Kailh BOX Switches Crack and Stress Keycap Stems

[Photoelectric]

MX Clears fit fine within caps.  There used to be an issue with earlier Signature Plastics DSA sets that could latch onto an MX Clear stem so tightly that you could pull out the stem from the switch when removing a keycap.  But I think that's been fixed on SP's end for a while.  MX Clears don't crack keycaps.

[Giorgio]

Show Image

****ed my shift key up. Box Blacks.

 :confused: I can't see the crack. Can someone point it out to me?

Has anyone had MX clears crack their stems? My clears are a much, much tighter fit than my box reds.

Look at the measurements.

[Bella_Hwang]

I made a post on Reddit after seeing this thread with some findings regarding some on the issues with the Kailh Box stem sizes. Although it is a low sample size, I am sure more people will start measuring there stems and find similar results.

https://www.reddit.com/r/MechanicalKeyboards/comments/91u9d4/kailh_box_switch_stem_measurements_and_possible/

Hi,I'm Bella from Kailh,actually when we got the message about the keycap crack,we urge the QA to random inspect the box stem of our stock,our datasheet as the post,haven't thicken to 1.40. could i Know when you get the switches and where you buy it?

[Bella_Hwang]

Hi Bella_Hwang,

I've taken some measurements of my Box Royal Hako True switches I got from Novelkeys a few weeks back.

You can find the measurements here: https://www.keebtalk.com/t/kailh-box-switch-stem-measurements-and-possible-problems/2926/16?u=manofinterests

I will be taking more measurements with sample sizes of at least 60 this weekend for the various box switches I have. Some have GMK sets with very small cracks, and some seem fine.

Hi,thanks for your kindly message,we'll notice the issue

[otanishock]

Hi Bella_Hwang,

I've taken some measurements of my Box Royal Hako True switches I got from Novelkeys a few weeks back.

You can find the measurements here: https://www.keebtalk.com/t/kailh-box-switch-stem-measurements-and-possible-problems/2926/16?u=manofinterests

I will be taking more measurements with sample sizes of at least 60 this weekend for the various box switches I have. Some have GMK sets with very small cracks, and some seem fine.

Hi,thanks for your kindly message,we'll notice the issue

If it turns out to be some kinds of QA issues where you release a massive amount of faulty switches (abnormal stem size) into the market, would you perform a recall on those from your supplier (like NovelKeys)?

Sent from my SM-G955U using Tapatalk

[Poesjuh]

I highly doubt it. Not only are there a lot of measurements out there already that really make it appear box stems are bigger. Plus, I actually have two different kinds of box blacks, both have the problem.

Look at the part where the led would be




Sent from my iPhone using Tapatalk

[MehAdviceGuy]

Speaking of measurements!
I have compiled a list of stem thicknesses of different switches that people has provided to me.
You can take a look at it here if anyone is interested: https://docs.google.com/spreadsheets/d/1Wc59WnMRJs0gVBrOyh2bRQcPl28yL7lim-p-N0l1daQ/edit?usp=sharing

If you have some measurements of stems then send them my way and i will update the list with what you give to me.

I also trimmed the BOX part of a BOX Jade keyswitch to see if the stem size widened at the end, I can say that on my sample size of one that it does not widen or narrow at the end.

[otanishock]

Speaking of measurements!
I have compiled a list of stem thicknesses of different switches that people has provided to me.
You can take a look at it here if anyone is interested: https://docs.google.com/spreadsheets/d/1Wc59WnMRJs0gVBrOyh2bRQcPl28yL7lim-p-N0l1daQ/edit?usp=sharing

If you have some measurements of stems then send them my way and i will update the list with what you give to me.

I also trimmed the BOX part of a BOX Jade keyswitch to see if the stem size widened at the end, I can say that on my sample size of one that it does not widen or narrow at the end.

Show Image

Thank you so much for the list. Now that we are able to compare the stem sizes of box switches to their Cherry counterparts', its no deniable that the stems on box switches are consistently larger than what they should've been.

Sent from my SM-G955U using Tapatalk

[ideus]

Hi,I'm Bella from Kailh,actually when we got the message about the keycap crack...

