Rivets versus hooks

[Richard Mellish]

At least two current threads make some reference to the alternative kinds of action, but it seems worth having a dedicated thread. I would like to understand why rivets are generally reckoned to be better.

 

In engineering generally, rivets are normally made tight, but that would be no good in a concertina because the levers need to move. But the rivets can't be too loose either, because then the levers would flop about. So I assume that the rivets must be carefully made just loose enough.

 

A hook action, on the other hand, would seem to be self-adjusting. I have in mind a comparison with the construction of a laboratory balance, which has the beam supported on a knife edge, whereas cheap scales might have some sort of axle.

 

Who can enlighten me?

[alex_holden]

With hook type actions, particularly on the shortest levers, they have a tendency to lift off the pivot point when you press the button quickly, leading to a slight delay in lifting the pad and a clicking sound as the lever comes back into contact with the pivot.

 

They also aren't as good at dealing with the side forces you get with levers that are cranked to go around an obstruction; in that case they tend to twist a bit when you press the button before they start to pivot, and the pad may not land flat on the board again when you release it. Sometimes notches wear in the sides of the lever over time, causing them to stick and not close reliably (Dave Elliott calls this "cranked lever syndrome").

 

Makers have come up with several different solutions to the problem of making a riveted action 'just tight enough'. The way I do it is to turn an accurate step in the shaft of the rivet so it can't go too far into the post and cause it to bind.

Edited October 14, 2020 by alex_holden

[Tiposx]

In my Meccano- taught brain I think that on the hook action the spring has two functions I.e. holding the lever against the pivot, and also keeping the pallet closed. In the riveted construction the spring only has to control the pallet, and can therefore be weaker if desired.

Further to this I notice that the hook action assembly is pretty much free to waggle around sideways when the end plate is off, and needs the button hole/ bushing to be in place to work properly. I have never taken the end plate of a rivet action instrument, but I imagine that it would waggle around less if the rivet is a good fit.
This leads me to think that there Is more bearing/guide surface, and therefore friction built in to the hook action, otherwise the pallet would tend to move around in use, and therefore be prone to air leaks.

It will be interesting to hear what other people think.

 

[alex_holden]

51 minutes ago, Tiposx said:

In my Meccano- taught brain I think that on the hook action the spring has two functions I.e. holding the lever against the pivot, and also keeping the pallet closed. In the riveted construction the spring only has to control the pallet, and can therefore be weaker if desired.

 

Yes, with a hook action it's sometimes necessary to use a very strong spring to get the shortest levers to work acceptably. It's less of an issue with longer levers.

[Richard Mellish]

Thanks for the answers. If I have further thoughts I'll come back.

[Tiposx]

I hadn’t thought about the need for stronger springs on short levers. That explains quite a lot of mystery - thanks Alex.

[Chris Ghent]

Hook action was driven by ease of assembly and economics rather than quality. 

[David Barnert]

My Dickinson/Wheatstone 46-key Hayden from the mid-1980s (SN: 60082) has a hook action. I know of at least one similar instrument from the same period that is riveted. Any idea why Dickinson would have done that?

[alex_holden]

24 minutes ago, David Barnert said:

My Dickinson/Wheatstone 46-key Hayden from the mid-1980s (SN: 60082) has a hook action. I know of at least one similar instrument from the same period that is riveted. Any idea why Dickinson would have done that?

 

I can't remember where I originally read this (most likely somewhere on this forum), but I seem to recall it had something to do with a batch of ten built by a new apprentice who knew how to make the hook actions but hadn't been taught how to make riveted levers yet.

[Don Taylor]

I have a 1950's Crabb with a hook (more specifically a 'crank') action.  I asked Geoffrey Crabb about the advisability of replacing it with a riveted action.  This is what he said:

 

The existing action design which, incidentally is similar to that used by Wheatstone and introduced both by them and my father in the late 1930's is quite good and far superior to the earlier Lachenal type.  Usually referred to as 'crank action' it's use enabled actions to be installed quickly by unskilled persons employed casually during periods of high demand.

