1920 Harley Model F

[T. Cotten]

Folks,

I used to try to test valves as you suggest, and some of my attempts are attached.

Just flip it upside down, and suck on it.
If it make a little pop when you pull your tonque off, it should do (long enough to shut the petcock, anyway).

It takes some talent, I admit.

....Cotten

[Steve Swan]

Folks,

I used to try to test valves as you suggest, and some of my attempts are attached.

Just flip it upside down, and suck on it.
If it make a little pop when you pull your tonque off, it should do (long enough to shut the petcock, anyway).

It takes some talent, I admit.

....Cotten

Tom, YOU ARE THE MAN !

[TechNoir]

Re: 1920 "F" Project

Hi folks, today i investigated the float valve in my carb using the advice given to me by Tom Cotten and Steve Swan.

I plugged the centre hole up using a bolt that I turned the head down to the same diameter as the flange on the stem of the carb and sealed it with an O ring. This way I also had a way to safely hold it in my vice.





Here is my needle valve



I didnt have a pin vice suitable to hold the needle so I had to make one. The advantage of having a long rod is that it is easier to keep it vertical.



Before I started i tested the valve to give me a start point. It filled up and overflowed in about a minute.

Then I lapped the needle in using polish, we dont have Silichrome in the UK so I used Solvol Autosol which I am pretty sure is the same stuff but with a different name. After 3 repetitions of a lapping in a similar manner to Steve Swans description it seemed to be holding so I left it for an hour.





As you can see it is much better but not 100% so I gave it one more repetition and it seemed to be OK.



Whilst I had the float off I decided to have another go at something that has been a problem for me so far.

The air valve is stuck solid. I dont have a proper wrench for it so I cut one out of 6mm (1/4") plate. I was then able to put loads of force on it but it is still solid. I put as much force as I dare on it, any more and I am sure something will break. It moved about 10 degrees but tightened up again.



I assume that it is not a left hand thread?

So for now I have left it soaking overnight in atf and acetone.



Is this a common problem? Am i doing something wrong? I assume it just unscrews.

John

[Robert Luland]

John, It just unscrews. Soak it. Sounds like a crud problem. Bob L

[Steve Swan]

John, is it the picture or is the float actually submerged in the gasoline ? i don't believe the float should be submerged like that. Either the float is gas-logged or the float valve is still leaking or both. Someone with more knowledge and experience can hopefully share their opinion.

[TechNoir]

Re: 1920 "F" Project

John, It just unscrews. Soak it. Sounds like a crud problem. Bob L

Thanks Bob. That's what I thought but you know sometimes when something is so immoveably stuck you have to stop and check in case you have missed something that is locking it.

John, is it the picture or is the float actually submerged in the gasoline ? i don't believe the float should be submerged like that. Either the float is gas-logged or the float valve is still leaking or both. Someone with more knowledge and experience can hopefully share their opinion.

Steve, you are correct. In the first picture above of the bowl with gas in it, the top of the float is about level with the top of he gas. In the second picture the gas has leaked past the valve after about an hour to get to the higher level.

After I had lapped the valve the final time the level of gas was a fraction below the top of the float but I have not pictured that.

I have learnt a lot about Schebler floats in the last couple of days after corresponding with yourself and Tom Cotten. A big thanks goes to both you and Tom for taking the time to provide some guidance on this issue.

My original cork float is 4 grams. The replacement Nitrophyl float is 7 grams. Both are similar sizes and shapes so I am going to assume similar volumes so therefore the new float is massively less buoyant then the cork one. Ok if I seal the cork one it will add to its mass but I reckon it will still be much less than 7 grams so I reckon a properly sealed cork float is better than a nitrophy "heavyweight" (see below).

Tom Cotten has stated that "genuine" nitrophyl replacement floats are about 3 grams. Tom describes mine as one of the Asian molded heavyweight knock-offs because there is a current patent on the genuine nitrophyl floats but there are lots of fake ones on the market that weigh lots more.


Here is a comparison of the weights






So that means that my new nitrophy float is only about half as buoyant as it should be. Therefore it is more submerged in gas than it should be but more importantly, by my reckoning, less bouyant means less pressure on the needle valve.

It should be OK for now but i am going to do 2 things. First I am going to re-seal my cork float using an Ethanol resistant sealer. Second I will get a genuine nitrophyl float but that might have to wait for a little while. I need a few other small bits and SWMBO has a business trip to the USA in January so I intend to get some stuff delivered to her hotel and make sure she has room in her bags. In the meantime I am going to see how the heavyweight performs.

John

[T. Cotten]

Just some notes for reference, Folks,

I just stacked six original 3/4"and 1" model HX floats on my scale, and they averaged 2.9g, but they all seem to have lost their collodion coating.

And I believe there were possibly thirteen 'genuine' nitrophyls since it was introduced in the '50s. With modern off-patent productions, the term has become somewhat generic.
Even the super heavy productions on the market smell right, though.

The material I machine has only been available since June of '08, and I replace any of my previous productions free of charge if they swell. The swollen float must be returned to me for verification first (because an 'Armstrong' and a Rubber Ducky have shown up instead).

The mystery is: If the ethanol content of USA P4gas has remained 10% over the years, why have only a small percentage of my 2000 to 2008 productions 'bloated'?

