Azio Levetron Clicker Mechanical Keyboard Questions

[ruhtraeel]

Hi,
Does anyone own this keyboard? It is really cheap, 30 dollars after a 10 dollar MIR on NCIX and it's mechanical. One thing I'm wondering is that right now since I'm using a Compaq SK-2860, would it be an improvement? I know that even though my Compaq keyboard right now is PS/2, it is not NKRO. It is maximum 6, but maximum 2 on certain areas on the keyboard; for example, I can press "a l s k d j" fine and it shows up on Microsoft's test, but "w a s" at the same time only registers 2. Is the Clicker keyboard like this, or is it 6 or 18 KRO no matter what region of the board is being pressed?


If it is of any relevance, I am a pretty avid gamer (mainly FPS's, MMORPG's and sometimes RTS's), and also a Computer Science student who programs on a regular basis.

[nullstring]

Hi,
Does anyone own this keyboard? It is really cheap, 30 dollars after a 10 dollar MIR on NCIX and it's mechanical. One thing I'm wondering is that right now since I'm using a Compaq SK-2860, would it be an improvement? I know that even though my Compaq keyboard right now is PS/2, it is not NKRO. It is maximum 6, but maximum 2 on certain areas on the keyboard; for example, I can press "a l s k d j" fine and it shows up on Microsoft's test, but "w a s" at the same time only registers 2. Is the Clicker keyboard like this, or is it 6 or 18 KRO no matter what region of the board is being pressed?

http://www.newegg.com/Product/Product.a ... 6823839014
Per this page, they apparently use XM switches?
http://deskthority.net/wiki/XM
It's apparently 6KRO, which is alot better than your 2KRO

I don't pay attention to newegg canada sales, but you might be better of getting a rosewill RK9000 when they go on sale for ~$64. People don't seem to be that happy with the XM switches.

[Daniel Beardsmore]

Where does that page say "XM"? It says APC BSW 055WH switches.

Xiang Min switches all have models starting with KSB. Also, Xiang Min only appear to have ever used white and green for the slider colour.

Granted, the APC BSW 055WH shown above is an awful lot like an XM internally, and helpfully like XM it's unbranded. There are other APC switches (search for "APC", near the bottom) that were actually branded APC and were much more similar to Simplified Alps Type III switches inside, and Type III was suggested to have been made by Strong Man. Strong Man went out of business some years ago — could be someone using the same tooling.

I've dropped Azio an e-mail to ask if they can shed some light on this.

[ruhtraeel]

So would anyone recommend this keyboard? Apparently they feel a lot like Cherry MX Blues. I tried my friend's Razer Blackwidow Ultimate, and it felt pretty good but I did the Microsoft NKRO test, and it seemed like the section for first person shooters (qwe,asd,zxc) was 6KRO but the right side with lkjh was only 2 or 3 key rollover. I tried typing the sentence "The quick brown fox jumps over the lazy dog" with both shift keys held down, and it didn't come out correctly, while my Compaq PS/2 keyboard came out properly.

However, I tried pressing "w a s" and "a s x" at the same time on my Compaq keyboard, and only 2 of them show up (this matters in games because I can't strafe diagonally while crouched). I'm thinking it's most likely because older keyboards had even key rollovers for each section, while newer gaming keyboards have prioritized certain sections over others.

I'm wondering if the Azio Levetron Clicker has this problem? Can someone with this keyboard try?

[Icarium]

Most people agree that NKRO isn't really necessary but more one of those hype thingies like gold contact audio cables.

[JBert]

I tried typing the sentence "The quick brown fox jumps over the lazy dog" with both shift keys held down, and it didn't come out correctly, while my Compaq PS/2 keyboard came out properly.

Off-topic: please stop spreading that joke. It's not a proper NKRO test, instead it just tests your controller firmware's shift key handling. You only use it to see who's a clueless newbie.

[Icarium]

Hm... if it doesn't work the board isn't NKRO. Seems an acceptable test to me.

[ruhtraeel]

I tried typing the sentence "The quick brown fox jumps over the lazy dog" with both shift keys held down, and it didn't come out correctly, while my Compaq PS/2 keyboard came out properly.

Off-topic: please stop spreading that joke. It's not a proper NKRO test, instead it just tests your controller firmware's shift key handling. You only use it to see who's a clueless newbie.

Fair enough, I had no idea.

However, why do people keep saying that PS/2 keyboards are NKRO? Mine is PS/2, and it clearly isn't, as the keys in an individual region is limited to 2.


Also, I did end up ordering it, as it actually affects gameplay in games when I can't strafe diagonally while crouched.

[JBert]

However, why do people keep saying that PS/2 keyboards are NKRO? Mine is PS/2, and it clearly isn't, as the keys in an individual region is limited to 2.