 

 

Geoffrey advised me not to replace the action.

[David Barnert]

3 hours ago, alex_holden said:

I can't remember where I originally read this (most likely somewhere on this forum), but I seem to recall it had something to do with a batch of ten built by a new apprentice who knew how to make the hook actions but hadn't been taught how to make riveted levers yet.

 

I’m surprised to hear that Dickinson ever had an apprentice. I just searched this site for the word “apprentice” and found many examples, but only one in a thread that mentioned Dickinson, and while not providing any definitive information, seems to suggest that he never worked with apprentices.

 

1 hour ago, Don Taylor said:

I have a 1950's Crabb with a hook (more specifically a 'crank') action.

 

Can you elaborate on what the difference is between a hook action and a crank action is? Your picture looks a lot like what I’ve got (the action, not the buttons or pads).

[alex_holden]

2 minutes ago, David Barnert said:

I’m surprised to hear that Dickinson ever had an apprentice. I just searched this site for the word “apprentice” and found many examples, but only one in a thread that mentioned Dickinson, and while not providing any definitive information, seems to suggest that he never worked with apprentices.

 

There's a mention of his apprentice here, though this isn't where I remember reading the detail about hook action:

 

[David Barnert]

9 minutes ago, alex_holden said:

There's a mention of his apprentice here, though this isn't where I remember reading the detail about hook action:

 

Sure, enough, there it is (quote from “Inventor” -- Brian Hayden):

 

Quote

Steve Dickenson had recently taken over Wheatstones and with a new apprentice was looking for extra work. - I ordered ten instruments.

 

I don’t know how my search missed it. Thanks.

Edited October 15, 2020 by David Barnert

[Don Taylor]

24 minutes ago, David Barnert said:

Can you elaborate on what the difference is between a hook action and a crank action is? Your picture looks a lot like what I’ve got (the action, not the buttons or pads).

David 

 

Those were Geoff Crabb's words, not mine, but he also says that Wheatstones used the same 'crank' action from 1930 onwards so that would explain why your action looks similar to the Crabb crank action.

 

I do not have a picture of a Lachenal hook action to compare to the crank action.

[Tiposx]

Those saddles would/could support the levers in the sideways direction. This would avoid the waggle I mentioned earlier. It would have to be an accurate fit to avoid waggle on one hand, or excess friction on the other.

[Don Taylor]

2 hours ago, Tiposx said:

Those saddles would/could support the levers in the sideways direction. This would avoid the waggle I mentioned earlier. It would have to be an accurate fit to avoid waggle on one hand, or excess friction on the other.

It is fairly easy to adjust the fit by prying them open a bit with a screwdriver or closing them with a pair of pliers.

[Tiposx]

That makes perfect sense. Looking at the photo I wonder what the bearing surfaces look like. If the post is a simple fold then it would have an elongated bearing surface. I can't see what the shape of the bearing surface on the lever is. If it is V shaped lengthways then it would have 2 points of contact and a vague location through its travel. Perhaps the lever is curved at the bearing point? Sorry for the ramble. I once owned an early Rock Chidley that had a version of the design, but it was over-engineered and impossible to adjust. It didn't have a smooth action either.

[Richard Mellish]

Having the hook with flat bearing surfaces that are (or can be adjusted to be) just a tad further apart than the thickness of the lever would seem to restrict wobble, so in that respect to be as good as a rivet and easier to make and adjust; but it would not avoid the problem mentioned by Alex (Wednesday at 10:52 AM).

 

You could avoid that by having something like the hook with flat faces but facing upwards, with the spring on the pad section of the lever pulling down, but that would be trickier to manufacture, particularly anchoring the spring to the action plate.

 

I have some sympathy for Steve Dickinson's philosophy of sticking to the design features from Wheatstone's classic period, after the various improvements and before the cost-cutting changes. On the other hand I also have some sympathy for other modern makers' changes for improving performance and/or facilitating manufacture.