There is an MD order for Kailh "Heavy" switches that is scheduled to be shipped in September. Will it be fixed? or Will those switches be certified that they will not break key cap stems?


Please be sure that you fix the wider stems. You shall review your specification considering the reported stems broken. The key cap set's cannot be the final go/no-go gages for oversized switch stems.

[LevelSteam]

I actually ordered a pair of high precision calipers just to look into this more myself, so I'll update here with a table of what I find too. I think I also have at least one of each of the box switches from a sampler pack so I'll be able to check them each out.

[mgsickler]

Here is my update. Sorry for just linking to reddit - https://www.reddit.com/r/MechanicalKeyboards/comments/935f6l/box_switch_updates_from_novelkeys/

[voight-kampff]

I have got a workaround for the problem, if one has some loved box switches.
Just put a metal keycap a few times on the box switch, and the stem will be a little tighter.

I used a cheap metal keycap which I got from ebay.

[Giorgio]

Here is my update. Sorry for just linking to reddit - https://www.reddit.com/r/MechanicalKeyboards/comments/935f6l/box_switch_updates_from_novelkeys/

Please read this carefully and thoroughly. I am only speaking on behalf of NovelKeys, and somewhat Kaihua (Kailh) with the information that I have been given. Please do not ask me questions about unrelated topics on this other than the BOX switches, or ask about other companies.

Some background and current information
BOX switches currently have an x axis that is 1.32mm (+- .02mm tolerance). According to Kailh's internal testing, their switches have all been in that spec.
BOX switches' x axis, however, is wider than most other MX style switches. Some users have reported higher measurements (I just got in a really nice set of calipers. All my testing shows that these are within the tolerances), but Kailh's official word is that the switches they have measured have been within spec.
The reason that the BOX switch has a 1.32mm X axis is because DAREU asked Kaihua to widen the BOX switch from 1.3mm to 1.32mm. At the time, DAREU was really the only company that was buying BOX switches. Because of this, Kaihua agreed to change the dimensions for DAREU.
Tolerances within keycaps are unknown, and there have been reports of this varying greatly.
NovelKeys has halted all BOX switch sales

Going forward
I have talked to Kaihua for quite some length about all of these issues. They have agreed to change the stems to 1.3mm (+- .02mm tolerance) for at least my orders.

This means that all current (NovelKeys) Massdrop orders and Novelias will be getting a 1.3mm (=- .02mm) x axis switch.

In regards to (NovelKeys) MD orders, I actually am receiving the switches I ordered today. I have already placed the corrected order and will eat the costs.

This will delay the batch of NovelKeys x Kailh BOX switches (this is the drop that had Dark Yellow, Pale Blue, Navy, and Jade switches). It shouldnt delay the other rounds.
In regards to Novelias, I actually am receiving the switches today as well. I have placed a replacement order for the updated stems.

This may cause a slight delay to this order.
Regarding my stock - I will slowly be getting replacements in for all my stock. This will take some time, and I currently do not have an ETA on this. Current stock may be relisted at some point, but I will have to figure some more things out for that.

[Kevadu]

While I'm glad that the problem is being addressed that's little comfort for those of us who already own a bunch of box switches...

I'm also curious why DareU requested a thicker X stem in the first place.

[PoochZag]

Will there be some kind of visual indicator of the new stems?  In the future it could be really difficult telling the difference when buying used boards or switches, and not everyone is going to have nice calipers to verify the difference in a few hundredths of millimeters

Edit: Just saw on the Reddit post that this is uncertain as of now

[ideus]

I have got a workaround for the problem, if one has some loved box switches.
Just put a metal keycap a few times on the box switch, and the stem will be a little tighter.

I used a cheap metal keycap which I got from ebay.

Show Image

Any upcoming Kailh order should include a cheap metal keycap as reworking tool. Just to be sure it will not brake any expensive keycap set.

[portbaron]

Well this is good news, at least steps are being taken. I'll want to get a hold of one of those 'reworking' tools because I really like box clicks I have now.

[pixelpusher]

I have got a workaround for the problem, if one has some loved box switches.
Just put a metal keycap a few times on the box switch, and the stem will be a little tighter.

I used a cheap metal keycap which I got from ebay.

