How reliable are mechanical keyboards?

[jamster]

I've been wondering this for a while now, as it comes up here. Usually in the form of casual comments about expecting a board to last 3, 5, 10 years. Or people commenting that mechanical boards are 'more reliable.'

The only two failures I've experienced are:

A heavily used Thinkpad scissor switch board which basically wore out after 5+ years of use. I wasn't surprised as the case was thin and flexy, and the PCB was being hammered every day.

A 1980s Northgate which succumbed to high heat and humidity in storage, and the casing went horribly nasty. It still worked.

Other than this, I might have had some random rubber domes fail over the decades, but they were so cheap that they wouldn't register in my memory. I have the overall impression that rubber domes are very reliable.

So, is keyboard reliably really an issue? I expect them to last... ages. At least 5-10 years, or more likely, I'd just replace them for some other reason. I've never gone with cheap boards with a bad reputation (e.g Razer), but these things are just the same switches soldered into PCBs. Seems to me that you'd have to have a very bad production process to screw this up.

[yui]

most boards i ever saw die were of liquid damage (user spilling their water or coffee) and controllers dying on very cheap rubber domes after a few years, i only ever managed to wear down 3 keyboards one was a rubber dome that lasted me for ever (it was the keyboard from my parent's 1st pc, lasted more than 20 years), it was my favorite board and when the sliders eventually went bad i switched to an M, and 2 laptops keyboards where the scissor switches on the most used keys have trouble registering (they have unknown history but are both at least 10 years old and were work computers before i bought them). so i guess that for most part keyboards are the most reliable piece of hardware in a pc, maybe less so with microcontroler based keyboards as those can get corrupt but then can always be flashed again.
(and i see a lot of spills on keyboard because i am working at IT and no one cares about their keyboards, or computer for that matter)

[jamster]

Yeah, I am deliberately excluding boards killed by abuse, that would be the main factor by orders of magnitude

[bbqkimchi]

I started getting mechs in 2009. My first MX variant, a Filco tenkeyless w/ cherry mx blues had a few keys that started chattering or registering two keystrokes. As the parts of a mech board are more or less replaceable, I was able to desolder the aging blues and put in some box jades. I anticipated that this would ultimately happen when I purchased it over a decade ago and naturally look forward to maintenance and repair tasks.  Repair options aren't abundant, if available at all, for rubber domes like they are for mechs. I have about ten mech boards now, 3 more on the way, that I anticipate will continue to function for at least the next 30 years as long as I take care of them. USB C and mini b connectors have to be completely obsolete and not compatible in any form w/ any amount of adapters for the boards to become useless. My keyboards will all outlive me.

[yui]

true that if you continuously fix what goes wrong it will outlive you, it is like cars

[Leslieann]

I've never bought a mechanical for it's durability, yes, you can fix it but the real benefit has always been the feel. Given equal conditions I wouldn't be surprised if a dome lasted just as long as any mechanical.

The biggest weakness of almost any keyboard is the owner above all else.

[Maledicted]

My Corsair K70 (for my gaming rig) has lasted me a good 6 or 7 years now. I clean it once every ... so many years it takes for the switches to literally start clogging up with enough debris to start not reliably registering, in which case I just pull the caps and clean the visible gunk off.

The closest I have had to seeing a keyboard die is the external plastic on/of switch ... shroud thing on one of my ancient Logitech wireless keyboard/trackpad combo boards (which already had shined caps when I bought it at Goodwill years ago, and I used daily as a HTPC keyboard) broke. I super glued the cheap plastic back together, but it broke again for good. The actual on/of switch is still fine within, so if I ever decide to stop being lazy, I'll probably mount a toggle switch or something on the case and wire it up to the PCB. For now, if I use it anymore, I flick it on and off by jamming a screwdriver or something inside.

I think one of my Model Ms may even have somewhat shined caps ... which are PBT, so I imagine that's had a hard life.

[rxc92]

Small things can break here or there on any board, but as long as you aren't smashing your keyboard or exposing them to (sticky, especially) liquids, there shouldn't be any problem. I've put my NerD and R2 in my luggage and book bag more times than I can count and dragged them all over the place with no issue, and I daresay a rubber dome would've held up just as well. Keyboards are durable things. Longevity is great as well, but considering how quickly people seem to cycle through keyboards in this community, it seems more like a secondary concern to enthusiasts.

[jamster]

A related question off this then would be: Are higher end boards longer lasting than budget boards? I recall a thread from weeks ago when someone was asking Keychron vs Leopold, and some answers were 'Leopold, better build, will last longer.

Personally, I'd expect longevity to be similar, minus factors guaranteed to degrade with time such as built-in batteries.

I still prefer higher quality builds for the tactile satisfaction, they simply feel better to hold.

[Maledicted]

A related question off this then would be: Are higher end boards longer lasting than budget boards? I recall a thread from weeks ago when someone was asking Keychron vs Leopold, and some answers were 'Leopold, better build, will last longer.

