"In the K0 and K1 CB750 with brass floats the floats were always staggered by 1mm or so. The "high" side float (relative to the bike leaning over on the kickstand) received the most-commonly recommended 26mm setting while the 'low" side float got 25mm instead. All [virgin] K0 and 657A series carbs were set this way, as far as I found. "
Mark
The confusion is in you mention just the low side, carb #1, 25mm, and the high side, carb #4, 26mm.
What about carbs #2&3? You don't mention those heights each time.
JHonda500 above has it right: it is the left & right float in each carb that I'm trying to describe.
Think of it this way: if the bike is on the sidestand the fuel is sloped inside the bowls, with the 'high' side toward the kickstand, That's the one that gets the 25mm setting, putting the float higher inside the chamber and thus using both floats to push the fuel valve shut. Without this little setting, only the kickstand-side float is pushing it closed, and this causes the float's bracket to dig into the opposite side's mounting post, which slowly builds up a tiny ridge there. Over enough miles, this causes the floats to not close at all when the float valve recesses into the soft brass float valve seat, and the bikes piddle fuel slowly with the petcock on when parked on the sidestand because the float valves are not being pressed closed all the way by just the one low-side float. I suppose the hope of the stronger float valve spring was to try to stop that (or else it was because it became a common part for ALL such carbs by making it as strong as the later carbs needed, IMHO).
If you have access to an old (and still OEM) set of the K0 or K1 carbs you'll see this in them, if they are factory-virgin, still. These also have tiny little edges adjacent to the notches in the float gasket surface: it is these little cast-in and also machined-off edges (on a per-carb basis!) that are the actual float reference Zero point. In the 657b and later carbs the middle body of the carb was raised a little bit and the float gasket surface, sans gasket, became the 'new' zero reference - IMHO this was done by a new engineer who was assigned to "cut costs" in the 750, and he didn't know the witchcraft of the early builders. It was then that the floats also became both 26mm height, and the Honda manual (the green one) was altered to show 26mm, even if you ordered one for the K0 bikes. This was never true, but over the years became the 'norm', causing much confusion (and arguments!) about setting the floats in what appears to be the same carbs from the K1 to the K6/F0-1 bikes.
I’m very interested in this argument, you are saying that in the green Honda manual the float height target is 26mm for both the floats. Do you have a reference for the earlier manual? Does someone has the genuine honda factory tool used for setting the float height and can measure if the two sides are different in height between each other?
The 'catch' about the earlier bikes (and original Honda manuals) is/was: the sandcast, K0 and Old Factory K1 carbs (at the least) have tiny slivers of a casting next to the notch of the float bowl's gasket surface. This little sliver is the zero reference seat for those early carbs, and was milled off quite precisely for each carb body. This feature vanished in the 657b and later carbs entirely (unless yours have the brass floats, like in the F0 bikes)) in favor of using the float bowl gasket surface as the zero reference.
In the mix of years and carbs: I have seen this little sliver in K0 carbs be 1mm higher on the "high" side, and I have also seen both of them being the same height. When they were staggered 1mm, that became our early 'clue' to what was going on, and about that same time Honda was issuing Service Memos (not the same as Service Bulletins for general public consumption) explaining this black magic. (There were a bunch of those for the sandcast bikes!) At that time, the 'problem' being solved was the dribbles under the bike's carb hoses from being parked on hot days on the sidestand, and we were busy trying 'higher' float settings, making sure the gas tank's vent wasn't plugged, and sometimes even rebuilding the carbs to stop it (it didn't.). While we accidentally discovered that setting the sandcast and K0 bike's floats to 26mm on both sides, using the float gasket surface as a zero, made for fewer fouled sparkplugs (those bikes ran too rich), it also made a slight flat spot just off idle that required setting the idle speed up to 1100 RPM to relieve. In the end of this saga, we drilled out the emulsifier holes (to as much as 0.039") to resolve the flat spot, but the bikes still piddled from the bowls if the bike was hot, the day was hot, and the fuel petcock was left open, and somewhat less when it was closed.
What happens is: the near-horizontal section of the fuel hose that reaches over to the #3-#4 carb fuel feed increases pressure when the bike is parked hot as the tank and fuel lines get heated by the engine. This causes both the air in the line and the fuel itself to expand to more PSI than the float valves can hold back, and it force-feeds (pumps) fuel to those carbs until it reaches the top of the overflow tubes. Since these tubes are halfway between the floats and about 24mm equivalent fuel-depth, they bleed the fuel out to the overflow when it reaches the same height as the 'low' side's gasket surface, hoping to not soak the low-side of the gasket in fuel while parked. It often does anyway...while the fuel tank's pressure relief valve is hopefully set below the pressure needed to push fuel past the tank's petcock seal, our modern gasolines soften and swell the petcock seal when wet, then make it hard and shrunken when dry, and it doesn't seal as well as they once did in 1970. So, today we can see it dribble when the bike is parked after a hot run, even with the petcock closed, so Honda (or Keihin) decided to stiffen up the float valves' spring to attempt to seal the valve a little tighter nowadays. While that actually worked, it introduced the bug we all now fight with float bowl fuel heights: the old softer springs let the fuel run deeper with road bumps and jiggles than the new stiffer valves will, so we end up with a lean spot at 1/8 throttle or so that makes for a small stumble in the throttle response because the float bowls don't fill as deep while riding.
I, for one, don't like that stumble, especially in city traffic (it is worse in the pre-657b series carbs) so I readjusted the emulsifier tubes' holes, air screw setting, ignition timing (delayed the advance start) and staggered the floats so as to make the bowls stay generally deeper. It works, but not without a bit of fussing to get it right. Throw in some seat wear for the idle-adjust screw, and the factory settings become merely a historical starting reference.
...and, yep, mine still piddles when parked in the sun with a hot engine and the fuel valve off, and can even leak in the garage with a cold engine and an open petcock when the barometer jumps as fast as it sometimes can here in the Mile High area. We can move 3 barometric points in as little as 15 minutes in Spring and Fall, enough to make one's ears pop! It's usually followed by high winds, like 100+ MPH...and gasoline smells from the garage...