Show Image

This is brilliant.  You would save me a ton of money if it works.  I'm wondering if someone could manufacture a metal one specifically to slim down box switches?

[mkarlsson]

I have got a workaround for the problem, if one has some loved box switches.
Just put a metal keycap a few times on the box switch, and the stem will be a little tighter.

I used a cheap metal keycap which I got from ebay.

Show Image

This is brilliant.  You would save me a ton of money if it works.  I'm wondering if someone could manufacture a metal one specifically to slim down box switches?

Great idea! Have to get one of those for my jades.

[ideus]

I have got a workaround for the problem, if one has some loved box switches.
Just put a metal keycap a few times on the box switch, and the stem will be a little tighter.

I used a cheap metal keycap which I got from ebay.

Show Image

This is brilliant.  You would save me a ton of money if it works.  I'm wondering if someone could manufacture a metal one specifically to slim down box switches?

Please do so. A sort of cutter or grinder in the shape of a key cap stem may do wonders as a rework tooling.

[Giorgio]

It's impossible to get any precision with such a tool. You'll do 50 keycaps perfectly, and the other 50 will be too lose or too thick.

[MehAdviceGuy]

Response from kailh if someone is interested: https://www.reddit.com/r/MechanicalKeyboards/comments/93c8js/response_for_kailh_box_switches_crack_and_stress/

[mkarlsson]

Here is my update. Sorry for just linking to reddit - https://www.reddit.com/r/MechanicalKeyboards/comments/935f6l/box_switch_updates_from_novelkeys/

Please read this carefully and thoroughly. I am only speaking on behalf of NovelKeys, and somewhat Kaihua (Kailh) with the information that I have been given. Please do not ask me questions about unrelated topics on this other than the BOX switches, or ask about other companies.

Some background and current information
BOX switches currently have an x axis that is 1.32mm (+- .02mm tolerance). According to Kailh's internal testing, their switches have all been in that spec.
BOX switches' x axis, however, is wider than most other MX style switches. Some users have reported higher measurements (I just got in a really nice set of calipers. All my testing shows that these are within the tolerances), but Kailh's official word is that the switches they have measured have been within spec.
The reason that the BOX switch has a 1.32mm X axis is because DAREU asked Kaihua to widen the BOX switch from 1.3mm to 1.32mm. At the time, DAREU was really the only company that was buying BOX switches. Because of this, Kaihua agreed to change the dimensions for DAREU.
Tolerances within keycaps are unknown, and there have been reports of this varying greatly.
NovelKeys has halted all BOX switch sales

Going forward
I have talked to Kaihua for quite some length about all of these issues. They have agreed to change the stems to 1.3mm (+- .02mm tolerance) for at least my orders.

This means that all current (NovelKeys) Massdrop orders and Novelias will be getting a 1.3mm (=- .02mm) x axis switch.

In regards to (NovelKeys) MD orders, I actually am receiving the switches I ordered today. I have already placed the corrected order and will eat the costs.

This will delay the batch of NovelKeys x Kailh BOX switches (this is the drop that had Dark Yellow, Pale Blue, Navy, and Jade switches). It shouldnt delay the other rounds.
In regards to Novelias, I actually am receiving the switches today as well. I have placed a replacement order for the updated stems.

This may cause a slight delay to this order.
Regarding my stock - I will slowly be getting replacements in for all my stock. This will take some time, and I currently do not have an ETA on this. Current stock may be relisted at some point, but I will have to figure some more things out for that.

Hi Giorgio,

I am not an engineer and may be totally wrong but...wouldn't it be safer to go with something like 1.28 mm? this way having error tolerances of +-0.02 you could end up with 1.26 or 1.30. Going for 1.30 could end up with 1.28 or 1.32 which would have the same problem that we currently have. Am I wrong?

[Giorgio]

Response from kailh if someone is interested: https://www.reddit.com/r/MechanicalKeyboards/comments/93c8js/response_for_kailh_box_switches_crack_and_stress/

LOL. English.

[Giorgio]

Here is my update. Sorry for just linking to reddit - https://www.reddit.com/r/MechanicalKeyboards/comments/935f6l/box_switch_updates_from_novelkeys/

Please read this carefully and thoroughly. I am only speaking on behalf of NovelKeys, and somewhat Kaihua (Kailh) with the information that I have been given. Please do not ask me questions about unrelated topics on this other than the BOX switches, or ask about other companies.

