[Thomas]

Excellent, Paul, that helps a lot. As soon as my worn keys are out (probably next week) I will measure the keyways (and the keys, if possible) and will give you the sizes.
Cheers, Thomas

[phutton]

They also sell small quantities via e-bay: https://www.ebay.co.uk/str/tasmanindustriesltd

P.

[phutton]

Hi Thomas,
Try this company: https://www.keysandpins.com/shop. I would like to stock keys in our Shop, so would be very interested in the sizes you need.

Paul

[Thomas]

Yes, Rick, it is a taper fit. But if the keys would be only for positioning the arm then one per shaft would be fine. So, because the old engineers knew what they did, two keys must take at least some load. This seems to be confirmed by their worn surfaces. I mean, many thousand hits by the pushrods will have a long-term effect, in any case. Unfortunately, I couldn't find respective information in the books, so, I have no better explanation.

Round ended parallel keys of 1/8 inch are indeed difficult to find. It seems that even in the UK they are all metric today. Of course, McMaster-Carr in the US offer them but charge you for shipping like crazy. But that isn't my biggest concern. I can buy 4mm width and machine them down to 3.175. Thank you for the material number. I will try to get that.

Cheers, Thomas

[Rick Parkington]

Hi Thomas,
That is interesting. As you say you would expect the keys only to locate the rocker not take the load - is it not a taper fit? Sunbeam used parallel keys to drive the clutch on earlier bikes, instead of a spline or taper - so maybe they liked the idea enough to use it on the rockers too!
I was going to say with the woodruff keys you will probably need to make them and I guess the same applies with the flat keys. I  would make them from a tough (but not hardened) steel such as EN16T (this is the old name - not sure what your equivalent would be) but this is a tough heat treated steel that is used for things like wheel axles and girder fork spindles.
Cheers Rick

[Thomas]

I was blinded and, hence, wrong! From the beginning I was sure to deal with Woodruff keys (don't remember why). But that are parallel keys with round ends. Such key is relatively easy to get out because I don't have to worry about damaging the half-moon sides of the keyway. I will go to the workshop again tomorrow to solve this. It will be easy to do with a Dremel. After that I'll get the right keys.

Cheers, Thomas

[Thomas]

Hi all,

I went to three local workshops to get the damaged Woodruff keys out. Even with brute force via a Dremel to take away the key interior out it is impossible. Like welded... One of the workshops discovered that the shaft is slightly worn, too. I now see only two options:

Either we go with a milling head from the thread towards the positions of the Woodruff keys and replace them with rectangular keys or I let them make completely new shafts including new Woodruff keys. The latter, though, might be impossible because it seems that the shaft and outer arm are one piece.

It seems surprising that the inner rocker arm can slightly rotate on the shaft when the nut is fastened. I would expect that the keys only define the position and do not take any load. On the other hand, there are two keys and it seems that they take load, indeed.

Does anybody have a godd idea about how to proceed?

Cheers, Thomas

[Thomas]

BTW: Is here somebody who knows the size of these Woodruff keys? And perhaps a supplier?

[Thomas]

Tony, your comment about the pushrod length made me thinking about the inner rocker arm position on its shaft. This position has the same effect as the pushrod length. And in contrast I discovered that I could slightly move the arm on its shaft. The locknut was not reasonable fastened! After fixing that I discovered that the arm still could be moved by hand. The reason can be seen in the attached pictures. The two Woodruff keys are worn. That's a surprise because I would have expected that my engineer who assembled it takes care for that before assembling. I unsuccessfully tried to remove them, so, I will go to a workshop close by and ask them to get them off. Then need to measure the size for new keys.

I learn from this discussion that final solutions are not the most important but the dialogue which might contain important hints.

[Thomas]

Yes, Tony, I agree. But where can I get them? Or how can I adjust their lengths?

My idea is not completely final. Shorter head bolts can be compensated for by respective washers. But shortening the two fins are final, of course.

But your comment brings me to another idea: The missing millimeters at the valve side could be compensated for by slightly rotating the inner rocker arm towards the push rod. Who knows, maybe the workshop did not correctly placed the inner arm onto the shaft, although there is a respective wedge...

[meddlesomebilly]

Hi Thomas,
Surely it would be simpler [and less final] to adjust the pushrod length.
Regards,
Tony.

[Thomas]

Hi Rick,

even moralic support and the knowledge that we all together try to find common solutions for our Sunbeam problems is encouraging. So, thank you very much for your input.

Valve bounce – I do not hear that in my motor. The valve clearance is as it should be (4 and 6thou, inlet and exhaust). In addition, our fellow Klaudius has given me good new valve springs (about 10% stronger than the old ones). However, the tappet adjuster situation sounds very similar. As you can see from my pictures, I have no real trouble with the square and the locknut but only with the geometry. The rest sounds similar to my problem, indeed. And I checked everything in the same manner as you did. It is puzzling!

