Fiddling the Mikunis on the Kat 750

[JMHJ]

Ah, Indeed. Thanks for the explanation. The art and science of combustion engineering (and airflow through the system, etc.).

I've thought on occasion over the years that if the engineers of these things would/could effectively pass on their knowledge to the end users it would make me at least, a much better mechanic. The guy that understands it all well (cars or bikes, etc.), or maybe even one significant part of it all, seems kind of rare. The engineer that understands all of it completely is probably kinda rare.

I guess that info's out there, if you know how to separate it from the myths and other extraneous stuff it may be surrounded by.

[KiwiAlfa156]

Ah, Indeed. Thanks for the explanation. The art and science of combustion engineering (and airflow through the system, etc.).

I've thought on occasion over the years that if the engineers of these things would/could effectively pass on their knowledge to the end users it would make me at least, a much better mechanic. The guy that understands it all well (cars or bikes, etc.), or maybe even one significant part of it all, seems kind of rare. The engineer that understands all of it completely is probably kinda rare.

I guess that info's out there, if you know how to separate it from the myths and other extraneous stuff it may be surrounded by.

Its the trick, isn't it? Understanding things from first principles, and applying logic. Doesn't mean you'll always get it right, but there's less chance of getting it wrong. Trusted sources of information are the answer to cutting through the noise. Over the years I've read a lot of guys like Kevin Cameron (Cycle World) and A. Graham Bell, both guys who were top tuners back in the day. Bell, in particular would support his tuning theories with dyno numbers. 'To measure is to know' is one of our fellows tagline and its true. What worked what didn't? They wrote/still write for the layman, and thus form that bridge for the highly technical and mathematical to the practical. Just the ticket for the curious. Some of their stuff may appear dated these days, old engines, and some old tech, but the principles haven't changed.

[KiwiAlfa156]

I don't own your model and I do appreciate the detailed work and trial and error you're sharing.
The leak at the coil and the compromised rubber parts are basic maintenance things, meaning you'll waste time if these things aren't checked first. Carburetion tuning always comes last after everything else is certain.
Just to mention, you said "the pilot circuit does draw fuel through the main jet". That's not correct. Your carbs aren't designed that way. I also question why the plugs vary so much in color, cylinder to cylinder. Did the bike pass a compression test? Did you carefully set the float levels? Any chance you mixed up the float valves and their seats while cleaning? They wear as a unit and mis-matching them will cause issues only new replacement will cure. Did you carefully vacuum synch the carbs?
One other thing, only my opinion, changing to full synthetic at 600 miles is too soon. Change at 600 yes, but conventional oil is best for full break in. Run an additional 2,000 miles minimum.

Hi Keith, sorry missed your post and questions. I'm carb tuning after a major rebuild of a junker. The resurrection journey is posted here.

https://www.thegsresources.com/_forum/forum/technical-forums/projects-rebuilds/91822-750-katana-resurrection

The coils, spark leads, and rubbers are all fitted new, so lack of maintenance wasn't likey to be problem.

I have to disagree regarding the pilot jet fuel supply. It is sealed by a nitrile bung and can only access fuel via a fuel supply port downstream from the main jet. This from the Suzuki factory workshop manual.


I fitted new float valves/needles and float levels set using the clear sight tube method. So uneven float levels was/is unlikely.

The pistons and bores were in good condition and measured within spec. I honed the barrels and fitted new rings. I replaced 3 valves and 3-angle cut the seats. The valves were lightly lapped in and the valves were leak tested on the bench.

Ring seating was the primary focus of running in, as along with the valve job, these were the only virgin parts in a 60,000 mi engine. I used a specialist running in oil and carefully varied load and rpm, with lots of heat cycles. IMHO there isn't much to be gained by extending the running in period as the bulk of ring seating and valve seating occurs in the first part of the running in period. Everything else in the engine needed limited time for previously mated parts to 'settle in' rather than 'wear in'.

I'm confident that the rings seated at this stage as there is no oil consumption and the engine is spinning freely and smoothly in the rpm range on the pilots and when the one of the coils isn't breaking down.

I'm now about 90% sure that the failing (new) coil is the cause of the uneven colour as it's affecting cyl 1 and to some extent 4. I didn't intially pick up the coil failure as I was in process of jetting when the failure caused changes to the the plug color that I assumed were the results of jetting changes. I have fitted a replacement coil today and will resume jetting and post the results. In the future, it's information that hopefully might be of use to someone else.