Some background and current information
BOX switches currently have an x axis that is 1.32mm (+- .02mm tolerance). According to Kailh's internal testing, their switches have all been in that spec.
BOX switches' x axis, however, is wider than most other MX style switches. Some users have reported higher measurements (I just got in a really nice set of calipers. All my testing shows that these are within the tolerances), but Kailh's official word is that the switches they have measured have been within spec.
The reason that the BOX switch has a 1.32mm X axis is because DAREU asked Kaihua to widen the BOX switch from 1.3mm to 1.32mm. At the time, DAREU was really the only company that was buying BOX switches. Because of this, Kaihua agreed to change the dimensions for DAREU.
Tolerances within keycaps are unknown, and there have been reports of this varying greatly.
NovelKeys has halted all BOX switch sales

Going forward
I have talked to Kaihua for quite some length about all of these issues. They have agreed to change the stems to 1.3mm (+- .02mm tolerance) for at least my orders.

This means that all current (NovelKeys) Massdrop orders and Novelias will be getting a 1.3mm (=- .02mm) x axis switch.

In regards to (NovelKeys) MD orders, I actually am receiving the switches I ordered today. I have already placed the corrected order and will eat the costs.

This will delay the batch of NovelKeys x Kailh BOX switches (this is the drop that had Dark Yellow, Pale Blue, Navy, and Jade switches). It shouldnt delay the other rounds.
In regards to Novelias, I actually am receiving the switches today as well. I have placed a replacement order for the updated stems.

This may cause a slight delay to this order.
Regarding my stock - I will slowly be getting replacements in for all my stock. This will take some time, and I currently do not have an ETA on this. Current stock may be relisted at some point, but I will have to figure some more things out for that.

Hi Giorgio,

I am not an engineer and may be totally wrong but...wouldn't it be safer to go with something like 1.28 mm? this way having error tolerances of +-0.02 you could end up with 1.26 or 1.30. Going for 1.30 could end up with 1.28 or 1.32 which would have the same problem that we currently have. Am I wrong?

Sorry for not quoting correctly. It wasn't written by me.

[mkarlsson]

Here is my update. Sorry for just linking to reddit - https://www.reddit.com/r/MechanicalKeyboards/comments/935f6l/box_switch_updates_from_novelkeys/

Please read this carefully and thoroughly. I am only speaking on behalf of NovelKeys, and somewhat Kaihua (Kailh) with the information that I have been given. Please do not ask me questions about unrelated topics on this other than the BOX switches, or ask about other companies.

Some background and current information
BOX switches currently have an x axis that is 1.32mm (+- .02mm tolerance). According to Kailh's internal testing, their switches have all been in that spec.
BOX switches' x axis, however, is wider than most other MX style switches. Some users have reported higher measurements (I just got in a really nice set of calipers. All my testing shows that these are within the tolerances), but Kailh's official word is that the switches they have measured have been within spec.
The reason that the BOX switch has a 1.32mm X axis is because DAREU asked Kaihua to widen the BOX switch from 1.3mm to 1.32mm. At the time, DAREU was really the only company that was buying BOX switches. Because of this, Kaihua agreed to change the dimensions for DAREU.
Tolerances within keycaps are unknown, and there have been reports of this varying greatly.
NovelKeys has halted all BOX switch sales

Going forward
I have talked to Kaihua for quite some length about all of these issues. They have agreed to change the stems to 1.3mm (+- .02mm tolerance) for at least my orders.

This means that all current (NovelKeys) Massdrop orders and Novelias will be getting a 1.3mm (=- .02mm) x axis switch.

In regards to (NovelKeys) MD orders, I actually am receiving the switches I ordered today. I have already placed the corrected order and will eat the costs.

This will delay the batch of NovelKeys x Kailh BOX switches (this is the drop that had Dark Yellow, Pale Blue, Navy, and Jade switches). It shouldnt delay the other rounds.
In regards to Novelias, I actually am receiving the switches today as well. I have placed a replacement order for the updated stems.

