Summary:
I've used a quadruple real-time clear tube float test shown in the attached picture to hopefully finally diagnose the engine problem that's been dogging me for two years now. I believe I need to replace (at a minimum) two float needles on a set of BS34's and I am wondering what everyone's recommendation is. Should I replace as a solid-tipped needle/seat set using OEM Mikuni parts at $37 per, or try just the aftermarket Viton-tipped needles that are out there? Or some other variation? The various toothpaste/polishing tricks didn't work well enough for me to eliminate the slow leaks (yes, I am already using a good quality filter upstream).
The full story:
I picked up this bike (the GPz 750 in my signature) a couple of years ago and as part of getting it back on the road I had cleaned and serviced the carburetors (replacing rubber components including diaphragms, Viton O-rings, etc.) while keeping the original Mikuni parts (or replacing with genuine Mikuni parts as necessary). At the end of that process I set the float levels using the time-honored clear tube method both on the bench and verified on the bike. The bike ran well but I mostly took short hops as I was repairing individual non-engine issues in the summer of 2016. Eventually (summer of 2017) I got ready for longer rides and discovered -- pretty consistently -- that after running at roughly 1/4-1/2 throttle for several (usually 5+) minutes, two cylinders seemed to cut out. After sitting it out for maybe 15-30 minutes I'd be able to re-start with all four cylinders but it usually wouldn't be until the next day that it would really run properly again. Since it seemed like a textbook thermal/electrical problem I worked my way through the usual suspects (maybe fouled plugs, maybe an intermittent coil, maybe an intermittent pickup, maybe a bad igniter unit, maybe bad plug wires, maybe bad ignition wiring harness, maybe a bad resistor in the plug caps, maybe low battery charging voltage, etc.) but to no avail. I also worked through all sorts of venting scenarios (clogged cap vent, vapor lock, too much pressure drop across the aftermarket fuel filter, faulty design of the aftermarket fuel tees, etc). Once I convinced myself that none of those things were at fault my thoughts eventually circled back to the carburetors. I had done a static clear tube test, but I didn't really know what was going on in the bowls when the problem occurred, hence the home-made getup shown in the picture. I could ride with the tops of the glass capillary tubing uncapped and once the problem occurred I could verify that the fuel levels were still correct on all four cylinders without first having to limp back home, get the tools out, etc.
Well, lo and behold on my first ride it acted up right on schedule and the fuel levels on #3 and #4 were way low (as shown in the picture). They had started out just fine at the beginning of the ride; right where I had set them via the clear tube method two years ago. Hmmm.
So my current thinking is that when I originally set up the float levels back in 2016 I didn't really pay attention to the speed at which I was performing it. I think I must have been waiting too long for the fuel levels to stabilize and adjusted the floats too low; by waiting too long for the levels to settle I had actually set them *based on the slow leaks*, rather than the "normal" filling rate. So - to fix this I need to re-set the float levels but I need to do it with needles & seats that don't have any leaks.
Oh, and In case anyone is concerned, I diligently use a Pingle tank petcock rather than the OEM vacuum petcock so no major worries about hydraulic locking the cylinders due to these leaks.