[KiwiAlfa156]

Ok 90% sure has become 90% uncertain. The coil gave smoother running, but the rich mid-range stutter is still there and the plug is still super rich on cylinder 1...

Spent some of my vacation thinking about the problem and possible causes. At this stage I'm thinking the main air bleed is clogged in carb number 1. I removed and cleaned the needle jet/emulsion tube last time I had the carbs out. But didn't check the air jet or air gallery to the emulsion tube. A blockage here would result in a rich mixture out of the needle jet.

So tomorrow the carbs get pulled again. Fingers crossed. This time.

As someone else has said, once you eliminate the obvious you're left with the cause.

[KiwiAlfa156]

So, central to this thread is the undertaking to be upfront about the process, warts and all.

I pulled the carbs this afternoon and removed the slide and dropped the bowl of carb number 1 with the intention of visually checking the clear flow through the main air bleed, which had become my prime suspect.

As soon as the bowl came off the cause of the rich running on cylinder 1 was obvious. The nitrile bung which seals the pilot jet was missing... As I'm the only one working on the bike, mea culpa. What a dick!

In the Mikuni BS the pilot jet draws its fuel from the main jet. Or from an alternative perspective the main jet meters fuel for both the slow (pilot) circuit, AND the high speed (needle/main) circuit.

If the bung is missing, when the fuel starts flowing through the needle jet, it sources it's fuel from the massive unmetered 'bung hole' rather than the main jet. Hey presto the mix entering the cylinder is as rich as buggery, as soon as your off the pilot jet.

That one goes straight into the memory banks. I'm not worried about making silly mistakes because it's part of the learning process. And this thread is a journey.

The bungs were a little shrivelled due to being 40 years old and unloved. I found a spare bung and 'reconditioned' them in Methyl Salicylate (Oil of Wintergreen) this soaks into the nitrile and makes is swell, giving a good seal as well as smelling nice



A blow through all the jets with aerosol carb cleaner and 'careful' reassembly. Also took the opportunity at 1,600 km after rebuild to reset all the valve clearances, there was adjustment needed, which is to be expected with hand re-cut seats and a couple of new valves settling in.



A quick run around the block and the bike is pulling cleanly across the rev range. But is a bit lumpy at idle. Reset of all pilot screws to 1.5 turns out improves things. Plugs look good and even across the cylinders. At some stage I'll set the pilot (fuel) screws using my trusty old colour tune plug.



I'll do some proper plug chops for needle clip and main tomorrow, but the carburation is really close. It's got the feel of a quick bike, so the attention to the squish clearance (0.7 mm) and careful inlet port clean up might be paying dividends.

[slayer61]

Well done Sir. Excellent find and thanks for sharing your find with us!

[KiwiAlfa156]

Well done Sir. Excellent find and thanks for sharing your find with us!

You're welcome!

So the weather's not playing ball, with the decaying remnants of Tropical Cyclone Lola hitting the northern part of NZ with gales and some rain. Lola had previously smashed the islands of Vanuatu, west of Fiji, tragically killing some folk. So final jetting runs will have to wait until conditions are closer to ideal. I did crack out the "ancient and venerable color-tune plug" and set the idle mixtures. I tried to take pics of the process, but I find that the easiest way to 'color-tune' is in complete darkness, making photography impractical.



The engine was fully warmed up and I had my blower fan keeping the engine cool. The secret is to keep the idle at 1,200 rpm to ensure that the bulk of the fuel being burnt is passing from the the pilot outlet; the only outlet controlled by the fuel screw.

The adjusting of the fuel screw of the while peering at the plug is an interesting experience. The colour changes from pale bluish white, through blue-ish to orange. Which is lean through correct to rich. More interesting is that you can watch the evenness of combustion. When lean you can see the combustion is erratic. Then it smooths out to become regular and blue-ish, then regular and orange. I set the screw where the regular and blue-ish burn occurs and then repeat with the next carb until all four are done. The idle is now smooth and responsive to the throttle.

What I do find really interesting is where is that each of the fuel screws ends up in a different place in turns out from seated. I can only assume that wear from adjustment, or just through use or abuse, means that "turns out" is at best a starting point only. And that on an old, used set of carbs, uniform, arbitrary screw settings aren't likely to be giving the correct mixture.

