Carb Tuning by CO Gas Analyser

[Sin_Tiger]

Not being a carburation    I've always struggled to get the idle air adjustment as good as it was from the factory and spent many hours riding / tweaking and often loosing track of how many turns out I had each one. I've tried using "Colourtune" plugs, even going to the expense of buying a multi cylinder set, yes it will run with three Clourtunes as plugs but I still found it tricky.

I know in the factory they used a gas analyser connected to each cylinder down pipe, that's what those little M6 hex screws are for in the downpipes. I'd read that they often strip as they come out or the threads get damaged and they won't stay in place when you refit them but I decided to give it a try and use high temp Thread Locker if it didn't go well.

I had a Gunson CO analyser sitting on the shelf that I'd used a handful of times but never very conclusively on bikes sampling out of the end cans. Reading what literature that came with it didn't expressly say it wouldn't be able to handle gas straight out of the downpipe so I decided there should probably be enough cooling of the gas along the 0.5m length of alloy tube and plastic hose not to melt the sampler pump.

I did not want to connect the hose directly to the downpipe for fairly obvious reasons, so I decided to make a set of spigots from some scrap brass bar with an M6 threaded bore on the outer end which would allow me to unplug only one cylinder at a time and not be screwing the plugs in and out of the downpipe, potentially wearing the delicate thread, waiting for the downpipe to cool or working with hot screws.

Slipping the hose onto the end of the spigot on the first test, it didn't take long to realise the end of the spigot was still too hot as the hose slowly melted. I managed to find a close enough piece of tube to allow me to attach the alloy pipe (about 200mm long) with the hose on the end. There was enough heat loss to prevent the hose even softening to the point where it would fold, so I decided that would probably be enough temp drop to prevent the sampler being damaged.

I warmed the bike up and set the analyser to calibrate in fresh air and adjust to a stable 2.0% CO. Once everything was stable and I had the carbs vacuum balanced I connected the analyser to the first cylinder. Starting from the usual base setting of around 2.5 turns out I had a reading of 6.5% CO and started making small (1/8th of a turn) adjustments waiting for a stable reading after each adjustment.

I was expecting the readings to be quite laggy but I was pleasantly surprised to note a change within 5-10 secs and just to be sure it wasn't a fluke, adjusting back out the same amount and getting within 0.1% of the original reading. Continuing in the same manner I managed to get the the reading to a stable 3.0% and returning to that after a short throttle burst, the CO increased as expected and  then reduced with a fixed RPM as you'd expect. I had screwed the jet in by less than 1/2 a turn.

I repeated the operation on cyls 2 (needing a bit over 3/4 of a turn) & 3 (needing a little over 1/2 a turn), did a small balance adjustment and allowed the analyser to sample fresh air in between each cylinder test and it returned to within 0.1% CO of the initial 2% setting each time. I repeated each cylinder test without adjustment and got within 0.1% CO of the previous samples without any adjustment.

I refitted new SUS screws with copper washers as the threads on the original MS screws were a little rounded and I didn't want to risk stripping the threads in the downpipes.

Subjectively, the engine is running as even sounding as any I've ever done, it revs without any hesitation and returns crisply to idle without any fluffing or hunting.

As a baseline, mileage is just under 36k miles, the engine has had a good clean out including the sump, fresh oil and filter, carbs stripped cleaned, clean air filter (split), clean plugs, inside of the down pipes de-burred, valve clearances all in spec and all within 0.05 of the factory shimming, I used 1.5 litres of E10 fuel with a dash of Seafoam, weather was overcast with almost no wind, about 16C ambient and about 65% RH. The Gunson's "Gastester" I used appears to have been superseded with a more compact version.

Some things I should have done, take more photos  , perhaps put the bike on a paddock stand (it has no centre stand) and check the downpipe temperatures. I'm looking forward to riding the bike, seeing how it behaves and getting some fuel consumption results.

[Lee337]

Far more patience (and experience) than me with all things carbed. This is probably why I ended up taking the trophy to a local company that tunes bikes.

An interesting read though, especially the process you followed to get the balance.

hats off to you. 

[ghulst]

And in this other post you said you were going off carbs...  The whole fiddling with the mixture etc. is the reason why I take my bikes to be set up. Yes, it does cost a bit of money, but it is a whole lot easier and the end result 1000% better than I could ever achieve.  