Personally, I'd expect longevity to be similar, minus factors guaranteed to degrade with time such as built-in batteries.

I still prefer higher quality builds for the tactile satisfaction, they simply feel better to hold.

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors? How many capacitors do you see in modern keyboards? I have 2 Unitek K151L boards from the late 80s that still work perfectly. Those things have literally the cheapest PCBs I have ever seen in a keyboard. I do believe that one did not work originally ... because some of the diode legs were literally rusted through, but I imagine that doesn't count? (lol) I have a Nan Tan keyboard from around the same period that literally feels like I could snap it in half with my bare hands. The screws on the bottom are rusted ... but it works perfectly.

I don't put much stock in anything somebody says when they recommend a brand, as opposed to a specific technology (capacitive buckling spring, hall effect, beam spring, etc), if that recommendation also includes some arbitrary claim about longevity vs some other brand using the exact same sensing technology.

[jamster]

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors? How many capacitors do you see in modern keyboards? I have 2 Unitek K151L boards from the late 80s that still work perfectly. Those things have literally the cheapest PCBs I have ever seen in a keyboard. I do believe that one did not work originally ... because some of the diode legs were literally rusted through, but I imagine that doesn't count? (lol) I have a Nan Tan keyboard from around the same period that literally feels like I could snap it in half with my bare hands. The screws on the bottom are rusted ... but it works perfectly.

I don't put much stock in anything somebody says when they recommend a brand, as opposed to a specific technology (capacitive buckling spring, hall effect, beam spring, etc), if that recommendation also includes some arbitrary claim about longevity vs some other brand using the exact same sensing technology.

I'm not entirely sold on MX switches, TBH. Not for any good reason beyond a friend having a switch fail mysteriously (not abuse) and anecdotal internet posts of "my one key stopped working." The 50M cycle thing just sounds too much like hard drive manufacturers and their old and very dubious MBTF figures

Opening them up, it is hard to see how one could fail.

[Maledicted]

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors? How many capacitors do you see in modern keyboards? I have 2 Unitek K151L boards from the late 80s that still work perfectly. Those things have literally the cheapest PCBs I have ever seen in a keyboard. I do believe that one did not work originally ... because some of the diode legs were literally rusted through, but I imagine that doesn't count? (lol) I have a Nan Tan keyboard from around the same period that literally feels like I could snap it in half with my bare hands. The screws on the bottom are rusted ... but it works perfectly.

I don't put much stock in anything somebody says when they recommend a brand, as opposed to a specific technology (capacitive buckling spring, hall effect, beam spring, etc), if that recommendation also includes some arbitrary claim about longevity vs some other brand using the exact same sensing technology.

I'm not entirely sold on MX switches, TBH. Not for any good reason beyond a friend having a switch fail mysteriously (not abuse) and anecdotal internet posts of "my one key stopped working." The 50M cycle thing just sounds too much like hard drive manufacturers and their old and very dubious MBTF figures

Opening them up, it is hard to see how one could fail.

If it helps, the Uniteks in those examples had MX blacks and the Nan Tan has Alps SKCM whites. I think MX is pretty reliable, myself. Definitely less susceptible to debri than Alps. Not everything can be a box switch .... or capacitive, etc.

[VP]

I have Das Keyboard for 7 years, used heavily, doesn't look like new, but works well, hope it continues to work well for another 7 years.

[fohat.digs]

what usually fails in any relatively small electronic?

That's easy. Sockets and plugs.

[nathanchere]

I have Das Keyboard for 7 years, used heavily, doesn't look like new, but works well, hope it continues to work well for another 7 years.

My full-size Das is a piece of ****. Solid build quality but hopeless PCB. They clearly haven't heard of debouncing.

[Rico]

I've used a Filco and a Topre Realforce for more than 10 years and they still doing stong.
On the other side I've broken countless Dell rubber domes (one per year) by gaming on them.

And for people that had problem with their mech keyboards I've always succeded repairing them by changing the swiches.

So not only mechs are reliable but most importantly they are repairable.

[VP]

My full-size Das is a piece of ****. Solid build quality but hopeless PCB. They clearly haven't heard of debouncing.

Did your Das stop working?

[jamster]

I have Das Keyboard for 7 years, used heavily, doesn't look like new, but works well, hope it continues to work well for another 7 years.

My full-size Das is a piece of ****. Solid build quality but hopeless PCB. They clearly haven't heard of debouncing.

I thought denouncing was a firmware level issue. Is there a way this is influenced by PCB design?

[Maledicted]

what usually fails in any relatively small electronic?

That's easy. Sockets and plugs.

I suppose that I mean outside of physical abuse.

[Hypersphere]

Regarding reliability of mechanical keyboards, I count Topre-switch boards as mechanical, and these have been employed in banks, where it is critical not to have issues with data entry. Likewise, the original IBM F62 "Kishsaver" keyboards were used in banks, attesting to their reliability.

In my own experience, I have had several rubber dome keyboards fail or partially fail (one or more keys stopped responding). One of these was a rather high-quality AST rubber dome keyboard.

