Alps Switches vs Cherry Switches - Which are more durable?

[iggysaps]

Considering that alps switches are usually rated at 20M actuations, case in point, Filco Zero, while cherry switches are usually rated at 50M actuations, case in point Filco's with any cherry switches.  Anybody actually feel that the alps were less durable than cherry switches?

A 30M difference in the number of actuations is a lot...

[lowpoly]

Only linear Cherrys are 50M. Blues and browns are less.

[ironcoder]

is that 20 thousand or 20 million?

[ch_123]

Million. You'd be in a pretty bad way if your keyboard could only withstand 20,000 keypresses...

[huha]

Even 20 Million is a lot. If you really need a durable keyboard, you have to go for capacitive, optical or reed switches.
Debating on the durability of a $100 keyboard is somewhat pointless. 20 Million keypresses is a lot, and should your keyboard fail due to reaching this number of keypresses, you can always just buy a new one, since it will long have payed for itself.

Durability should only be a concern if you want to install the stuff somewhere extremely inaccessible, where it is often used and should last a long time.

-huha

[cchan]

Only linear Cherrys are 50M. Blues and browns are less.

I believe that the latest spec for Cherry MX switches quotes every type (linear, tactile, ergo tactile, or clicky) to have a 50 million actuations rating. The non-linear ones used to be 20M though.

[itlnstln]

I believe that the latest spec for Cherry MX switches quotes every type (linear, tactile, ergo tactile, or clicky) to have a 50 million actuations rating. The non-linear ones used to be 20M though.

Seconded.  This is correct.

[IBI]

Has anyone seen any cherry/alps switches that are worn out?

[itlnstln]

Has anyone seen any cherry/alps switches that are worn out?

I've seen them worn in, but never worn out.

[keyb_gr]

Has anyone seen any cherry/alps switches that are worn out?

Well, I have a board with two Cherry blues not clicking properly... but I guess that's dust/dirt inside (which can enter during actuation), since things got better for one of them at least after an attempt at blowing it out with lungpower. Should get a can of compressed air one of those days. One would assume the tactile switches to be more susceptible to dirt than the linear ones.

Cherry blacks seem to be very tough. They do get somewhat lighter (which is quite nice actually), but that's about it. The lasered lettering on PBT keycaps fares less well...

[itlnstln]

The lasered lettering on PBT keycaps fares less well...

Wow, really?  I have never had any problems with lasered lettering before.

[keyb_gr]

Wow, really?  I have never had any problems with lasered lettering before.

Nor do I - but on my newer G80 with blacks the lettering on the numpad enter is worn very thin in part, not to mention the number of shiny keys and the shiny spots on the case. Yes, this one was used a lot in its former life (which only was 7 years max). It also needed a complete teardown badly, as there was this black dust/dirt (toner?) everywhere. Works like a champ though...

So it's not like the lasered lettering would be bad, but like I said those blacks are tough...

[itlnstln]

Nor do I - but on my newer G80 with blacks the lettering on the numpad enter is worn very thin in part, not to mention the number of shiny keys and the shiny spots on the case. Yes, this one was used a lot in its former life (which only was 7 years max). It also needed a complete teardown badly, as there was this black dust/dirt (toner?) everywhere. Works like a champ though...
 
So it's not like the lasered lettering would be bad, but like I said those blacks are tough...

The G80 I have doesn't quite have the polish of the Filco, but I find that I type a little faster on it.  It seems that they use a lighter plastic on the G80 than Costar uses on their 'boards, and that gives the keys a little lighter feel, IMO.  I have a hard time choosing which one I like better.

[ironcoder]

Million. You'd be in a pretty bad way if your keyboard could only withstand 20,000 keypresses...

Agreed, but a lot of stuff made today is ****e. Case in point, I got some freebie Chicony keyboards with some somewhat expensive new boxes and 2 out of 4 of them went bad in less than a week. I would have been happy if they would have lasted 2,000 keypresses.

[watduzhkstand4]

The alps feel more durable to me but I think Cherry lasts longer

[itlnstln]

IMO, Alps switches, real or fake, don't feel as refined as Cherrys.  I don't think either one is more durable than the other, though.  Alps feel more industrial and brute-force where Cherrys feel more like butter.  Kinda like a GMC Yukon vs. a Cadillac Escalade.  Same size, durability, performance, etc., but one is more refined than the other.

[talis]

Someone may have done this calculation before but :

If you assume English, e is the most common character at 12.702% (ignoring the space bar, capitols and all that).

That means a key rated at 20,000,000 activations will fail after a total of 157,355,518 total key presses (if 12.7% of them are on a single key).

Using the standard definition of "word" at 5 characters that's a total of 31,491,103 words.

If you can type at a constant 80wpm, the most use key will hit its life after 393,639min (273 days) of constant typing.  A key rated at 50M key presses should last 2.5x as long.

Seems like the new best thing will come along long before either 20M or 50M activation keys will wear out =P.