This may cause a slight delay to this order.
Regarding my stock - I will slowly be getting replacements in for all my stock. This will take some time, and I currently do not have an ETA on this. Current stock may be relisted at some point, but I will have to figure some more things out for that.

Hi Giorgio,

I am not an engineer and may be totally wrong but...wouldn't it be safer to go with something like 1.28 mm? this way having error tolerances of +-0.02 you could end up with 1.26 or 1.30. Going for 1.30 could end up with 1.28 or 1.32 which would have the same problem that we currently have. Am I wrong?

Sorry for not quoting correctly. It wasn't written by me.

Oh sorry... to mgsickler then

[ptiede]

According to Mike 1.3 is the same as the Kailh speed switches which haven't shown any issues.

On another note, something else Mike said that seems emminently reasonable is there is evidence that part of what's going on isn't just that the spec was 1.32 but that the gmk caps have much looser tolerances for the cross size and those can get pretty small. So the stems cracks when you have a larger box stem and a narrow cap.

[ideus]

According to Mike 1.3 is the same as the Kailh speed switches which haven't shown any issues.

On another note there is evidence that part of what's going on isn't just that the spec was 1.32 but that the gmk caps have much looser tolerances for the cross size and those can get pretty small. So the stems cracks when you have a larger box stem and a narrow cap.

If the were true it may happen with other switches and not only with Kailh's.

[ptiede]

According to Mike 1.3 is the same as the Kailh speed switches which haven't shown any issues.

On another note, there is evidence that part of what's going on isn't just that the spec was 1.32 but that the gmk caps have much looser tolerances for the cross size and those can get pretty small. So the stems cracks when you have a larger box stem and a narrow cap.

If the were true it may happen with other switches and not only with Kailh's.

Maybe, but not necessarily. 1.34 may be the limit where the GMK tolerances start to go haywire. The thing is nobody knows what gmk's tolerances are since they haven't told anyone yet, and that information isn't readily available. But at the end of the day, this whole thing is a mess and there are probably many factors that are going into the stems cracking.

[mkarlsson]

According to Mike 1.3 is the same as the Kailh speed switches which haven't shown any issues.

On another note, there is evidence that part of what's going on isn't just that the spec was 1.32 but that the gmk caps have much looser tolerances for the cross size and those can get pretty small. So the stems cracks when you have a larger box stem and a narrow cap.

If the were true it may happen with other switches and not only with Kailh's.

Maybe, but not necessarily. 1.34 may be the limit where the GMK tolerances start to go haywire. The thing is nobody knows what gmk's tolerances are since they haven't told anyone yet, and that information isn't readily available. But at the end of the day, this whole thing is a mess and there are probably many factors that are going into the stems cracking.

I still think is risky going to the limit at 1.32. I was just suggesting why not going to a safer size closer to standard cherry switches.

[ptiede]

I think their point is none of kailhs non-box switches have cracked stems. So they are just making box stems have the same average thickness at the non box ones so it really isn't risky. Actually the box ones will now have tighter tolerances than non-box. That is, speeds are 1.30±0.03 and the new box stems will be 1.30±0.02

[ideus]

Kailh's cross design is not MX COMPATIBLE as it is designed now. It is oversized and out of actual MX stack of tolerances. The only option is to retooled it down to standard MX tolerances.

[ptiede]

That's not what I said... Im sorry if it reads like that. All I said was that there may be a few factors in the stems cracking. One of course are the Kailh stems, it just may not be the only but it is definitely a large part.

[chuckdee]

The reply on deskthority comes with their own measurements.

https://deskthority.net/keyboards-f2/response-for-kailh-box-switches-crack-and-stress-keycap-stems-t19539.html

[MehAdviceGuy]

We now have people reporting in that XDA Keycaps have started to stretch and crash from BOX switches: https://www.reddit.com/r/MechanicalKeyboards/comments/93jwkq/psa_xda_caps_will_break_on_box_switches/

I have also heard that cheap keycaps should be having issues. (I have destroyed some from a old AJAZZ keyboard by putting the keycaps on BOX Jade switches.)

[Giorgio]

According to Mike 1.3 is the same as the Kailh speed switches which haven't shown any issues.