Again, from my own experience, I have original IBM XT and Model M keyboards that are still working perfectly. The XT boards date from 1984 and the Model M boards from the early 1990s. My Topre boards are much more recent, but I have been hammering away at my main keyboard -- a lubed and silenced HHKB Pro 2 -- for five years, and it is still functioning perfectly. The same goes for my RealForce boards, although they have not seen as much use as my HHKB.

I have less experience with other types of mechanical keyboards, but my perception is that those with Cherry mx switches are quite reliable -- I have some rather old Wyse boards with Cherry black switches that still function smoothly. I prefer Alps over Cherry, but Alps switches are more susceptible to dirt.

[Findecanor]

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors?

The IBM Model M2 is notorious for faulty electrolytic capacitors failing over time. (Not from use but from not being used). I haven't heard of another keyboard model with this problem. Most capacitors I've seen have been ceramic.

Modern keyboards with more sensitive low-voltage microcontrollers may actually use more decoupling capacitors, but all are tiny SMD caps. There's no reason to use electrolytic capacitors in modern electronics unless you're handling high currents, which keyboards don't.

My full-size Das is a piece of ****. Solid build quality but hopeless PCB. They clearly haven't heard of debouncing.

Do you mean a few of your keys are chattering? Or are all of them?

[Kavik]

I can't specifically recall any keyboards that have died personally, but a former coworker had issues with multiple Dell rubber dome keyboards that would stop registering a single key, which is repairable on a mech with discrete switches but ruins and entire rubber dome keyboard. I think the same thing has happened to my family's old keyboards, but I don't really remember.

That said, I have had more issues with mechanical keyboards with faulty switches or bad PCBs or not being grounded or whatever.

[Missingnumber]

I know the conversation is mostly about prebuilt boards, but I thought I would throw in my experience. Getting in to the custom scene, I decided to get a KBDFans Tofu while waiting for a group buy board.

I haven't had a good experience with their reliability, since I've had two separate DZ65 RGB PCBs break on me from normal use. It seems the model had an issue with static discharge frying the controller. I was unable to fix them, but KBDFans was good enough to send me a replacement. There's a new version of the PCB out now, and I hope it has better ESD protection. 

Edit: Just wanted to add that this is not a problem unique to mechanical keyboards and their switch style, but it is a potential problem with the PCBs many custom mechanical keyboards use. It's something I never worried about with a rubber dome or even my previous prebuilt mechanical gamer-style board.

[Maledicted]

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors?

The IBM Model M2 is notorious for faulty electrolytic capacitors failing over time. (Not from use but from not being used). I haven't heard of another keyboard model with this problem. Most capacitors I've seen have been ceramic.

Modern keyboards with more sensitive low-voltage microcontrollers may actually use more decoupling capacitors, but all are tiny SMD caps. There's no reason to use electrolytic capacitors in modern electronics unless you're handling high currents, which keyboards don't.

That's pretty much where I was going with my comment.

[Polymer]

A related question off this then would be: Are higher end boards longer lasting than budget boards? I recall a thread from weeks ago when someone was asking Keychron vs Leopold, and some answers were 'Leopold, better build, will last longer.

Personally, I'd expect longevity to be similar, minus factors guaranteed to degrade with time such as built-in batteries.

I still prefer higher quality builds for the tactile satisfaction, they simply feel better to hold.

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors? How many capacitors do you see in modern keyboards? I have 2 Unitek K151L boards from the late 80s that still work perfectly. Those things have literally the cheapest PCBs I have ever seen in a keyboard. I do believe that one did not work originally ... because some of the diode legs were literally rusted through, but I imagine that doesn't count? (lol) I have a Nan Tan keyboard from around the same period that literally feels like I could snap it in half with my bare hands. The screws on the bottom are rusted ... but it works perfectly.

I don't put much stock in anything somebody says when they recommend a brand, as opposed to a specific technology (capacitive buckling spring, hall effect, beam spring, etc), if that recommendation also includes some arbitrary claim about longevity vs some other brand using the exact same sensing technology.

For what type of boards?  We know that early on there were some severe issues with soldering for some mechanical keyboard brands...there seem to be fewer problems now than there were but you still rely on soldering for many of these boards which means some will be done to a better standard than the other.

Comparing the technology and saying things like like MX boards are the same because they use the same type of switches is pretty ignorant given the past data...although there might be a point where that statement is true.

[noisyturtle]

I have never in 10 years of mech keyboarding had a single switch go bad or fail on me.
And I know if one does I can easily replace it in a matter of minutes.
Literally no downsides.

[Maledicted]

A related question off this then would be: Are higher end boards longer lasting than budget boards? I recall a thread from weeks ago when someone was asking Keychron vs Leopold, and some answers were 'Leopold, better build, will last longer.

Personally, I'd expect longevity to be similar, minus factors guaranteed to degrade with time such as built-in batteries.

I still prefer higher quality builds for the tactile satisfaction, they simply feel better to hold.