[Jinks]

G'day.

[KEITH KRAUSE]

Kiwi: As soon as the bowl came off the cause of the rich running on cylinder 1 was obvious. The nitrile bung which seals the pilot jet was missing... As I'm the only one working on the bike, mea culpa. What a dick!

In the Mikuni BS the pilot jet draws its fuel from the main jet. Or from an alternative perspective the main jet meters fuel for both the slow (pilot) circuit, AND the high speed (needle/main) circuit.

If the bung is missing, when the fuel starts flowing through the needle jet, it sources it's fuel from the massive unmetered 'bung hole' rather than the main jet.

You were right about the pilot circuit drawing fuel through the main jet. I had "VM" carbs on my mind and should be more careful before I reply.
I also misunderstood the location of the "bungs" as you call them, compromising my advice.
The BS pilot circuit does draw fuel through the main jet, but the main jet does not meter or regulate fuel flow for the pilot circuit, nor the needle jet circuit. Fuel flow in ANY circuit is always regulated by the SMALLEST port or opening in the circuit. At closed or small throttle openings, the pilot jet regulates flow. For the needle jet circuit, the gap, or space between the needle jet and jet needle regulates fuel flow until the jet needle raises high enough inside the needle jet (approximately 3/4 throttle position) that the gap between them becomes larger than the main jet opening and that's when the main jet regulates fuel flow.
With the missing bungs, the open "holes" were not regulating fuel flow because they were not the smallest opening, but they would introduce a secondary flow that would result in a rich mixture.
I'm glad it appears you've sorted the jetting. I've been there with all the thinking, trial and error.

[KiwiAlfa156]

Thanks Jinks, but I ain't got em sorted yet and there's still plenty of scope to eff things up!

But with that in mind, took a ride down the local 'test facility' today and gave it some high speed testing. Which was cruising on the needles at around 140 kph (~87mph) and then winding it WFO and holding it. The idea was to see if I could feel if: (a) it felt right on the needles at higher throttle and (b) if I could feel a change as it shifted onto the main jet.

At the moment, I'm happy with the carburation off-idle through to revs of around 5000-6000 rpm. It's smooth, crisp and clean. So running on the by-pass and transition onto the needle is good. Idle needs some work and tonight I pulled the fuel screws and pilot jets, cleaned them and the pilot orifices. There was a bit of crud on the fuel screws, which may be the problem. Will re-colortune the fuel screws again once I'm happing with the main jetting.



The high speed runs has me thinking that the clip position is right at no.3, the standard position. The bike carbureted cleanly at higher speed/rpm and during changing throttle positions/acceleration. As I wound on to WFO held it there, there felt like there was a slight glitch as the speed built up and the revs got closer to redline. And after a few seconds a very gentle and erratic 'surge' -like your just starting to run out of fuel. I'm thinking lean. Currently the mains are 115s and are only one size up from the stock jetting. I've already run 120s and these gave rich plug reads. So when I got home I swapped the 115s for 117.5s, and hopefully these will be on the money. Will plug chop to confirm. It would make sense that the less restrictive filter set up would only really flow more air at WFO.

I bang on about the pilot by-pass orifices so I thought I'd show a pic.



The 'huge' port is the outlet from the choke/enrichment circuit (essentially a mini-carburetor integrated into the carb body). The single hole nearest the edge of the carb is the pilot outlet. This is the only orifice who's flow is adjusted by the fuel screw. The three holes upstream are the by-pass, metered only by the pilot jet/pilot air jet. You can see where the throttle plate sits when at idle, closing off the by-pass holes. As soon as throttle plate opens the by-passes are in operation, so all 4 holes are passing fuel/air. It's obvious why the fuel screw setting will have much diminished effects once above idle.

What is also interesting is how big the gap under the slide is when fully lowered in the bore. VMs, in comparison, have a crack of daylight underneath them.



This makes sense as it would be difficult to create sufficient air flow/vacuum though too small a gap. The slide is lifted by the vacuum created by the air flowing under the slide creating low pressure in the area above the diaphragm via a hole in the bottom of the slide. There must be a reasonable amount of butterfly rotation/opening before the air velocity under the slide is sufficient for it to start lifting. During this period the fueling is the pilot mixture + air under the slide and fuel drawn up past the needle. As the butterfly opens further it increases the volume of air passing under the slide which has to travel faster under the slide, creating more vacuum, raising the slide and letting in more air and fuel from the needle jet/main jet. As the slide rises, the speed of the air passing under slide slows back to what it was at the previously setting, but because the area under the higher slide is greater, more air/fuel is flowing. This is the constant velocity part of the CV carb. The carb seeks to keep the air speed under the slide constant at all throttle opening, once it takes over from the pilot circuit. More air flow at the same velocity = more vacuum at the venturi.

