am bound to get a 12.5 peak 'cos I set idle at 12.5 earlier!
So I graph across a shorter section of the same run and selected a portion where I was not idling (the rpm and tps data allowed me to do this)
Not as good eh? This is what I hope to improve with Keith's HypoJets.
If I get a chance I will do some more controlled runs at higher speeds but need to get out of town to do this.