You guys have already provided a LOT of help (and I've provided a lot of pictures) in THIS thread about a modern Mikuni replacement for a '75 DT250:
yamaha-enduros.com/index.php/forums/74-7...alternative?start=20

Long story short, I tried going that route, but discovered the '75 carb has different dimensions from the '74 that I have, so I'm back in my rut of trying to solve my float issue. I have made some interesting discoveries, though, and I'm hoping that the clues I've discovered helps trigger some little "gotchas" that this group has seen before.

As I show in the post above, my carb is "like new"... bead blasted, ultrasonic cleaning (many, many times over), new brass (some genuine Mikuni, other not.. the float needle is NOT Mikuni).

Here's the basic problem... the carb either refuses to fill with gas OR it overflows. It's like the needle is either stuck on or stuck off.
photos.app.goo.gl/rknZV9gkrs4VXkh79

That's really the bottom line. I spent 4 hours yesterday trying to change every single variable I can imagine, to no avail. (I would guess that I took the float bowl off 30 times just yesterday.) Here's what I can unquestionably confirm:
1. The needle itself is smooth and shuts off fuel flow 100% when I lift the float arms by hand. It takes very little pressure to do this. As soon as the needle "seats", fuel flow stops... not one single drop leaks past. Conclusion: The needle and seat is GOOD. (I have the carb hooked to the fuel line with the float bowl off to do this.)
2. The floats FLOAT and do not bind (at least when the carb bowl is off. They don't have pinholes that cause them to sink. Short video of the floats: photos.app.goo.gl/C6Z6oibQQvWsKyEm7

Those are the only two things that I'm 100% certain on at this point. The floats float and the needle has the ability to full shut off fuel flow. Here's some of my trial and error experiences:
1. Float height set to 17.2mm (factory spec). Result: OVERFLOW. (SANITY check here, please I should be measuring the TIPS of the U shaped brass float arm to the base of the carb, correct?)
2. Spec allows for +/- 2 or 2.5mm... since it leaked badly at 17.2mm, I decided to increase this to 19.5mm (which means the floats don't need to be as high to shut off fuel flow). Result: OVERLOW.
3. For grins, I continued to raise the float height in 0.5mm increments. (Is my terminology right here? is going from a measured 17.2mm gap to a 19.5mm gap "raising" the float level? Because in reality, the floats should shut off fuel at a LOWER height... even if my terminology is backwards here, I hope you can follow along.) Result: at a whopping 23mm gap, I was able to stop the overflow condition! The problem? I got NO fuel. It's like the needle was closed the whole time.
4. Back to a 22.5mm gap. Result:OVERFLOW
5. Back to 23mm gap. Result: FLEETING success. It filled with fuel and didn't overflow. I could tilt the carb to the side and it would overflow. Came back to level and it shut off. Secured it back to the rubber cylinder boot and turned the gas on, and it overflowed. :( This was the one and only time (out of ~30) that the carb even TRIED to act like it would function.
6. I flipped the brass float arm (the U-shaped arm shown in the attached picture) upside-down, thinking that perhaps I installed it wrong at some point. Largely, the same results as above held true. QUESTION: Which side is "up" on this brass arm in my picture? What I THINK is correct is that the side that is showing in my picture faces the sky when it's installed. The "tab" shown on the lower right of the arm, i believe, is to keep the arm from going down too far when the carb is empty. I theorize this because when this is upside-down, the arm can swing down so far, the needle falls all the way out.)



I'm mainly left scratching my head (down the skull, at this point!). My best guess is that when the float bowl is put on, SOMETHING is binding. Which makes me wonder if the floats and the brass float arm are even the correct parts for this carb. (this bike has had some other frankenstein parts on it for sure!).

Any tips? Anyone with a working '74 DT250 carb that they would sell and/or take this one on trade? I'm out ideas, patience, and even hope at this point. This is a BEAUTIFUL DT250 that I've spent years getting up to snuff... 12v PowerDynamo system and everything, and this stupid, mechanically simple issue has plagued me for far too long!

Thanks in advance, guys!