Even bigger valve caps can't change my situation. The simple question is: How can I get the geometry right?

As I said, I am close to shorten the head bolts by 2.5mm and grind off 3mm from the two problematic fins. Not really satisfying but I see no other possibility.

Cheers, Thomas

[Rick Parkington]

Hi Thomas,
I'm sorry this is not a helpful reply but I thought it was maybe worth saying I had a problem like this with my then partner's 1937 250cc Model 14 Series 2. When we bought the bike it was very prone to valve bounce - a nasty clatter at  high revs caused by the valve being slow enough closing that it hits the rocker coming back down again. The valve angle was also very wrong and the tappet adjusters had to be screwed all the way down to get the right clearance.

I cannot remember the whole story now but I remember fitting a new set of valve springs to cur the bounce. Tappet clearance was  always a problem,  on the exhaust particularly - the square only just poking out of the locknut to close up the clearance as much as possible. During a week at the Isle of Man, the tappet gap got bigger and bigger with no more adjustment left and stupidly thinking a loose tappet was better than a tight tappet, I decided to ignore it. With 120 miles to go on the ride home the Sunbeam dropped it's exhaust valve and holed the piston. This taught me that a wide tappet gap prevents the cam lowering the valve onto its seat and allows it to smack shut putting excess tension on the stem.
Putting it back together I was determined to find an explanation for the problem but couldn't. I checked the rocker arms had not slipped on the spindles, bought new valves but the stems were the same length as the old ones and besides, if the stems had been too short the valve spring tension would have been higher and I wouldn't have had bounce from the old springs. I wondered whether the pushrods had been shortened and when I saw a dismantled engine advertised in a magazine rang the seller and apologised that I didn't want to buy his engine - just ask how long the pushrods were! He was kind enough to do so and they were the same length as mine.
I remember suspecting cam and follower wear because the cam followers had been built up with weld but they looked right and I remember noticing that the rocker arms would have hit the top of the rocker box if they had lifted any higher so I don't think the problem was on the lift side.

In the end all I could think to do was make some thicker hardened valve caps and considerably longer tappet adjuster screws which was the end of the problem if not an actual solution. I only ever saw one other model 14 and the valve angle on that looked perfectly normal so as I say, no help at all except to offer moral support and say you're not alone in having a problem, Thomas!
Cheers Rick

[Thomas]

David, I guess you can find such an app in the respective app stores for free.

• The cams: According to my estimations we talk about 2-3mm differences in height for remedy. I would see such wear immediately. But both cams are fine. And, yes, it cannot occur suddenly.
• Oversize heads: No, the dish diameters are identical to the old valves. And it actually doesn't matter. If the valve sinks into the seat and its tip is at a higher position the non-axial position of the rocker arm collet would not alter because of the 'constant' position of the push-rod tip. From my point of view the valves themselves are not the problem.

Overall, either the push-rods are (suddenly) too short or something is basically wrong with the height of the rocker-box. The only thing I changed in the configuration was a 0.15mm gasket at the barrel base. But that is too thin for the problem. The only way I see is shortening the head bolts by about 2.5mm to lower the box and, hence, lower the outer rocker arm side. But I must admit, I am not completely sure about all that. And it seems that this is a unique observation no one can share.

[DavidG]

I agree with your conclusions: That is really strange! (But what a brilliant phone app.). I wonder how other bikes from this year would measure up? Worn cams and/or followers might account for some change over time, but not suddenly; I wonder if the old valves had oversize heads, so the new ones are effectively sinking too far into into 'pocketed' seats?

[Thomas]

Hi all!

I talked to my motor engineer about the rocker arm again. Her agrees that the present situation is not perfect and that additional force is applied to the guide (which is out of a special bronce needing less lubrication). He also mentioned that he had made extra washers to lift the rocker box because the box touched the head and wobbled without them! He didn't tell me and I didn't realised it when assempling the box again. So, I dismantled everything again and could see that two top fins slightly touched the rocker box. The scratches are from tinkering the box into the frame.

I also checked the rocker arm angles for different setups with my smart phone protractor. For that I used 2mm washers on the head bolts to estimate the influence of the box height above the head on the rocker arm angles.