[itlnstln]

Please do. I'm curious to see if that fixes it or if you have to disassemble the key and then see wear points.
 
First actual Cherry Key (or buckling spring for that matter) that I've seen that may be worn out.
 
My kid tears through Logitechs but that's partly because he's a slob at the keyboard and games constantly. Wonder where he inherited that from?

This issue used to pop up on brand new Das IIIs.  There would be some dust or something that would keep the little white piece on the switch from popping up or sliding down.  Once the user would press the switch rapidly several times or take the cap off, press the switch down (so the internals were accessible), and blow it out with compressed air, it would click again.

[IBI]

I think Cherry lasts longer

Based on what?

My kid tears through Logitechs but that's partly because he's a slob at the keyboard and games constantly.  Wonder where he inherited that from?

Perhaps you should use that as an excuse to get the spillproof cymotion expert.

[watduzhkstand4]

Based on what?

it says it on the site that it lasts 50 million operations.

[itlnstln]

it says it on the site that it lasts 50 million operations.

So are Cherrys.  The 20 million is an old number; it has since been updated.  Really, once you get over 15-20 million or so, the number is just a SWAG, IMO.  Honestly, you'll buy another keyboard before you wear a switch out.

[chimera15]

Has anyone seen any cherry/alps switches that are worn out?

Bob from Northgate keyboard repair sent me a bag of about 100 alps switches that had gone bad.  Eventually the spring that actuates against the switch to actually produce the contact gets bent out of position and fails.  They also, at about half-3/4ths their lifespan get really sloppy, stiff, or aberrant in their behavior in one way or another, especially the simplified fakes.   Alps switches are *super* easy to fix or replace though.  

  I would say that the comparison was different depending on the alps switch.  If you're talking alps complicated, I think they have a much longer lifespan than simplified fakes will, especially in their consistency of feel over their lifespan.

[ch_123]

Seems like the new best thing will come along long before either 20M or 50M activation keys will wear out =P.

Considering that most of us are typing on keyboards whose key-switch technology has remained unchanged since the 1980s, I'm not sure about that one...

Just to throw it out there - Unicomp's old websites give a 25 million figure for their buckling spring keyboards, and something like 4 million for their M4s.

[keyb_gr]

So, digging up this eldely thread... anyone found any worn-out MX blacks yet?

I bet that would make a good PR gag for Cherry - send us your board with one or more really worn-out (failed due to normal usage) switches and get a new equivalent type in return, or something like that. I don't think they'd lose a lot of money...

[ch_123]

I should point out that after using more than one worn out AT102W, Alps don't last forever. I have never used any Cherry board other than my brand new Filco so I can't really give a comparison there.

[megarat]

If you assume English, e is the most common character at 12.702% (ignoring the space bar, capitols and all that).

I don't think you can ignore the other keys.  I suspect that the switches most likely to go first would be those associated with the space bar, enter and/or delete, and the modifier keys (shift/command/control, etc.)

It would be interesting to see what some real-world metrics come out to.  Hey, that sounds like an actual non-mischievous use for a keystroke recorder!

[namelessguy]

I don't think you can ignore the other keys.  I suspect that the switches most likely to go first would be those associated with the space bar, enter and/or delete, and the modifier keys (shift/command/control, etc.)

It would be interesting to see what some real-world metrics come out to.  Hey, that sounds like an actual non-mischievous use for a keystroke recorder!

Well, this is what I heard;
Topre keyboards in banks (they are renouned for their quality and durability, and are the de-facto standard among Japanese bank staffs now) are known to fail after a year or two of sustained heavy use.
Usually the space bar is the first one to break, sometimes the enter key, though I don't have information on how they are getting broken.

Looks like 30,000,000 keypresses are far from enough in these places.

[keyb_gr]

Assuming an 8 hour work day with an average of 50 wpm (at that point I'd be more worried about wear on humans rather than keyboards) and a 15% space bar duty cycle, you "only" get about 13 million keypresses in 2 years. Hmm.

Anyone got some figures for medical transcriptionists?

[ch_123]

Also not good since I guess it's the official keyboard of the "Japanese Data Center" whatever that is.

Wiki

IIRC, the exact quote was "National data centers", so I assume those would be the data centers belonging to the various agencies of the Japanese government?

[BucklingSpring]

Only linear Cherrys are 50M. Blues and browns are less.

Nope... All Cherry MX's are good for 50M including tactile as well.

I know because I had an argument not long agao with a Deck rep claiming their new tactile board will use 50M Cherry switches. But when I looked at CherryCorp.com, MX taciles were rated 20M.

So I wrote to Cherry support and got a reply back from them confirming that all MX's are 50M.


From:  Keyboard Support
To:  me
Date:  Wed, 17 Dec 2008 10:38:47 -0400

Hello,

I apologize but our website has not been updated yet. All MX series
switches are rated at 50 million actuations.

Best regards,

Ed Ferraton
Senior CID Technical Services
Cherry Electrical Products
"

They finally updated their site with the correct information:
Click here to see for yourself