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors? How many capacitors do you see in modern keyboards? I have 2 Unitek K151L boards from the late 80s that still work perfectly. Those things have literally the cheapest PCBs I have ever seen in a keyboard. I do believe that one did not work originally ... because some of the diode legs were literally rusted through, but I imagine that doesn't count? (lol) I have a Nan Tan keyboard from around the same period that literally feels like I could snap it in half with my bare hands. The screws on the bottom are rusted ... but it works perfectly.

I don't put much stock in anything somebody says when they recommend a brand, as opposed to a specific technology (capacitive buckling spring, hall effect, beam spring, etc), if that recommendation also includes some arbitrary claim about longevity vs some other brand using the exact same sensing technology.

For what type of boards?  We know that early on there were some severe issues with soldering for some mechanical keyboard brands...there seem to be fewer problems now than there were but you still rely on soldering for many of these boards which means some will be done to a better standard than the other.

Comparing the technology and saying things like like MX boards are the same because they use the same type of switches is pretty ignorant given the past data...although there might be a point where that statement is true.

What is it, specifically, that's ignorant about my statements? I believe what you're talking about stems from a comment chain with an example of Leopold vs Keychron, saying that The Leopold will last longer based solely on the brand's reputation when both boards are most likely using the same switch technology. Even if we were talking about the entire history of random keyboards pitted against each other in this regard, have you seen any keyboards with failed solder joints that were not the result of physical abuse? I haven't yet, even on really cheap boards.

I think a cold joint, etc, would be a pretty wild exception to the general rule. Bad grounding and/or junk controllers or something might make more sense to me, and those would still be exceptions in my opinion.

[Polymer]

A related question off this then would be: Are higher end boards longer lasting than budget boards? I recall a thread from weeks ago when someone was asking Keychron vs Leopold, and some answers were 'Leopold, better build, will last longer.

Personally, I'd expect longevity to be similar, minus factors guaranteed to degrade with time such as built-in batteries.

I still prefer higher quality builds for the tactile satisfaction, they simply feel better to hold.

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors? How many capacitors do you see in modern keyboards? I have 2 Unitek K151L boards from the late 80s that still work perfectly. Those things have literally the cheapest PCBs I have ever seen in a keyboard. I do believe that one did not work originally ... because some of the diode legs were literally rusted through, but I imagine that doesn't count? (lol) I have a Nan Tan keyboard from around the same period that literally feels like I could snap it in half with my bare hands. The screws on the bottom are rusted ... but it works perfectly.

I don't put much stock in anything somebody says when they recommend a brand, as opposed to a specific technology (capacitive buckling spring, hall effect, beam spring, etc), if that recommendation also includes some arbitrary claim about longevity vs some other brand using the exact same sensing technology.

For what type of boards?  We know that early on there were some severe issues with soldering for some mechanical keyboard brands...there seem to be fewer problems now than there were but you still rely on soldering for many of these boards which means some will be done to a better standard than the other.

Comparing the technology and saying things like like MX boards are the same because they use the same type of switches is pretty ignorant given the past data...although there might be a point where that statement is true.

What is it, specifically, that's ignorant about my statements? I believe what you're talking about stems from a comment chain with an example of Leopold vs Keychron, saying that The Leopold will last longer based solely on the brand's reputation when both boards are most likely using the same switch technology. Even if we were talking about the entire history of random keyboards pitted against each other in this regard, have you seen any keyboards with failed solder joints that were not the result of physical abuse? I haven't yet, even on really cheap boards.

I think a cold joint, etc, would be a pretty wild exception to the general rule. Bad grounding and/or junk controllers or something might make more sense to me, and those would still be exceptions in my opinion.

The switch is not the source of failure in most cases...so stating they (different brands) use the same switch therefore will have the same reliability doesn't stand up to the experience people in the hobby have had to date.

And yes, people have simply had their keyboards fail without abuse..poor soldering is a big reason.  Just because you haven't seen it yourself, frankly, means little to nothing because factually, it has happened.  You see so much less of it now than you did maybe 6-7+ years ago... Quality is far better, on average, nowadays...probably because the cost of decent QC is far lower...

[Maledicted]

A related question off this then would be: Are higher end boards longer lasting than budget boards? I recall a thread from weeks ago when someone was asking Keychron vs Leopold, and some answers were 'Leopold, better build, will last longer.

Personally, I'd expect longevity to be similar, minus factors guaranteed to degrade with time such as built-in batteries.

I still prefer higher quality builds for the tactile satisfaction, they simply feel better to hold.

I'm no engineer, not even an electrician, but short of abuse, what usually fails in any relatively small electronic? Electrolytic Capacitors? How many capacitors do you see in modern keyboards? I have 2 Unitek K151L boards from the late 80s that still work perfectly. Those things have literally the cheapest PCBs I have ever seen in a keyboard. I do believe that one did not work originally ... because some of the diode legs were literally rusted through, but I imagine that doesn't count? (lol) I have a Nan Tan keyboard from around the same period that literally feels like I could snap it in half with my bare hands. The screws on the bottom are rusted ... but it works perfectly.