• No washers on the head bolts: 92° for the open valve, 107° for the closed valve with respect to the valve cup (inner rocker arm on the push rod). Hence, about 15° between the two extreme positions (nowasher_open_closed.jpg).
• 2mm washers on the head bolts: 114° for the closed valve (inner rocker arm on the push rod).
• Difference between washer and no-washer positions for closed valve:  107° from 1. and 114° from 2. (washer_nowasher.jpg). Hence, 3.5°/mm.
• Small angles: Linear interpolation possible. Hence: Extreme positions for 15° interval and shortened head bolts

100° and 85° (2mm off)	= +10°,-5°
98.25° and 83.25° (2.5mm off)	= +8,25°,-6,75°
96.5° and 81.5° (3mm off)	= +6.5°,-8.5°

So, if my considerations and especially my rough measurements are correct (error about 2°) I should shorten the bolts by 2.5mm. In addition, I need to grind two fins off by 3mm.

I must admit, though, that all that confuses me. We can assume for good reason that the works had adjusted everything well taking all necessary parameters into account. This includes the right angles and a matching rocker box to the head. Why is it now necessary to make a compromise between the correct angles and the box-head adjustment? Why do I now need to grind the fins? Something is not as it should be. For instance, push rods too short could be an assumption. But why should the suddenly be too short? The valves too long? I have taken the old valves as a draft for GS. Very strange...

Cheers, Thomas

[DavidG]

Well my comments may not apply to your bike - I think they would up to 1936 - but if you spaced the rocker box further away from the head, it should improve the angle of incidence of the rocker on valve. Thinking about it, if you follow my suggestion you'd also have to lengthen the pushrods. Not sure about 1937 rockers - maybe no scope to alter in that way.
The reference to shims is confusing - but on some engines there are shims inside the valve caps, so one can adjust clearances (and angles?) that way, so that may be what he's thinking of. Sorry can't help more. Now running a sidevalver... :-)

[Thomas]

I got a message from GS Valve who made my pair and those for the club:

If they are the valves we do for the club (58889), then these valves have a Stellite 12 hard tip, and will run perfectly OK with the tappet, they are also plasma nitrided so will be fine with the guide. The only issue could be is if I was not told they were removing the shim, the valve assembly could be shorter, which will affect the rocker geometry which could have an impact on some of the parts. The length of the valve is 3.905", not sure if the is the length of the valve, or the length of the valve plus the shim. If I had to guess I would say it is without the shim as the groove is 0.225" from the tip, and with a shim would make it very close. If this is the case, the valve and stem should still be OK, just the rocker geometry would not be as std.

The length he mentions are without the cap (I believe that is what he means with "shims"). My valves have 4.142 inches taken from the old valves. The difference to the club valves is small but significant. Either my technician welded too much onto the head bolts or the valves in the club shop should be preferred.

So, David, I am not sure what you mean with "...put a washer under the rocker box bolts; with my (earlier) bikes, it was to lengthen the pushrods..." but because the top pushrod side is a constant I would rather lower the rocker box to get closer to the best valve position.

Cheers, Thomas

[DavidG]

I've only had experience of earlier (1929-31) Model 9s, but I'd think the same principle applies: you want the rocker to be at right angles to the valve half way through its movement. I had one bike that was way out - pushing the valves outwards - and in less than 1000 miles the guides were shot! Less crucial on your bike, as i think they have a bit of lubrication up there? But still a good principle.

The caps, as I recall, were used on the exhaust valve only, to protect the austenitic steel valve from wear (I may be wrong..) The simple remedy with yours would seem to be to put a washer under the rocker box bolts; with my (earlier) bikes, it was to lengthen the pushrods, by un-soldering one end, putting a fine copper washer under the hardened end, and soldering it back together. The end results have done thousands of miles, so i think it's worthwhile.

[Thomas]

Thank you for your important information, Paul!

Yes, there are caps on top of the stems. Interestingly, they were not even hardened before. My motor workshop technician realized that and then hardened them. Hm... So, somebody must have placed them on top later. There is only one which then confuses me: The domed screw tips at the rocker arms are much broader then the valve stems. Why, when they push slim valve tips? Or do I exaggerate with this consideration?

Cheers, Thomas

[phutton]

It does look wrong, Thomas! Is there a hardened cap between the end of the valve stem and the tappet adjusting screw? If there is, it should not be there - there is no such part included in the spares list. The tappet should bear directly onto the hardened end of the valve stem.

P.

[Thomas]

Is there no one who can say something about this?

[Thomas]

Hi all,

after overhauling the valve guides of my 1937 Model 9 I assembled everything again. I now wonder about the rocker arm positions on the valves. Attached are two pictures I've taken today. They show the arm position for the closed and open exhaust valve. It seems to me that for both valves the rocker is not at 90 degrees relative to the valve stem when the valves are half open but rather when they are fully open (the situation for the inlet valve is slightly better). As a consequence, the force is less perpendicularly applied to the valves. The engineer waived the washers (no shims) on top of the head bolts but welded them to an even length. I don't expect trouble here (I mean, he is the motor expert) but at least I noticed it. Can somebody tell me if that is the normal situation?

Cheers, Thomas