On another note, something else Mike said that seems emminently reasonable is there is evidence that part of what's going on isn't just that the spec was 1.32 but that the gmk caps have much looser tolerances for the cross size and those can get pretty small. So the stems cracks when you have a larger box stem and a narrow cap.

Yeah. GMK never had a problem in 30 years, and suddenly kailh appears and teaches us how to do keycaps. And switches. After making those switches bigger.

[MehAdviceGuy]

According to Mike 1.3 is the same as the Kailh speed switches which haven't shown any issues.

On another note, something else Mike said that seems emminently reasonable is there is evidence that part of what's going on isn't just that the spec was 1.32 but that the gmk caps have much looser tolerances for the cross size and those can get pretty small. So the stems cracks when you have a larger box stem and a narrow cap.

Yeah. GMK never had a problem in 30 years, and suddenly kailh appears and teaches us how to do keycaps. And switches. After making those switches bigger.

Its wonderful how they change stuff, others stuff break, and they start accusing others of not having their tolerances in check. Fun times to have just put a board together with BOX Jade's...

[Giorgio]

I think their point is none of kailhs non-box switches have cracked stems. So they are just making box stems have the same average thickness at the non box ones so it really isn't risky. Actually the box ones will now have tighter tolerances than non-box. That is, speeds are 1.30±0.03 and the new box stems will be 1.30±0.02

Yeah, sure. 0.01mm of difference, that's 10 microns. No thanks. I'll box them and pack them.

Just think that with a temperature difference of 50 degrees celsius (like going from summer to winter), a length of 1.3mm of ABS becomes 0.00702mm larger, supposing that they're using the highest quality of abs available.
So the thermal expansion is equal to the new tolerances set by kailh (0.01mm vs 0.007mm which is a difference smaller than 40%), thus making the new tolerance absolutely insignificant and useless.
You leave your keyboard by the window in summer or winter? Boom, keycaps explode. This is aggravated by the fact that you have to take into account the fact that keycaps become smaller and bigger faster than the stem, and that have a bigger surface.

1.3mm * 50 degrees celsius * 108 Linear Temperature Expansion Coefficient /1000000= 0.00702mm
1.3*50*108/1000000

[KHAANNN]

I ordered these from Aliexpress, but I don't think the seller's going to ship, and it's a very slim chance that any of them would even have the ability to file inside the box stem - I've checked so many possible file's - none of them were that slim

The metal keycap idea is awesome, indeed if someone produces a metal part designed to slim down stems, that would be beyond awesome

Actually, we can design the part, and order it from Shapeways, with a sand-blasted material, the sand-blasting should turn the part into a grinder

Unrelated, DareU, more like, FYourKeycapsFU, with the amount of bad karma they got now, their lives will probably be ruined forever

[MehAdviceGuy]

(Attachment Link)

I ordered these from Aliexpress, but I don't think the seller's going to ship, and it's a very slim chance that any of them would even have the ability to file inside the box stem - I've checked so many possible file's - none of them were that slim

The metal keycap idea is awesome, indeed if someone produces a metal part designed to slim down stems, that would be beyond awesome

Actually, we can design the part, and order it from Shapeways, with a sand-blasted material, the sand-blasting should turn the part into a grinder

Unrelated, DareU, more like, FYourKeycapsFU, with the amount of bad karma they got now, their lives will probably be ruined forever

Why not just remove the box part with a razor blade?

[KHAANNN]

(Attachment Link)

I ordered these from Aliexpress, but I don't think the seller's going to ship, and it's a very slim chance that any of them would even have the ability to file inside the box stem - I've checked so many possible file's - none of them were that slim

The metal keycap idea is awesome, indeed if someone produces a metal part designed to slim down stems, that would be beyond awesome

Actually, we can design the part, and order it from Shapeways, with a sand-blasted material, the sand-blasting should turn the part into a grinder

Unrelated, DareU, more like, FYourKeycapsFU, with the amount of bad karma they got now, their lives will probably be ruined forever

Why not just remove the box part with a razor blade?

I assume they are not designed to work that way, haven't tested, let me know if you make it work (It needs to be wobble-free etc.)