[KiwiAlfa156]

Kiwi: As soon as the bowl came off the cause of the rich running on cylinder 1 was obvious. The nitrile bung which seals the pilot jet was missing... As I'm the only one working on the bike, mea culpa. What a dick!

In the Mikuni BS the pilot jet draws its fuel from the main jet. Or from an alternative perspective the main jet meters fuel for both the slow (pilot) circuit, AND the high speed (needle/main) circuit.

If the bung is missing, when the fuel starts flowing through the needle jet, it sources it's fuel from the massive unmetered 'bung hole' rather than the main jet.

You were right about the pilot circuit drawing fuel through the main jet. I had "VM" carbs on my mind and should be more careful before I reply.
I also misunderstood the location of the "bungs" as you call them, compromising my advice.
The BS pilot circuit does draw fuel through the main jet, but the main jet does not meter or regulate fuel flow for the pilot circuit, nor the needle jet circuit. Fuel flow in ANY circuit is always regulated by the SMALLEST port or opening in the circuit. At closed or small throttle openings, the pilot jet regulates flow. For the needle jet circuit, the gap, or space between the needle jet and jet needle regulates fuel flow until the jet needle raises high enough inside the needle jet (approximately 3/4 throttle position) that the gap between them becomes larger than the main jet opening and that's when the main jet regulates fuel flow.
With the missing bungs, the open "holes" were not regulating fuel flow because they were not the smallest opening, but they would introduce a secondary flow that would result in a rich mixture.
I'm glad it appears you've sorted the jetting. I've been there with all the thinking, trial and error.

Thanks Keith, yes the VMs are different kettle of fish all together. Your right that the main doesn't provide the final metering for the pilot circuit. This is the job of the pilot jet and pilot air jets. However all the jets in the BS are 'in series' and I'm convinced that each jet has an effect on fuel flow/rate of the jets further down the fuel circuit. This is why I suspect Mikuni literature recommends that the first step in tuning carbs is to select the correct main jet before selecting the needle jet, needle, setting needle clip, etc.

The stuttering richness I experienced with the bung missing occurred as soon as the slide began to rise and about 1/4 - 1/3 throttle 4,000 -4,500 rpm -where metering was controlled by the needle/needle jet clearance. It was rich as soon as fuel began to flow past the needle. This could only be because fuel was being supplied directly from the bowl to the needle jet from the open bung hole under the pilot jet back down though the pilot supply gallery -therefore by-passing the main jet. The very same symptoms occurred when (again through my stupidity) a main jet unscrewed itself and fell into the bottom of the float bowl (bung was in so no secondary fuel path) - stuttering at mid-throttle and sooty black plug as soon as it was off the pilots. A 1 mm diameter main jet hole is going to provide more resistance to fuel movement in any situation than a 6mm hole. Incidentally, it was also rich at idle (sooty plugs) on the cylinder with both the bung missing and main jet out. The evidence suggests that the main jet is pre-metering the fuel to the needle jet (and to some extent the pilots); by-passing the main jet or removing it altogether resulted in super rich mixture exiting the needle jet regardless of the fact that the gap between needle and jet are the smallest restriction in the fuel flow.

If, as you say "Fuel flow in ANY circuit is always regulated by the SMALLEST port or opening in the circuit." Then you should be able to have a bike that carburates perfectly apart from WFO with the main jets removed as they are only the smallest port or opening at may be the last 1/8th throttle. From experience I don't think this is the case. But willing to be proven wrong.

So I can only conclude from the evidence from my unintended experimentation is that each jet has an effect (large or small) on the metering of any jet that is follows it in the fuel circuit. The first jet pre-meters the subsequent jet. The flow out of a jet must be related to the flow into that jet (after all that's how the main jet sets WFO mixture through the needle jet). An example of 'input affecting output' is how raising float level richens fuel mix across the all the metering in a carb. The higher fuel level reduces the signal strength (vacuum) required to lift the fuel out of the bowl and through a jet. The jet size is unchanged. Venturi vacuum is unchanged. The restriction to flow is fixed in brass. Yet more fuel flows because of the conditions at the mouth of the jets has changed. Fuel flow rate is variable subject to conditions. The size of the restriction at the jet or needle/jet isn't the sole determinant of fuel flow. Pressure differential, air bleed ratio, etc. all have an effect.