[Beernard]

Thanks Sin. That's a great read and I applaud your curiosity and obvious tenacity and skill. I truly am interested in the road test. I'm expecting happy days!
Regards from Oz,

[Sin_Tiger]

Well after a clean MoT and the first 100 miles I can report that this technique appears to work getting it right straight out of the box. Very smooth with none of the usual tell tale hesitancy around 2500 - 3000 rpm, no pops or farts and    response is as good as I've ever experienced even after a dyno set up, not saying a dyno run might get it even better fine tuning the needle positions 

None of the screw plugs have fallen out but the first 100 mm of "VHT" gloss black BBQ paint has burned off, I had expected that as only matt black VHT seems to work, on the plus side the clean areas on each pipe are very consistent in length and pattern. I forgot to mention, I did polish off the excess weld bead on the inside of the downpipe flange joint.

It will be interesting to see the fuel consumption results after a couple of tank fills but overall I'm happy with the throttle response in all conditions and won't have to spend hours trying to get it settled.

[ssevy]

Nicely done!

[London_Phil]

No more pops n farts eh?, but what about the bike?
Looking good Sin, if they did an efi conversion, I would never have sold mine...
I wonder what happened to the guy who did one, a few years ago?, he was on the forum..

[ghulst]

The big question is whether there is any EFI kit you can fit from another model, or do something with a programmable ECU. In cars this has become more of a thing in the past five years. You can convert anything now, but on bikes it seems to still be lacking.

[London_Phil]

https://tigertriple.com/forum/index.php/topic,4348.msg27293.html#msg27293

[Sin_Tiger]

No more pops n farts eh?, but what about the bike?
Looking good Sin, if they did an efi conversion, I would never have sold mine...
I wonder what happened to the guy who did one, a few years ago?, he was on the forum..

That was the guy in Greece    It was a nice job, especially considering the available technology at the time, he did seem to be something of an electronics specialist though.

The big question is whether there is any EFI kit you can fit from another model, or do something with a programmable ECU. In cars this has become more of a thing in the past five years. You can convert anything now, but on bikes it seems to still be lacking.

I'm working on it 

[Sin_Tiger]

No more pops n farts eh?, but what about the bike?

Just back from a 600+ mile round trip up to the Assynt area, more on that in a ride report.

Some of the most challenging tarmac roads I have ever ridden and mostly truly awful weather on the way back, thank you Meteo you got it completely wrong 

Bike performed flawlessly, absolute joy to ride, after a time away from Steamers it has rekindled a spark  Returning mid to high 40 UK MPG, possibly still a tad rich but that may be down to worn jets or emulsion tubes which didn't get replaced.

Photo is Suilven in the background.

[Beernard]

Sounds like a great result, Sin. Fantastic that your effort has been converted to riding bliss! All the best from here. Bern.

[nightrunner]

I have not checked in on this forum in many years and I log on tonight to find this. Very cool. I'm surprised and impressed to read that someone else out there went to the this trouble. I did as well. I work in a research lab so am able to borrow a CO analyzer. I decided to check CO levels before rejetting and balancing. I made an adapter fitting to attach a long stainless tube to the sampling ports on the Tiger's pipes. I was shocked to find that all three cylinders were running at about 0.1% CO. No freakin' wonder my bike was so hard to start! The manual says something like 1 - 2.5% or something like that. So I pulled the carbs, cleaned them, and installed the larger main and pilot jest and shimmed the needles. I also rigged up a differential U-tube manometer using clear vinyl tubing and water. [And yes I had a reservoir to make sure no water got sucked into the carbs.] Now here's the secret. Most mechanics will use those mercury carb sticks (or worse, the vacuum gauges) where you measure the total vacuum on all three carbs simultaneously. The problem is you don't have much resolution comparing the mercury levels side by side. I measured differential vacuum between two carbs. The absolute vacuum is not something you need to know in balancing carbs, just that they are pulling the same vacuum. So my gadget gets connected to the reference carb and one of the others (a slave carb). The water in the U-tube gets pulled one way or the other. You adjust the slave carb until the levels in the U-tube are the same. That means both carbs are now pulling the same vacuum. Then repeat for the other slave carb. It even works for a pre-FI Goldwing with six carbs.

Once they were balanced, I connected the CO meter. The stainless tube was to cool the hot gas before it reached the analyzer. Then it was simple enough to adjust the mixture screw on each carb to get 1% CO. I went for 1%. Maybe it was 1.5.  This was like 15 years ago. Anyway after the rejet and tune, she starts easy and purrs like a big kitty. No flat spots in the RPM range.

Anyway, that was a nice write-up. Glad it worked out for you as well as it did for me.

[Sin_Tiger]

Good share  , the differential vacuum adjustment makes sense, I'll give that a try next time and I agree about the mercury gauge resolution.

[Madruss]

Good read, thanks for the effort