I don't enjoy the thought of repeatedly disassembling my IBM cases in order to get the tension correct. Has anybody tried using spring disc washers in order to provide a wider range of screw turns setting within the proper tension range? Sprung washers stacks are used in tuning some centrifugal clutch assemblies for small engines to set engagement rpm and they work fine in that much more difficult environment.
Or has anybody used plastic or rubber washers? It seems to me that any of these three types of washer would be superior to a completely rigid setup. Also I worry if I drop or otherwise stress the board there is zero compliance after the mod is done. I would feel better if there were some way to place at the least a small amount of compliance in this retention system.
Sprung washer stack for bolt or screw mod?
-
- Location: Serbia
- Main keyboard: IBM Model M
- Main mouse: Logitech M575
- Favorite switch: Matias Quiet Click
I probably won't provide you any useful advice but i'll let myself jump in with some basic consideration from an engineering POV.
So, as you said, what we want is "a wider range of screw turns setting within the proper tension range". To achieve that, we would need some kind of constant force spring (first diagram), which we won't find in a form of washer. For any variation in deflection (i.e. tightening angle) we will get a variation in tightening force. How much wil it vary depends on how steep the F-delta curve of our spring is.
Belleville spring washers are designed for high loads and small deflections, so the F-delta curve will be steep and nut turning range pretty narrow.
Crescent-shaped washers should have a linear F-delta diagram and they should be somewhat softer. They might be a usable solution.
I would exclude plastic and rubber ones right away. A rubber washer (or some kind of o-ring), when tightened, will have a residual stress in a counter-clockwise direction which could loosen the nut/bolt. Plastic ones (e.g. nylon) won't make much of a difference.
Now, we don't get perfect results from tightening by feel either, so all this talk might be pointless.
There is also another possible problem - are these spring washers even available in M2/M2.5 sizes and how easy they are to obtain.
To wrap this up, I would probably use an ordinary split lock washer to prevent the bolt from loosening and still tighten by feel. To do this, you need a 10mm bolt (8mm is too short), and it should be oriented so the nut is on the barrel plate (if you don't want to grind off ends of the bolts which would stick out on the back otherwise).
I've bolt modded a Model M once, and since then I only do a necessary partial repairs for a good reason - the whole process is insanely time consuming.
So, as you said, what we want is "a wider range of screw turns setting within the proper tension range". To achieve that, we would need some kind of constant force spring (first diagram), which we won't find in a form of washer. For any variation in deflection (i.e. tightening angle) we will get a variation in tightening force. How much wil it vary depends on how steep the F-delta curve of our spring is.
Belleville spring washers are designed for high loads and small deflections, so the F-delta curve will be steep and nut turning range pretty narrow.
Crescent-shaped washers should have a linear F-delta diagram and they should be somewhat softer. They might be a usable solution.
I would exclude plastic and rubber ones right away. A rubber washer (or some kind of o-ring), when tightened, will have a residual stress in a counter-clockwise direction which could loosen the nut/bolt. Plastic ones (e.g. nylon) won't make much of a difference.
Now, we don't get perfect results from tightening by feel either, so all this talk might be pointless.
There is also another possible problem - are these spring washers even available in M2/M2.5 sizes and how easy they are to obtain.
To wrap this up, I would probably use an ordinary split lock washer to prevent the bolt from loosening and still tighten by feel. To do this, you need a 10mm bolt (8mm is too short), and it should be oriented so the nut is on the barrel plate (if you don't want to grind off ends of the bolts which would stick out on the back otherwise).
I've bolt modded a Model M once, and since then I only do a necessary partial repairs for a good reason - the whole process is insanely time consuming.
- Attachments
-
- (Very crude) force-deflection diagrams.
- F-delta_diagram.png (40.68 KiB) Viewed 1417 times
- Yasu0
- Location: hawaii
- Main keyboard: dull grey ibm selectric
- Main mouse: vertical ergonomic old man mouse
- Favorite switch: unicomp m, spring and rubber in perfect harmony.
Input appreciated as I do not have an engineering background. I figured a coil type spring would add too much height to the system to be acceptable. I have not yet done this bolt mod, so I just assumed so.
I understand the washer spring form factors do not offer an ideal solution. But, if one selected the best suited progressive rate washer of what is available... I thought it would at the least be superior to a completely rigid solution where a slight turn of a bolt makes a huge difference in tension. Also the compactness and cheapness of the washer type spring is unbeatable. "Do not let the ideal be the enemy of the good." Good in this case meaning an improvement over the status quo of bolt modding.
If I open up an a couple of M's maybe I will try stacks of a few cheap spring washers alternating with flat regular washers and see how it tensions up. Side by side with the standard method it would be interesting to compare.
I understand the washer spring form factors do not offer an ideal solution. But, if one selected the best suited progressive rate washer of what is available... I thought it would at the least be superior to a completely rigid solution where a slight turn of a bolt makes a huge difference in tension. Also the compactness and cheapness of the washer type spring is unbeatable. "Do not let the ideal be the enemy of the good." Good in this case meaning an improvement over the status quo of bolt modding.
If I open up an a couple of M's maybe I will try stacks of a few cheap spring washers alternating with flat regular washers and see how it tensions up. Side by side with the standard method it would be interesting to compare.
-
- Location: Serbia
- Main keyboard: IBM Model M
- Main mouse: Logitech M575
- Favorite switch: Matias Quiet Click
Well, that's the best way to make an improvement.If I open up an a couple of M's maybe I will try stacks of a few cheap spring washers alternating with flat regular washers and see how it tensions up. Side by side with the standard method it would be interesting to compare.
My main concern about bolt modding is will the bolts loosen from vibrations and moving the keyboard around (i.e. when you tweak the tension once will it stay like that permanently). There are various solutions to that: spring washers, lock washers, Loctite threadlocker. How well do they work - the only way to know is to try them out.
As for the tensioning itself, a spring washer of adequate strength will probably make the process easier. The best way to find out what exact kind of washer would work the best is to get several different ones and to try them out.
- anthonymak
- Location: Hong Kong
- Main keyboard: IBM Model F101
- Main mouse: Microsoft arc
- Favorite switch: Buckling Spring
- DT Pro Member: -
I guess screw tension has minimal impact on membrane performance. The internal black rubber layer has already serve as tension buffer unless you ditch the rubber. I have done bolt or screw mod to several model m and none of them has screws falling out even without the rubber layer. I would say screw tension does affect capacitive pcb performance.