[MehAdviceGuy]

(Attachment Link)

I ordered these from Aliexpress, but I don't think the seller's going to ship, and it's a very slim chance that any of them would even have the ability to file inside the box stem - I've checked so many possible file's - none of them were that slim

The metal keycap idea is awesome, indeed if someone produces a metal part designed to slim down stems, that would be beyond awesome

Actually, we can design the part, and order it from Shapeways, with a sand-blasted material, the sand-blasting should turn the part into a grinder

Unrelated, DareU, more like, FYourKeycapsFU, with the amount of bad karma they got now, their lives will probably be ruined forever

Why not just remove the box part with a razor blade?

I assume they are not designed to work that way, haven't tested, let me know if you make it work (It needs to be wobble-free etc.)

I'll try my circumcised BOX jade tonight and give you a update on how much it wobbles. But again, that is a sample size of one.

[ptiede]

According to Mike 1.3 is the same as the Kailh speed switches which haven't shown any issues.

On another note, something else Mike said that seems emminently reasonable is there is evidence that part of what's going on isn't just that the spec was 1.32 but that the gmk caps have much looser tolerances for the cross size and those can get pretty small. So the stems cracks when you have a larger box stem and a narrow cap.

Yeah. GMK never had a problem in 30 years, and suddenly kailh appears and teaches us how to do keycaps. And switches. After making those switches bigger.

Its wonderful how they change stuff, others stuff break, and they start accusing others of not having their tolerances in check. Fun times to have just put a board together with BOX Jade's...

Ok so I am not affiliated with kailh or have anything do with them. So don;t put my words in their mouth.

I think their point is none of kailhs non-box switches have cracked stems. So they are just making box stems have the same average thickness at the non box ones so it really isn't risky. Actually the box ones will now have tighter tolerances than non-box. That is, speeds are 1.30±0.03 and the new box stems will be 1.30±0.02

Yeah, sure. 0.01mm of difference, that's 10 microns. No thanks. I'll box them and pack them.

Just think that with a temperature difference of 50 degrees celsius (like going from summer to winter), a length of 1.3mm of ABS becomes 0.00702mm larger, supposing that they're using the highest quality of abs available.
So the thermal expansion is equal to the new tolerances set by kailh (0.01mm vs 0.007mm which is a difference smaller than 40%), thus making the new tolerance absolutely insignificant and useless.
You leave your keyboard by the window in summer or winter? Boom, keycaps explode. This is aggravated by the fact that you have to take into account the fact that keycaps become smaller and bigger faster than the stem, and that have a bigger surface.

1.3mm * 50 degrees celsius * 108 Linear Temperature Expansion Coefficient /1000000= 0.00702mm
1.3*50*108/1000000

Cool you calculated a linear thermal expansion coefficient and assumed it applies to a range of 50 degrees. But also your calculation doesn't even apply like you said it did. So you have a 50 degree spread. That means you want the temperature in your house to be fifty degrees warmer than in Kailh's factory when they do their measurement. So let's assume that Kaihua's factory is at 15 degrees Celcius (way lower than the average ambient temperature but ok) then your house must be 65 degrees Celcius...

I am NOT saying Kailh isn't at fault here, of course, they are, I can see with my own two eyes that stems are cracking. The point I was trying to make was that not every stem is cracking, so some tolerances are too loose or there is a compounding issue here. Like some of GMK's caps are a little smaller than normal and combined with the Kailh switches that are on the larger size of the tolerances can lead to cracking. Materials science isn't easy. Also, there is a massive selection bias in this reporting. People are only going to report if their caps cracked (as they should it is still way too common),

[Giorgio]

LOL.

It's an isotropic material so any other coefficient is a combination of it.

And about the 50 degrees. That's obviously a simplification, but it's no unreal at all. Considering that they measure their mold when it's cold. And that when it's hot for the heat of plastic, it becomes bigger.  But hey, maybe I found where the problem lies.

But that's just to show that thermal expansion and their new tolerances are in the same order of magnitude. And that's what's important when you talk about superimposing different effects.

[Brammm87]

Just took keycaps of 5 boards with BOX switches (black, white, brown and navy), no cracks but all showed signs of stress on the stems... This ****ing blows.