The more fiddling I do with carbs the more I understand that carburetors and are a dynamic system, and that although we are taught to think of them as being discrete but overlapping systems (because it helps make sense of the function of individual parts), they are sometimes best thought of in more holistic terms where a combination of factors affect their operation in any given state. Best advice I ever read was, "carbs don't read manuals". My trial and (much) error journey continues to convince me of that is truth in that.

[KiwiAlfa156]

Update.

Didn't do a WFO plug chop today. But did an hour or so of mixed riding after work with a short motorway run and through the local hills where I live. I am very happy with the carburation through the mid-range and up close to redline. It's very responsive and crisp. So just a plug chop to confirm the main jet is right, but it's running sweet.

After the ride, while the engine was still warm, I re-adjusted the fuel screws with the 'ancient and venerable colortune. I'd been tweaking the screws by ear and to be honest just made things worse. The idle was rough and the uneven burn across the cylinders made it vibrate coasting on a closed throttle. Yuck.

Great success with the fuels screws. Did a quick test ride and she's now smooth and pulls seamlessly from idle,and is silky coasting on a shut throttle. It's become a real pleasure to ride.

So almost done.

[KEITH KRAUSE]

Carbs are fun. Re-jetting for intake/exhaust mods are even more fun.
Not sure I understand your reasoning, but whatever works for you. There is an "over lap" effect when you transition from one jetting circuit to the next. There are so many variables in re-jetting for mod's. Sometimes there is no perfect solution and you have to accept minor issues. Sometimes the mod's we make are not the best match and no amount of tinkering is going to change that.
Just hang in there and don't get frustrated. You're not alone.

[KiwiAlfa156]

Just hang in there and don't get frustrated. You're not alone.

Thanks Keith, the tune on the carbs is close to perfect. So the tuning process has been a success. I'm just going to confirm the the main doesn't need to go up one more step, just to be sure.

I guess if I were to encapsulate my reasoning (which probably hasn't been that well articulated), it is 'what you learn from doing in practice should always modify the theory and beliefs that you bring with you before you turn a wrench'. I always believed that the smallest restriction in a fuel circuit did the metering -end of story, common sense, universal knowledge. Except I was confronted with (repeatable) evidence to the contrary. Measurable real life proof is hard to deny and the theory you hold has to change to accommodate the observed facts (the earth ain't flat). And when in light of what you have observed, you can change a variable and it results in a predicted outcome... well, that's science, isn't it? I'm glad to say I learnt something.

[KiwiAlfa156]

Update. Re-adjusted the 750's fuel screws with the colortune, as it was still not as smooth as my GS650G. It was rougher than the 650 decelerating on a closed throttle and had a slight glitch on a just cracked throttle at low rpm.

I noticed that the band of screw adjustment in which the mixture looked correct was relatively broad. So I adjusted for the most consistent and regular blue flame combustion at idle.

The closed throttle roughness and off-idle performance was improved but still not as good as the 650 which is a pleasure to ride around town because of good carburation at low rpm.

So I pulled the plugs after chopping in the glitch zone and got this.



The plugs are 4,3,2,1 and 2 is perfect. 4 is a touch rich, 3 is richer than 4, and 1 is the richest. With the plugs back in and the engine running, I added about 1/4 turn to 1, and the idle smoothed out and raised a little. I added about 1/16th turn in to 4 and about 1/8th to 3, with a small but noticeable improvement in smoothness and idle speed increase.

A test ride confirmed smoothness decelerating on a closed throttle and much improved responsiveness at small throttle openings.

A quick rev-out to redline top test top end in 1st and 2nd was positive for the main jet size with good power close in the 8,000+ rpm range and no sensations of either a rich or lean mixture. But the rev counter has exhibited some needle stickiness and I don't trust whether is is reading accurately. I might have to run a trusted accessory tacho in tandem to confirm accuracy.

I will video the final jetting testing results. But this was how it was running on Sunday after the colour tune but before this evening's idle plug chop.