You never know what is going to come out of the next pump.

....Cotten

[TechNoir]

Re: 1920 "F" Project

Quick update and another question or two.

The air valve was very difficult to get started even after a night of soaking but once it moved a bit then gentle coaxing back and forward had it off with threads intact.



I can now see the valve. In the original build report it just says that a new valve has been made but no mention of material.

This is made of 1 piece of aluminium. Am I right in thinking it should be a sandwich of 2 pieces of metal (aluminium or steel?) with leather, peek or teflon in the middle. I.e. in my picture the bit that is machined at an angle should be the bit that does the sealing.






My bike has been hard to start so far, would this be the reason? I have been searching this forum for Schebler info and this seems to be something that has been commented on before.

I am back off to put it all together now and see how it runs now that the float is doing its job.

John

[TechNoir]

I also meant to ask what thickness should the leather/peek/teflon be? I assume the 2 metal washers are not a critical thickness.

John

[T. Cotten]

John!

What's holding the airvalve open?
Aluminum might work, but does the carrier that rides upon the airvalve shaft have a bushing?
The shaft looks different, too.

Not all had leathers (some early ones were just a brass disc), but they appear to have been 3/16" before compressed fore and aft by washers.

Tensioning the airvalve spring corrects for variances. Following Tommo's advice, I "weigh" the tension as shown in the attachment.
Most original springs read either an even 3 oz or 5 oz, with the adjusting screw in the middle of its travel.
This assumes, of course, that the mass of the carrier assembly resembles the original.

...Cotten

[TechNoir]

Re: 1920 "F" Project

John!

What's holding the airvalve open?
Aluminum might work, but does the carrier that rides upon the airvalve shaft have a bushing?
The shaft looks different, too.

Not all had leathers (some early ones were just a brass disc), but they appear to have been 3/16" before compressed fore and aft by washers.

Tensioning the airvalve spring corrects for variances. Following Tommo's advice, I "weigh" the tension as shown in the attachment.
Most original springs read either an even 3 oz or 5 oz, with the adjusting screw in the middle of its travel.
This assumes, of course, that the mass of the carrier assembly resembles the original.

...Cotten

Cotten,

I am holding it open, see here.



Hare are all of the parts



The air valve is machined from a single piece of aluminium, no bushes





Here is the shaft.



With no reference point I have no idea what parts are as per original and what parts are replacements.

I like your idea about weighing the spring tension. It looks like a simple rig to replicate. I dont have any small weights but I do have lots of scrap and a lathe so I can easily make some.

Do you know what the stock carrier assembly weighs? I assume the carrier assembly is the shaft, spring, split pin, nut and the washers and seal.

I am assuming that a bush is a good idea? If I am going to make new parts then does anyone have any pictures of the correct setup?

John

[T. Cotten]

Well John,

Tommo gets the credit for weighing tensions,...
Because the pedestal has some weight, I filled a couple of vials with shot to total 3 and 5 ounces.

Is the shaft steel?

A fossil carrier I have with the leather gone is ~17g, so ~20g would be a good guess at un-bushed assemblies. No two seem alike, of course.
The variety of springs is unfathomable.

....Cotten

[TechNoir]

Re: 1920 "F" Project

Hi Cotton,

I guess first I must say thanks to Tommo for the spring tension idea. I had read some stuff that talks about spring tension but hadn’t figured out how to do it. Some shot in a small container is perfect for a weight. Thanks.

The shaft is steel.

I checked the original engine build report from 1989 and it says:

"New air balance valve & spindle assembly made up & seating machined to suit"

Which I think is this air valve

It also says under the list of new parts made:

"Carburettor

1 Throttle spindle
1 spindle lock pin
2 spindle butterfly screws
1 spindle butterfly
1 choke valve spring
1 high speed air flap lever
1 high speed air flap screw
1 Needle spindle retaining screw
1 float chamber base nut
1 choke valve
1 choke valve spindle
1 choke valve end nut"

I think it is referred to in the parts list as the choke valve.

I will order some teflon or peek and make a new valve. Then tension the spring using your estimated weight as a starting point and see how it goes. At the moment my shaft, spring and nut weigh 20 grams without the aluminium valve so I will have to adjust as necessary for the complete assembly.

Thanks again,

John

[T. Cotten]

Just for clarity, John,

The weight I gave was for the carrier alone, as the shaft, spring, and nut don't move.

I could cut one-piece carrier/seals like yours entirely from PEEK or whatever, and it would save a lot of time, but my mandate is to stay as close to the original design as practical.
Only a Cannonballer would demand such innovation.

It would please me greatly to know how the aluminum assembly performs, but only after a few thousand miles. I wonder if your difficulty in removing the assembly was due to the aluminum beating the brass outward, binding the threads. Three or five ounces doesn't seem like much, until it is applied pulsing at a couple of thousand RPM.

.....Cotten

[TechNoir]

Re: 1920 "F" Project

I don’t think the difficulty had anything to do with the one piece valve. It was stuck when I got the engine and it had never been run.

I will let you know how I get on with it. I am away on business until the weekend and have fallen out of love with the bike for a couple of days because I seem to have turned a working bike into a non runner when I failed to get it going this afternoon. I will tension the spring properly and see what difference it makes. If no luck I will try an assembly similar to the original.

John