ergo clears by cutting instead of replacing the springs?

[jacobolus]

The only two relevant properties of a spring in a keyswitch like cherry MX or Alps, at least where feel of the switch is involved and assuming that the spring diameter is correct &c., are (1) spring constant, and (2) initial length of the spring. What do I mean by that? Well, springs have a close to linear relationship between displacement and force (Hooke’s Law). This means that wire gauge, number of turns of the spring, material properties, etc. can be summarized by a single number, the spring constant k.

Inside a switch, the the spring is initially compressed until it fits in the available space, so the amount of force the spring exerts is «spring constant» * («initial spring length» - «space inside the switch»). So there are two ways to reduce the amount of force required in the top part of the keypress: either reduce the spring constant, or reduce the spring's initial length.

One of the advantages that clear springs have over other cherry MX springs is that their force curve is steeper over the range of motion of the switch. This is because clear springs have a higher spring constant than red/blue/brown springs, but a shorter initial length.

If we just replace the clear switch with a lighter cherry MX spring [or with an Originative spring e.g.], then we take away that advantage. Additionally, ergo clears made with light MX springs tend to stick, having a bit of trouble springing back up past the tactile bump.

We can fix these problems by cutting the spring shorter instead of swapping it out. If we cut the winds at the very end, the ones with no space between them, we shorten the initial spring length (very slightly) without changing the spring constant. If we keep cutting, into the spread out winds, then we reduce the initial length and also start reducing the spring constant. In practice, this makes the switch lighter at the top, but still with some oomph near the bottom, enough to avoid ever “sticking”, even without lube.

Unfortunately, cutting springs is a kind of fiddly process compared to just swapping them for new ones. To that end, does anyone have an idea about good sources for tiny custom springs? It seems to me that a spring with about the same spring constant as the current clear spring, or maybe slightly less, but initially shorter, would be better for many folks [also maybe for brown/blue switches]. It seems sad that the current spring options are so limited.

[CPTBadAss]

One of the advantages that clear springs have over other cherry MX springs is that their force curve is steeper over the range of motion of the switch. This is because clear springs have a higher spring constant than red/blue/brown springs, but a shorter initial length.

There is a lot in this post we're taking your word for. For example, this quote above. Would you mind backing up some of your statements with links or references?

If we just replace the clear switch with a lighter cherry MX spring [or with an Originative spring e.g.], then we take away that advantage. Additionally, ergo clears made with light MX springs tend to stick, having a bit of trouble springing back up past the tactile bump.

I was cutting springs over the weekend as an experiment. If you cut off the ends with no spacing, the switch is very prone to jamming.

We can fix these problems by cutting the spring shorter instead of swapping it out. If we cut the winds at the very end, the ones with no space between them, we shorten the initial spring length (very slightly) without changing the spring constant. If we keep cutting, into the spread out winds, then we reduce the initial length and also start reducing the spring constant. In practice, this makes the switch lighter at the top, but still with some oomph near the bottom, enough to avoid ever “sticking”, even without lube.

This has been discussed a bit in another thread. Just because you change the spacing, doesn't change K. See this post.

It seems to me that a spring with about the same spring constant as the current clear spring,
or maybe slightly less, but initially shorter, would be better for many folks [also maybe for brown/blue switches].

Have you tried any of the aftermarket springs yet? There's been a lot of different springs made which people have given a lot of opinions on already. This seems to be a blanket statement, with no links proving your statement.

To that end, does anyone have an idea about good sources for tiny custom springs?  It seems sad that the current spring options are so limited.

And there's an IC to make new custom springs. But AFAIK, there's not a ready source of spring manufacturers.

[Lanx]

say no to cutting springs. (they're really tough)

[Grendel]

Some data. Cutting clear springs makes me cringe tho, I've come to appreciate them in blue switches..

[1pq]

I think you need some help, buddy.

[jacobolus]

One of the advantages that clear springs have over other cherry MX springs is that their force curve is steeper over the range of motion of the switch. This is because clear springs have a higher spring constant than red/blue/brown springs, but a shorter initial length.

There is a lot in this post we're taking your word for. For example, this quote above. Would you mind backing up some of your statements with links or references?

Yeah, sorry. This is mostly basic high school physics, but I really should make some diagrams, but that’s a pain in the ass and slow; I’ll try to make some later. [Also, I’m no mechanical engineer, so take what I say with some grain of salt.]

The basic idea though is: springs generally behave like their idealized model. i.e. within some range of motion, the force they exert is roughly proportional to the amount they've been squished. This is called Hooke’s law.

If you open up an MX clear switch and a brown switch, say, and pull the springs out, you’ll notice that the clear spring is stiffer but shorter than the brown spring. The stiffness of the spring can be summarized by one parameter, the “spring constant”, k, for that spring. Being stiffer means that the clear spring has a higher spring constant, which means for every bit that the spring gets compressed, the amount of force it exerts increases more than for the brown spring. Because it is shorter, when placed inside the switch, the clear spring is less compressed in its initial position in the switch. As a result, the initial force that the clear spring exerts is, while more than the brown switch for sure, less than it would be if the two springs had the same length.

I was cutting springs over the weekend as an experiment. If you cut off the ends with no spacing, the switch is very prone to jamming.

Hm, that’s possible. I have only ever tried cutting off one end. Cutting the spring definitely does have the downside that that end of the spring no longer sits entirely flat. [Since the spring ends are flattened with a grinder at the factory.]