I don't put much stock in anything somebody says when they recommend a brand, as opposed to a specific technology (capacitive buckling spring, hall effect, beam spring, etc), if that recommendation also includes some arbitrary claim about longevity vs some other brand using the exact same sensing technology.

For what type of boards?  We know that early on there were some severe issues with soldering for some mechanical keyboard brands...there seem to be fewer problems now than there were but you still rely on soldering for many of these boards which means some will be done to a better standard than the other.

Comparing the technology and saying things like like MX boards are the same because they use the same type of switches is pretty ignorant given the past data...although there might be a point where that statement is true.

What is it, specifically, that's ignorant about my statements? I believe what you're talking about stems from a comment chain with an example of Leopold vs Keychron, saying that The Leopold will last longer based solely on the brand's reputation when both boards are most likely using the same switch technology. Even if we were talking about the entire history of random keyboards pitted against each other in this regard, have you seen any keyboards with failed solder joints that were not the result of physical abuse? I haven't yet, even on really cheap boards.

I think a cold joint, etc, would be a pretty wild exception to the general rule. Bad grounding and/or junk controllers or something might make more sense to me, and those would still be exceptions in my opinion.

The switch is not the source of failure in most cases...so stating they (different brands) use the same switch therefore will have the same reliability doesn't stand up to the experience people in the hobby have had to date.

And yes, people have simply had their keyboards fail without abuse..poor soldering is a big reason.  Just because you haven't seen it yourself, frankly, means little to nothing because factually, it has happened.  You see so much less of it now than you did maybe 6-7+ years ago... Quality is far better, on average, nowadays...probably because the cost of decent QC is far lower...

I have piles of boards ranging in age from almost 40 years old to brand new, from IBM down to Red Dragon, etc. All of the boards I have that I can test with a modern computer have worked perfectly outside of abuse. Some specific low-end brands may have a higher failure rate than the likes of Filco and Leopold, it stands to reason that most do not. Just about anything has happened in the past, regardless of who made the board.

[gipetto]

I've had new dirt cheap 3 euro rubber domes die in under a year. before that i tended to use second hand gifted rubber domes from neighbours who upgraded. the second hand ones were better quality but they all lost the use of one key. the 3 euro rubber domes either lost one key or broke the usb cable. no i wasn't abusing it but I do move my keebs around a lot.
I bought a ducky zero shine as my first mech keyboard and the h key would be intermittent after 6 months. I would hammer on it with my finger to bring it back to life. eventually i got fed up with the controller and decided to handwire it. when i inspected the key i noticed bad corrosion on the contact, the rest of the keys were shining. I'm not sure whether it was a spill or a mixed batch of switches from the factory. i was getting fed up of mx blacks in any event.

[Polymer]

I have piles of boards ranging in age from almost 40 years old to brand new, from IBM down to Red Dragon, etc. All of the boards I have that I can test with a modern computer have worked perfectly outside of abuse. Some specific low-end brands may have a higher failure rate than the likes of Filco and Leopold, it stands to reason that most do not. Just about anything has happened in the past, regardless of who made the board.

There have been numerous brands, some good, some bad...the lesser ones had lesser quality and lesser care put into them..they had a higher failure rate.  What we see today is, in general, better quality control..probably because the cost basis is better.  Better scale, better understanding of what to do, better understanding of where they fail, etc, etc....and yes, buying X low end board today is probably less risky than it was in the past.  But none of the validates your statement.  In fact, by your own statements, you've proven yourself wrong. 

Just because they use the same switch does not mean the reliability is equivalent....There is more to the reliability of board than the switch itself and that has been a fact for awhile.   As I said before, the gap is smaller and the number of quality producers is growing but so what?  You still try to find a reasonably good quality producer and that still remains.  To boil it down to they both use the same switch so therefore the reliability is the same remains wrong. 

[Maledicted]

I have piles of boards ranging in age from almost 40 years old to brand new, from IBM down to Red Dragon, etc. All of the boards I have that I can test with a modern computer have worked perfectly outside of abuse. Some specific low-end brands may have a higher failure rate than the likes of Filco and Leopold, it stands to reason that most do not. Just about anything has happened in the past, regardless of who made the board.

There have been numerous brands, some good, some bad...the lesser ones had lesser quality and lesser care put into them..they had a higher failure rate.  What we see today is, in general, better quality control..probably because the cost basis is better.  Better scale, better understanding of what to do, better understanding of where they fail, etc, etc....and yes, buying X low end board today is probably less risky than it was in the past.  But none of the validates your statement.  In fact, by your own statements, you've proven yourself wrong. 

Just because they use the same switch does not mean the reliability is equivalent....There is more to the reliability of board than the switch itself and that has been a fact for awhile.   As I said before, the gap is smaller and the number of quality producers is growing but so what?  You still try to find a reasonably good quality producer and that still remains.  To boil it down to they both use the same switch so therefore the reliability is the same remains wrong.

