Slashdot Mirror
Cherry MX Mechanical Keyboard Switches Compared
crookedvulture writes "Keyboards with mechanical key switches are enjoying a renaissance of sorts. They're prized by gamers, coders, and writers alike, and Cherry's MX switches are the most popular on newer models. There are MX blue, brown, black, and red switches, each with a different tactile feel and audible note. This comparison of four otherwise identical Rosewill keyboards details how each switch type feels and sounds, complete with audio recordings of the various colors in action. Recommended reading for anyone considering a mechanical keyboard or one of the Rosewills, which cost about $100. Looks like the removable USB cord on these particular models is prone to breakage."
Sounds like an ad, but seriously I found the best advice there.
from the company that holds the original design patent from IBM for the Model M, and inhereted the process from Lexmark. The keyboard is still built and serviced in Lexington, Kentucky. They ship internationally. have a nice rennaissance.
http://www.pckeyboard.com/
Good people go to bed earlier.
Hands down the best set of information have ever seen about mechanical keyboards was this forum post on the topic it has everything you want to know about the subject.
You need one of these: PS/2 to USB Adaptor
Support Right To Repair Legislation.
PS/2 has unlimited simultaneous keypresses (n key rollover/NKRO). USB, depending on where you look, either cannot, or needs a workaround to do more than 6 keys and a certain number of modifiers, generally 4 (shift/alt/ctrl/super).
Keep on knockin'
https://robbiecrash.me
I have no idea what the hell any of this is about. Mechanical keyboards?
So you have to start them up like a chainsaw or something?
Are they like the old C-64/VT100 keyboards. Cause you needed kung-fu death fingers to type on those.
Sig. Sig. Sputnik
Accurate, and, on top of that, USB is polling based, while PS/2 is interrupt based. USB will check the port every "n" milliseconds to see if there's data waiting, while with a PS/2 keyboard, when you press a key, an interrupt is generated. The delay is very short (depending on how often a given keyboard's driver polls, it might only be 10 milliseconds or so), but it's worth noting.
PS/2 is still significantly better than USB for keyboard technology. But USB is usually good enough if you're not a gamer.
It is worth noting too, that just because you have PS/2 doesn't mean you have NKRO automatically, as depending on the technology 'underneath the hood' of a keyboard, it might not allow certain combinations of simultaneous keypresses. But that's an issue that's resolved by getting a mechanical keyboard, since they have the individual switches.
Shameless plug: Steelseries 7G. It's a big investment for a keyboard, but it provides NKRO, is a native PS/2 keyboard, is well constructed, and has audio jacks and USB ports on the keyboard, as well as a removable plastic wrist rest.
I have a DAS Keyboard with brown Cherry switches. They are 45g to activate, and not clicky. The keyboard is not silent, but it is certainly not noisy, either. This is the best feeling keyboard I own, and maybe the best feeling one I have ever owned. I have pounded the crap out of it for over a year and it still works, as one would expect. Every time I shop for a laptop, I just cringe when I feel the keys and I think about spending all day using it.
Friends help you move. Real friends help you move bodies.
Never forget: 2 + 2 = 5 for extremely large values of 2.
I'm a considerate coworker, and so when I brought my Loner to work I passed out icepicks, iodine solution and cotton balls to everyone within earshot.
Emacs users the world over disagree with you.
You can get foam "landing pads" which eliminate the bottoming-out clackiness of all the cherry-type keyboards. That makes all but the blue switches almost entirely silent.
http://elitekeyboards.com/products.php?sub=access,slpads
And why would anyone want to use PS/2 in a new design, anyway?
Some users of KVM switches or older computers may find it useful and I doubt there is any signficiant cost difference between a chip that can scan a keyboard and output over USB and one that can scan a keyboard and autodetects USB or PS2.
Also IIRC there is a design flaw in the USB keyboard specs which limits the number of keys that can register as down at the same time.
note: i'm known as plugwash most places but i screwd up registering that here somehow in the past and now can't register
It's a software (bios) issue. Many older motherboards kept the keyboard port active regardless of if one was plugged in on boot (although almost all of them disable the ps/2 mouse port if a mouse isn't detected on startup. Especially annoying with 'passive' vga/ps2 kvm switches, where you have to switch over and wait for bootup on any system you reboot/power on or else find out you need to reboot when you switch over later and the mouse isn't working). But if the ports are 'always on' in the bios, you can swap around with impunity.
Sadly the only way you're likely to get this is running an intentionally hacked bios, or a copy of coreboot, since I've never in 20+ years seen a PC that allows it (most assumed if a PS2 mouse wasn't hooked up that you'd manually configure a com-port based one, and thus disabled the PS2 mouse port so it wouldn't interfere, if I'm remembering correctly.) Not really an issue for modern systems, but like so much cruft in the bios it's still there because nobody got off their ass and fixed it.
The mechanical keyswitches are better to type on.