[Giorgio]

Just took keycaps of 5 boards with BOX switches (black, white, brown and navy), no cracks but all showed signs of stress on the stems... This ****ing blows.

Mother of god. I can't imagine how upset I would be in seeing thousands of my dollars at risk for the gross incompetence of a third party.

[Brammm87]

Just took keycaps of 5 boards with BOX switches (black, white, brown and navy), no cracks but all showed signs of stress on the stems... This ****ing blows.

Mother of god. I can't imagine how upset I would be in seeing thousands of my dollars at risk for the gross incompetence of a third party.

I'm trying to be reasonable. It looks like all keycaps will be able to go back on boards and I never really intended to resell caps (I'm in this hobby for the collecting aspect). I am however considering trying to get my money back, or at least my switches replaced once the manufacturing process has been updated. I really love my box blacks and whites (not a big fan of the browns and navy's so far).

[ideus]

This information may be of interest here. It is the answer I got from MD's customer service on the BOZ switches's latest drop, that it is scheduled to be shipped in one and a half month or so.


Marriah D. (Massdrop Community Support)
[size=0pt]Aug 2, 9:42 AM PDT [/size]
[size=0pt]........
 
 Marriah here. Thank you for contacting Massdrop Community Support. I understand your concern with the recent issues with BOZ switches and I will be happy to provide more information.
 
 At this time, Novelkeys and Kailh are aware of this problem and are working on getting the switches corrected. It has been communicated that they are working on getting this issue fixed prior to this drops shipment date in order to ensure that members receive the corrected versions of the BOX switches.
 
 If a delay does occur, we will be sure to communicate this out to members as soon as we are aware.
 
 I hope this information is helpful. Please feel free to reach back out with any other questions or concerns.
 
 Best Wishes,[/size]
 [size=0pt]Marriah[/size]

[Giorgio]

This information may be of interest here. It is the answer I got from MD's customer service on the BOZ switches's latest drop, that it is scheduled to be shipped in one and a half month or so.


Marriah D. (Massdrop Community Support)
[size=0pt]Aug 2, 9:42 AM PDT [/size]
[size=0pt]........
 
 Marriah here. Thank you for contacting Massdrop Community Support. I understand your concern with the recent issues with BOZ switches and I will be happy to provide more information.
 
 At this time, Novelkeys and Kailh are aware of this problem and are working on getting the switches corrected. It has been communicated that they are working on getting this issue fixed prior to this drops shipment date in order to ensure that members receive the corrected versions of the BOX switches.
 
 If a delay does occur, we will be sure to communicate this out to members as soon as we are aware.
 
 I hope this information is helpful. Please feel free to reach back out with any other questions or concerns.
 
 Best Wishes,[/size]
 [size=0pt]Marriah[/size]

Doesn't matter. All the "fixed" stems are still completely wrong if the do what they said.

[ptiede]

This information may be of interest here. It is the answer I got from MD's customer service on the BOZ switches's latest drop, that it is scheduled to be shipped in one and a half month or so.


Marriah D. (Massdrop Community Support)
[size=0pt]Aug 2, 9:42 AM PDT [/size]
[size=0pt]........
 
 Marriah here. Thank you for contacting Massdrop Community Support. I understand your concern with the recent issues with BOZ switches and I will be happy to provide more information.
 
 At this time, Novelkeys and Kailh are aware of this problem and are working on getting the switches corrected. It has been communicated that they are working on getting this issue fixed prior to this drops shipment date in order to ensure that members receive the corrected versions of the BOX switches.
 
 If a delay does occur, we will be sure to communicate this out to members as soon as we are aware.
 
 I hope this information is helpful. Please feel free to reach back out with any other questions or concerns.
 
 Best Wishes,[/size]
 [size=0pt]Marriah[/size]

Doesn't matter. All the "fixed" stems are still completely wrong if the do what they said.

You have zero actual evidence to support that claim. In fact, there is evidence against it. Kailh is again just changing the box stems to match their other ones. My Kailh Halo stems don't crack or stretch my caps, so why would the box ones? Like you have an argument why it may work, and hey maybe it is a decent one, but again what matters is what we see in the real world and so far we have no evidence that Kailh's other stems are cracking caps.