That's not what I said, so I'm not sure who you're arguing against here. The point was just that brand A vs brand B is less important in terms of reliability/longevity than switch technology is, since we know something capacitive, hall effect, etc, will inevitably outlast most other common switch technologies. The second component to that is that people make way too big of a deal out of brand loyalty based on the anecdotal, no actual evidence at all, and/or group think. With the (most likely) hypothetical example above of Leopold vs Keychron, what are the odds that the suggestion of Leopold over Keychron is based even on the commenter's use of even 1 Leopold vs 1 Keychron keyboard used for years, to the point of failure? I would say they're very low, and even if they were, what sort of useful data is that? Not very. Leopold is an established name, Keychron is not as well established, therefore the Leopold user's crystal ball has determined that a Leopold will last longer. This is what I'm arguing against, yet you say that I have proven myself wrong ... somehow.

Here's a simple example: The IBM Model M2 was pretty much garbage in terms of long-term reliability. I imagine we can all agree on that, right? Most of their other mechanical designs were fantastic (though I also dislike the Model M in general). Are we going to say that all IBM boards suck, even based on the thousands of those vs random anecdotal accounts commonly used today to arbitrarily promote or disparage a brand? All brands have RMA divisions. I think I remember reading on this very website about some Realforce boards, of all things, that may have had warped molds to the point that they didn't sit straight on a desk.

I don't think that brand loyalty, in any guise, makes any sense.

[Polymer]

That's not what I said, so I'm not sure who you're arguing against here. The point was just that brand A vs brand B is less important in terms of reliability/longevity than switch technology is, since we know something capacitive, hall effect, etc, will inevitably outlast most other common switch technologies. The second component to that is that people make way too big of a deal out of brand loyalty based on the anecdotal, no actual evidence at all, and/or group think. With the (most likely) hypothetical example above of Leopold vs Keychron, what are the odds that the suggestion of Leopold over Keychron is based even on the commenter's use of even 1 Leopold vs 1 Keychron keyboard used for years, to the point of failure? I would say they're very low, and even if they were, what sort of useful data is that? Not very. Leopold is an established name, Keychron is not as well established, therefore the Leopold user's crystal ball has determined that a Leopold will last longer. This is what I'm arguing against, yet you say that I have proven myself wrong ... somehow.

And yet you're arguing the switch technology is what determines the reliability....maybe not 100% the same argument but really it is.

Secondly, I do agree, some of what our evidence is, is anecdotal..but not by one person...just the number of different problems we saw as a community.  Maybe you weren't here but there were a number of detailed tear downs quite awhile ago..which showed the poor quality of some brands...That's still somewhat anecdotal but it is hard evidence, at least with the boards that were examined.  But isn't all that (people's experience plus looking under the hood) how we generally come up with reliability?  Not a perfect example but good enough...

Here's a simple example: The IBM Model M2 was pretty much garbage in terms of long-term reliability. I imagine we can all agree on that, right? Most of their other mechanical designs were fantastic (though I also dislike the Model M in general). Are we going to say that all IBM boards suck, even based on the thousands of those vs random anecdotal accounts commonly used today to arbitrarily promote or disparage a brand? All brands have RMA divisions. I think I remember reading on this very website about some Realforce boards, of all things, that may have had warped molds to the point that they didn't sit straight on a desk.

I don't think that brand loyalty, in any guise, makes any sense.

Brands aren't blindly followed, is that the argument you're trying to make?  There are plenty of examples of good keyboards from a brand and some bad ones from the same brand..Coolermaster was a good example of that...and you know, the realforce example really shows you don't know what you're talking about.  The molds were NEVER warped.  What it is is that the case is clipped together so you can actually twist the board itself somewhat.  So yes, some didn't seem straight but you just had to move the case a bit and it would sit straight.  That bothered some, but really it is just how the case is put together...they're one of the best built most reliable keyboards out there. 

[nathanchere]

I thought denouncing was a firmware level issue. Is there a way this is influenced by PCB design?

When they don't provide any firmware updates *ever* (Das Keyboard 4 Professional) I consider it one and the same, though technically you're right it's probably more accurately described as hopeless firmware.

Do you mean a few of your keys are chattering? Or are all of them?

For science...

Here is an example of text typed on a Tofu keyboard with Box Jades, presented without any corrections. I chose this for comparison as it's the only other clicky board I have and want to minimise the impact different switch feel has on the results.

Her is ann example of the same txt typeed ono the Das keybooard witth Cherrey Blues. also presnted witthout any correctttions. I think the reesults prety cleearly speak for themeslves.