This has been discussed a bit in another thread. Just because you change the spacing, doesn't change K.

I think I’m not making myself clear enough, maybe. I’m not suggesting that k will be variable. Only that having larger k on a shorter spring yields a better result across the range of motion of the switch.

Have you tried any of the aftermarket springs yet? There's been a lot of different springs made which people have given a lot of opinions on already.

Do you have good links to where to find them? All I’ve heard of are the Originative ones, which don’t seem particularly interesting to me.

AFAIK, there's not a ready source of spring manufacturers.

I’m guessing there has to be somewhere in the SF Bay Area to source some custom springs. I’ll try to ask around a bit.

[jacobolus]

say no to cutting springs. (they're really tough)

Hooray for quality wire cutters, I guess.

[CPTBadAss]

Yeah, sorry. This is mostly basic high school physics, but I really should make some diagrams, but that’s a pain in the ass and slow; I’ll try to make some later. [Also, I’m no mechanical engineer, so take what I say with some grain of salt.]

The basic idea though is: springs generally behave like their idealized model. i.e. within some range of motion, the force they exert is roughly proportional to the amount they've been squished. This is called Hooke’s law.

If you open up an MX clear switch and a brown switch, say, and pull the springs out, you’ll notice that the clear spring is stiffer but shorter than the brown spring. The stiffness of the spring can be summarized by one parameter, the “spring constant”, k, for that spring. Being stiffer means that the clear spring has a higher spring constant, which means for every bit that the spring gets compressed, the amount of force it exerts increases more than for the brown spring. Because it is shorter, when placed inside the switch, the clear spring is less compressed in its initial position in the switch. As a result, the initial force that the clear spring exerts is, while more than the brown switch for sure, less than it would be if the two springs had the same length.

Not everyone took high school physics so linking to say another Hooke's Law article helps. That's all I meant. I agree with what you're saying above.

Hm, that’s possible. I have only ever tried cutting off one end. Cutting the spring definitely does have the downside that that end of the spring no longer sits entirely flat. [Since the spring ends are flattened with a grinder at the factory.]

Hm, are you sure about this? This post shows a typical CNC spring machine, no grinders.

Have you tried any of the aftermarket springs yet? There's been a lot of different springs made which people have given a lot of opinions on already.

Do you have good links to where to find them? All I’ve heard of are the Originative ones, which don’t seem particularly interesting to me.

Originativeco.com and kbdmod.com are what come to mind. I'm not exactly sure what you're not interested in. Care to explain? They're a great starting point for what you're proposing. Some good reading is in this thread.

AFAIK, there's not a ready source of spring manufacturers.

I’m guessing there has to be somewhere in the SF Bay Area to source some custom springs. I’ll try to ask around a bit.

I think the issue is trying to find a place that will do relatively low MOQ. But I'm not 100% sure.


In general, I doubt very much that simply cutting the springs will help make ergo-clears. Why? Because even if the K value is improved as you suggest, the spring needs to return the stem to the proper location in the housing. There's a reason why that X distance is there; it's so the switches have a proper travel distance inside the casing. Otherwise you'd get a switch that only returns part way up.

Just FYI, I am a mechanical engineer and I've tried this in keyboards. So if you could prove me wrong or show me more clearly what you're thinking that would be great. Not trying to shut you down; more trying to get you to explain exactly what you're thinking.

[jacobolus]

Hm, are you sure about this? This post shows a typical CNC spring machine, no grinders.

Okay, that could be. I think I’ve seen some keyswitch springs that were ground flat, I’ll go look more closely at the cherry springs specifically. Either way, I can see how cutting the tightly coiled ends off could make the spring sit less flat. I’m just agreeing with you that cutting the spring all the way past that section could lead to problems w/ the spring pushing at an angle or something.

Originativeco.com and kbdmod.com are what come to mind. I'm not exactly sure what you're not interested in. Care to explain?

As far as I can tell, these are about the same initial length as the stock springs (and longer than the clear springs). As a result, a similar weight spring from Originative or wherever might differ marginally from a stock spring, but won’t have an especially noticeable difference in force curve through the travel of the spring. i.e. might have slightly different weight and/or slightly better consistency, but that’s not really what I’m looking for.

I’m actually more interested in playing with Alps springs than Cherry MX springs, but the latter seem to be more popular so I figured might start a better conversation. Anyway though, what I want to see is less force near the beginning of the key press, and more force near the end. One way to accomplish this is with a slightly stiffer spring cut slightly shorter.

To that effect, I think a cherry heavy spring or clear spring, cut slightly shorter, might yield a better feeling “ergo clear” switch than using a cherry light spring, or one of the Korean springs.

I think the issue is trying to find a place that will do relatively low MOQ. But I'm not 100% sure.

For sure.

In general, I doubt very much that simply cutting the springs will help make ergo-clears. Why? Because even if the K value is improved as you suggest, the spring needs to return the stem to the proper location in the housing.

Well of course. But note, we can get noticeable differences with very subtle changes. For instance, just cutting 1mm off the length of the spring makes a substantial difference in its weight, especially obvious near the top of a keypress.

For an MX clear switch, all you really need is for there to be sufficient force upward to reliably overcome the tactile bump in the plastic slider.

[gameaholic]

I knew someone who did this to their rice rocket.  It made the ride rougher but it looked better when it was parked.