The thing that sticks out the most is the above statement... could be your whole problem? (Mikuni float valves are somewhat universal parts.... can be found at place like Niche Cycle Supply pretty cheap.)

Other than that, the question that come to mind is what is that relationship of the float arm to the carb body with the arm just closing the valve? Parallel to the float bowl gasket surface? Angled up? Angled down?

Well, that all depends on how far I've bent the tab that makes contact with the needle.
At spec (17.2mm), it's parallel to the carb body, yet if you assemble it this way, the carb will overflow.
When I bend the tab enough to get it to stop overflowing, it angles DOWN, like this picture (although obviously nowhere as exaggerated as this photo):


I looked at Niche cycle, and they have float needle assemblies for the '75, but not the '74. This '74 was an oddball carb, perhaps. I have a DT400, and the carb from it has a completely different float needle assembly... different threads and all.

Parallel is correct...

It looks like the floats are sort of an oval shape? I'm going to have to do some looking but i seem to remember the early carbs that that the pin-mounted floats were perfectly round. Wondering if you have the wrong floats and they hit the carb body before shutting off?

That's my best guess... as seeing it now assembled it looks like needle.seat is probably okay.

Seeing your photo of the carb , I see the float arm is installed upside down .
Flip that over , and as Mark previously posted adjust the tang so that the arm
is parallel to the bowl gasket surface . That should put you close to factory
float height . Patience Grasshopper !!!

I think it's just off-angle perspective of the floats. They are perfectly round, but from my camera angle (slightly to the right) you see the height of the cylinder which makes them look oval in that pic.

In going back through my service manual, I MIGHT have found a clue. The service manual shows a washer on the threaded float needle assembly. But as I look at my assembly (also shown in the picture in my first post), I have TWO washers. One is thin metal (I believe that's the "right" one). The other is about a 1mm red fiber washer.



So if there's 1mm (or heck, even 0.5mm) extra thickness in the needle assembly, that means the "head" of the float needle is sitting 0.5mm - 1mm lower than normal. Extrapolate that 1mm down the length of the brass float arms (maybe 50mm??) and that could be a good 5mm difference at the tips of the arms (where you make the measurement.

I have something to try this evening!

I should have specifically called that out in the photo. Yes, I know that's upside down. I mentioned in the original post that I flipped it upside down JUST IN CASE. It's not that way now, nor was it that way when I've been messing with it most of this time.

Patience... PATIENCE??? Is that a joke? This has been my final issue now for a YEAR. It SOUNDS like hyperbole to say that I've had it apart 100+ times. It is not. I will rack my brain trying to figure it out until I finally give up in exasperation for a month or two. Again, I've gone as far as ordering a new OEM Mikuni, but was unfortunately a bad fit. I've had the patience of Job on this. But I'm now at a point where I need to sell it (within the next 2 weeks), and I simply can't sell it with this nagging issue. Everything else on it is nearly perfect. I can't let a mechanically simple thing like a float make it a "project bike". I'll get it, though. I know I will. I appreciate your input!

How far is the tang bent when the float arm is installed the correct way? The second washer is odd too, I'd certainly take it out. Is the overflow from the carb bowl seal or the overflow orifice?

Also odd it's a metal washer on the banjo bolt which holds the main jet, I'd expect it to be a fiber washer or copper at worst, but that looks like too hard of a metal to seal well without a lot of torque.

One other thought, have you left the floats sitting in liquid for an extended time to see if it's a slow leak?

The Float valve should only have one washer .
Originally they were probably fibre , but more recently are supplied with an
aluminium washer .You only need one , either will do the job .
You do talk of taking the measurement at the arm tips .
I believe you should measure where the float pins would touch the arm .
However if the arm was set parallel , the measurement should be the same .

If you decide to replace your float valve . You said Niche cycle had one listed for
a 75 . I've checked my parts books , but don't have a book for a DT 250 A.
A DT3 takes a 2.0 valve and DT 250 B takes a 2.5 . They are identical other
than apperture size . Partzilla might tell you which size yours will require ,
I'd suspect it's the 2.5 .