[Maledicted]

That's not what I said, so I'm not sure who you're arguing against here. The point was just that brand A vs brand B is less important in terms of reliability/longevity than switch technology is, since we know something capacitive, hall effect, etc, will inevitably outlast most other common switch technologies. The second component to that is that people make way too big of a deal out of brand loyalty based on the anecdotal, no actual evidence at all, and/or group think. With the (most likely) hypothetical example above of Leopold vs Keychron, what are the odds that the suggestion of Leopold over Keychron is based even on the commenter's use of even 1 Leopold vs 1 Keychron keyboard used for years, to the point of failure? I would say they're very low, and even if they were, what sort of useful data is that? Not very. Leopold is an established name, Keychron is not as well established, therefore the Leopold user's crystal ball has determined that a Leopold will last longer. This is what I'm arguing against, yet you say that I have proven myself wrong ... somehow.

And yet you're arguing the switch technology is what determines the reliability....maybe not 100% the same argument but really it is.

Secondly, I do agree, some of what our evidence is, is anecdotal..but not by one person...just the number of different problems we saw as a community.  Maybe you weren't here but there were a number of detailed tear downs quite awhile ago..which showed the poor quality of some brands...That's still somewhat anecdotal but it is hard evidence, at least with the boards that were examined.  But isn't all that (people's experience plus looking under the hood) how we generally come up with reliability?  Not a perfect example but good enough...

Respectfully, are we going to get to a point in this conversation where you refute anything I have said? Switch technology is the biggest factor, unless there's some indication that one brand's quality is severely lacking vs another. I don't care what the guts look like if the board is going to continue to work for 10+ years. Those Uniteks I mentioned look horrible inside. I have never seen a cheaper PCB in a keyboard. Alas, that hasn't hampered their function in the last 30+ years. I imagine the Nan Tan looks terrible too.

Here's a simple example: The IBM Model M2 was pretty much garbage in terms of long-term reliability. I imagine we can all agree on that, right? Most of their other mechanical designs were fantastic (though I also dislike the Model M in general). Are we going to say that all IBM boards suck, even based on the thousands of those vs random anecdotal accounts commonly used today to arbitrarily promote or disparage a brand? All brands have RMA divisions. I think I remember reading on this very website about some Realforce boards, of all things, that may have had warped molds to the point that they didn't sit straight on a desk.

I don't think that brand loyalty, in any guise, makes any sense.

Brands aren't blindly followed, is that the argument you're trying to make?  There are plenty of examples of good keyboards from a brand and some bad ones from the same brand..Coolermaster was a good example of that...and you know, the realforce example really shows you don't know what you're talking about.  The molds were NEVER warped.  What it is is that the case is clipped together so you can actually twist the board itself somewhat.  So yes, some didn't seem straight but you just had to move the case a bit and it would sit straight.  That bothered some, but really it is just how the case is put together...they're one of the best built most reliable keyboards out there.

Brands definitely are blindly followed. They're blindly followed by society in general as well as within individual communities. If brands were not blindly followed, Apple either would not exist, their products would cost half as much, and/or their ecosystem would actually be user-friendly, etc.

Yeah, I haven't looked into the distinct specifics of every single aspect of Realforce production. I just know I saw a thread a few months back, which had comments from 4-5 long-term Realforce users, none of which mentioned what you have said, in regard to their misaligned, warped, or whatever else, cases. I'm already seeing brand loyalty right here because in one breathe you talk about there being no brand loyalty and in another you give Topre an out for an obvious (if unimportant) design flaw. I'm not saying that Realforce boards suck. I don't even particularly like tactile switches, but I have an 87u and can appreciate its qualities. There is no perfect keyboard, that was the obvious point I was making with that example.

You're the same guy that wrote essays trying to argue with me that everybody hates MX brown because of how they feel when pressed slowly, not because of the inherent characteristics of the switch. You're obviously a smart guy, but you deal too much in absolutes. There are very few absolutes in any situation.

I don't want to hijack this thread, so if you would like to continue, PM me.

[Polymer]

Respectfully, are we going to get to a point in this conversation where you refute anything I have said? Switch technology is the biggest factor, unless there's some indication that one brand's quality is severely lacking vs another. I don't care what the guts look like if the board is going to continue to work for 10+ years. Those Uniteks I mentioned look horrible inside. I have never seen a cheaper PCB in a keyboard. Alas, that hasn't hampered their function in the last 30+ years. I imagine the Nan Tan looks terrible too.

Or we're measuring different things.  You can only just look in this thread.  How often is the reason for a problem the switch itself?  How many cherry mx switches have you see fail?  I think most people have seen very few relative to the number of problems a keyboard has.  But yet we know some manufacturers are more reliable than others.  Why is that?
Now if you're just saying yes there are difference in keyboards with different brands (But then why the whole thing about brand loyalty etc, that makes no sense) but that an MX switch vs. Alps is a bigger factor in reliability..sure...ok..but then everything else you said about PCBs, etc, quality, brands, etc, is pretty irrelevant...Not that I'm saying every keyboard from a brand is great..we know they can have different models with varying degrees of quality.  Again, Coolermaster was a perfect example of that.  Same switches, some of their versions were great, some were terrible. 