Anecdotal : my dear grey-haired mother used to be a legal secretary, so she would type most of the day for all of her professional career. When I saw that she was getting arthritis in her finger joints, as any good son would, I replaced the shitty generic standard-issue membrane keyboard that came with her office computer with a decent Cherry. Her arthritis improved significantly within weeks. (Both keyboards were non-ergonomic standard layout). When she retired, she took the keyboard with her and continues to use it even today.
The main problem with a membrane keyboard is a lack of positive feedback. You have to wait until a character appears on screen and your brain has processed this event to know whether you have successfully hit the key. For a fast typist, this is way to slow, so instead, they start to mash the keys harder than necessary in order to be sure of positive contact. Hence the finger joint arthritis.
With a mechanical keyswitch, there are two forms of positive feedback that you have successfully hit the key. First, there is the characteristic "click" sound. Secondly, there is the moment in the keystroke when the key switch "gives". This means that your finger can sense precisely how much effort is required to achieve positive contact, which means you only expend as much effort as necessary. In my humble opinion, ergonomic keyboards are a really stupid response to most typing RSI issues, and it's probably a better idea to get a keyboard with proper keyswitches - all the ergonomic boards I've touched still have the same shitty membrane switches.
The thing that got to me the most was her IT departments disgruntled response about having to install a new keyboard - "Why do you need a special keyboard?" (she shut them up by informing them her son was both a doctor and a computer professional). A professional typist (whether you're typing code, or legal papers) should use a professional keyboard. IBM understood this when they developed the Model M. It's a shame that we've lost that in the quest for cost savings..
Few years old, but still contains lots of interesting stuff
http://www.overclock.net/t/491752/mechanical-keyboard-guide
This isn't true of Maltrons -- I have a friend who's a huge fan, owning 3 generations of the product. According to Wikipedia, they use Cherry MX switches.
I can't believe no one mentioned WASD keyboards:
http://www.wasdkeyboards.com
You can fully customize the keyboard, from what is printed on each key (and what font is used) and what color the key is, to what type of switches they use.
They don't cost much more than other mechanical boards.
I highly recommend them!
It's a lurching zombie, damn it. Shambling is for scarecrows.
In my humble opinion, ergonomic keyboards are a really stupid response to most typing RSI issues, and it's probably a better idea to get a keyboard with proper keyswitches - all the ergonomic boards I've touched still have the same shitty membrane switches.
90% of the 'ergonomic' keyboards are only ergonomic because their marketing department decided to call it that. There's no actual testing, no actual thought on what would help the typist, they just copied what everyone else calls ergonomic.
If you are going to get an ergonomic keyboard, look for the ones where they actually thought about it - they'll cost more, but they'll actually be good as well. Personally, I use a Kinesis, and it's very good. They also all have mechanical switches - some models even offer a choice of which switch you want.
On the other hand, you'll never find them in a store...
'Sensible' is a curse word.
Actually, USB HID devices (keyboards, mice, game controllers, etc) use the USB "INTERRUPT" transport mode. This mode allows the USB device to specify a scheduled update rate at which it will send button/etc updates without having to wait for a driver or application to request updates from it. This isn't *exactly* interrupt timing straight from the key/button press, but it can be close. The USB driver for the HID devices/keyboard may be choosing to poll instead of utilize the USB data arrival interrupt, but that is likely to be OS specific.
You are correct that *other* USB devices are polling based (ones using the BULK and ISO transports) and must wait for a driver to request data from them before transmitting.
Furthermore, there is not an inherent 4 or 6 "key at a time" limit in the USB HID descriptor reports.
1) The modifier keys are sent as a bitfield separate than the keyboard keys, so they wouldn't count toward the max number of keys pressed at a time. :
USB HID 1.1 spec :
"Since only one array element can be reported in each array field, modifier keys
should be reported as bitmap data (a group of 1-bit variable fields)."
2) While a given keyboard may choose to limit its HID report to only transmit an array of a few keys at a time, it is possible to transmit a much larger number of keys at a time (report sizes greater than 64 bytes can be spread over multiple USB transactions if desired).
HID Usage Tables 1.1 spec :
"N selections of a set. More than one selection (button) can be valid at a time. Multiple selections
can be returned to the system at one time in a multi-byte array. The “n selections of a set” form is
defined by a Main item with the Array flag set and the Report Count set to n, where n is the
number of selections that can be reported in a single report. An example is a keyboard."
HID 1.1 spec again :
Rather than returning a single bit for each button in the group, an array returns an index in each field that corresponds to the pressed button (like keyboard scan codes). An out-of range value in and array field is considered no controls asserted. Buttons or keys in an array that are simultaneously pressed need to be reported in multiple fields. Therefore, the number of fields in an array input item (Report Count) dictates the maximum number of simultaneous controls that can be reported. A keyboard could report up to three simultaneous keys using an array with three 8-bit fields (Report Size = 8, Report Count = 3). Logical Minimum specifies the lowest index value returned by the array and Logical Maximum specifies the largest. The number of elements in the
array can be deduced by examining the difference between Logical Minimum and Logical Maximum
(number of elements = Logical Maximum - Logical Minimum + 1).