Brands definitely are blindly followed. They're blindly followed by society in general as well as within individual communities. If brands were not blindly followed, Apple either would not exist, their products would cost half as much, and/or their ecosystem would actually be user-friendly, etc.

Are we talking keyboards or other stuff?  I'm talking keyboards.  I'm not seeing blind following of any brands in keyboards...I have seen loyalty when it comes to a particular board.  But are people blind buying boards from a particular keyboard brand?  Doesn't seem like it...doesn't seem like any of them have that level of consistent brand appeal. 

Yeah, I haven't looked into the distinct specifics of every single aspect of Realforce production. I just know I saw a thread a few months back, which had comments from 4-5 long-term Realforce users, none of which mentioned what you have said, in regard to their misaligned, warped, or whatever else, cases. I'm already seeing brand loyalty right here because in one breathe you talk about there being no brand loyalty and in another you give Topre an out for an obvious (if unimportant) design flaw. I'm not saying that Realforce boards suck. I don't even particularly like tactile switches, but I have an 87u and can appreciate its qualities. There is no perfect keyboard, that was the obvious point I was making with that example.

You haven't looked into it because you don't know.  It is how the case is put together, it allows you to twist the insides as parts slide within the case.  Obviously I'm talking about something I know about and the fix is simply, twist it back or place all four corners on the desk and pretty down.   It has been discussed many many many times over the years.  Is the case twisting a flaw?  To some it is...to some it isn't.  In what way is this a flaw?   The board allows for a certain amount of give, for whatever reason.  What's the flaw in that?  Flex can be both a good and a bad thing depending on someone's point of view.  But given you actually didn't understand the issue here, you're speaking from inexperience.  I'm just calling it out because you're speaking as if you knew what you were talking about when it was obvious, you didn't.  If somehow you're claiming that's brand loyalty, I call complete BS.  I'm calling it exactly how I see it.  If something is wrong, I have no issues calling it out.  If it is just wrong to some people, that's for them to decide.  But to show you my point.  RF 87U.  What I dislike about it.  Wobbly bottom left corner key is a bit weird.  The text for some keys is really bad for the TKL cluster.  The ABS spacebar sucks.  I used to hate the look of the case but somehow it has grown on me.  But when people think the case is warped or the keyboard is uneven, they just aren't educated on the keyboard itself.  For whatever reason the keyboard had been twisted in shipping or handling and requires you to twist it back..its not actually bent metal but how the pieces slide within the case itself.  Pretty easy to fix.  When I first saw it, I thought it was pretty lame but seeing the insides and how it is put together, I actually really appreciate the design.  I've also found that I didn't mind that it flexed a bit..and it has probably helped, if anything, the few times it has been banged around.

You're the same guy that wrote essays trying to argue with me that everybody hates MX brown because of how they feel when pressed slowly, not because of the inherent characteristics of the switch. You're obviously a smart guy, but you deal too much in absolutes. There are very few absolutes in any situation.

I don't want to hijack this thread, so if you would like to continue, PM me.

Actually that wasn't my argument.  I DID NOT say everybody hates browns because of that.  Maybe that's just how you're reading it...

[funkmon]

To get back to the point of the thread, I have a Dell Quietkey that I've used on and off now for 25 years, mostly off, but I do occasionally whip it out still. I have an ancient Dell Bigfoot that's fine as well. These are the only long term use keyboard I have that haven't developed issues.

My Model M developed some kind of membrane problem over the years and I had to send it into Unicomp.

My scissor switch keyboards have died due to the scissory parts going bad. The S keys used to always die on the cheap replacements I bought between 2004 and 2010 and I don't know why. I have probably had 10 rubber dome keyboards go on me over the years for no obviously discernible reasons.

I don't, however, consider it to be a rubber dome problem, but a cost problem. I would simply buy the cheapest keyboards I could find and they all sucked butts. All the expensive ish or OEM rubber domes I've had prior to 2010 still work and are fine. The keyboards my parents have used on the other hand in the past 10 years, they've all died and been replaced with $30 Logitechs. My dad swore by Logitech, and I understand why. They make a good reliable product for the guys who don't care that much about keyboards, and for some even who do. Every Logitech he bought still functions. He used to just buy new ones because he liked buying stuff.

Anyway, so since basically the least expensive mechanical keyboard you can buy still costs $40, you're getting way way better build quality than the $2 rubber domes people go with.

I don't think that a Leopold is going to last longer than a $50 board in any practical reality, but when we start to get into vintage keyboards, we can see the build quality really play a factor.

I use a keyboard made in 1981 on the regular. That's 40 years old. I hope to still be using it in another 40 years. Genuinely. For most of the guys asking "hey should I buy a mechanical keyboard?" they're not considering what they're going to use when they're retired, so again, I don't think there's a huge practical benefit to them. But for me, there is.

I want a board with longevity, heritage, mainstream switches, and excellent service options, because I want to use the board when I'm 50. That's part of the reason I use Models M. I'm always going to be able to get those keyboards and as long as Unicomp stays in business another 10 years, I'll probably still be able to get them in good working condition until I die.