I guess it is confusing. NKRO at first depends on the keyboard technology used (simple membrane contacts, switches without diodes only allow 2KRO). If the technology has a limit, it won't get better than that when you use USB or PS/2.

When your keyboard does support NKRO, you run into interface-limitations (The page link here tells you that most USB keyboards impose an upper limit of 6KRO due to compatibility reasons. PS/2 never had a compatibility mode and allowed one to "stream" n keys to the computer). So if you have an "NKRO" keyboard connected through USB, likely only 6 keys can be pressed at the same time (yet that's quite a lot already). If you want more, you need to check the manual to see if it can work in PS/2 mode, then connecting it using a PS/2 adapter might mean that you can now have 10-20 keys.

[Icarium]

And those 6 keys for USB don't include modifiers. Those are a bonus.

Also note that you can cleverly shape the matrix to get better results with switches without diodes and membranes and so on. While they are technically still 2KRO you can often press many keys at once without problem. There is just no better term for it.

[ruhtraeel]

However, why do people keep saying that PS/2 keyboards are NKRO? Mine is PS/2, and it clearly isn't, as the keys in an individual region is limited to 2.

I guess it is confusing. NKRO at first depends on the keyboard technology used (simple membrane contacts, switches without diodes only allow 2KRO). If the technology has a limit, it won't get better than that when you use USB or PS/2.

When your keyboard does support NKRO, you run into interface-limitations (The page link here tells you that most USB keyboards impose an upper limit of 6KRO due to compatibility reasons. PS/2 never had a compatibility mode and allowed one to "stream" n keys to the computer). So if you have an "NKRO" keyboard connected through USB, likely only 6 keys can be pressed at the same time (yet that's quite a lot already). If you want more, you need to check the manual to see if it can work in PS/2 mode, then connecting it using a PS/2 adapter might mean that you can now have 10-20 keys.

So why exactly does my PS/2 keyboard not have close to NKRO, and 2KRO in some cases? From what I can understand, it's not a hardware issue, rather an issue in assembly language and how a keyboard's input stream is handled. Since a keyboard returns a series of bits, each with respective information and flags as to which key is being held down, overflow, etc, it makes sense to me that there are limitations in how Intel's assembly language is handled. For example, once you read in a key's pressed flag bit into a register like the high and low bits of each 32 bit register such as al and ah in eax and such, you only have so many registers you can store the state of each key being pressed before you need to start using your memory buffer. Since memory to memory operations don't work in assembly, it seems like once you have exhausted all your usable registers, there needs to be some way to limit how many bits go into your memory buffer before it is completely exhausted, or the keyboard interrupt is being called for the keyboard to still have efficient operation, thus having the region-specific key limitations?

So I guess it makes sense that the keyboard can "stream" N keys to the computer, but it's the computer that may not be able to handle it?

Or is there some way that Intel assembly gets around this and still manages to register every key?


I am purely drawing from my experiences in writing a mouse driver in x86 Intel assembly language, but I'm not too sure if keyboards work the same way.

[JBert]

You're overthinking this. Common PS/2 drivers have a buffer of 10-20 keys in memory. They likely get an interrupt from the PS/2 controller in the computer, read one key into a register and then immediately copy it into memory to let the Operating System deal with it, so the computer hardware is not limiting PS/2. Beware though, I'm not a driver programmer.


The problem with your keyboard is that its key detection method (matrix scanning) likely has ghosting issues (read our fine NKRO article in the wiki). Because "ghost" keypresses are bad, the keyboard's firmware blocks certain combinations from happening, which means that some key combinations of 3 random keys results in only 2 being pressed. It doesn't matter that other random key combinations go up to 6 or more, your keyboard is still only 2KRO as key roll-over is measured by the worst case scenario.


If you understand that, you can start looking at USB: legacy drivers or the BIOS only have a buffer of 6 keys + modifiers, if your keyboard wants to send more then those keys will not register.

[ruhtraeel]

So my Azio Levetron Clicker appeared in the mail today.

First thoughts:

Super super loud (Can be a problem for some people, I know my parents in the room next to me won't be thrilled about this)
Super satisfying to type on
European style enter key doesn't bother me much because my laptop keyboard is like that as well
Rubberized keys don't affect regular typing too much
ASX still only registers as AS... goddamnit. I'm starting to think that no keyboard will register those 3 keys. I hate using Ctrl to crouch in games because then I can't reach the Q key
Keyboard feels very light/plasticy, doesn't feel solid at all but that's what I expected from a $30 dollar keyboard

For 30 bucks including a 10 dollar rebate, I would say this is definitely worth it. Feels similar to my friend's Blackwidow, which cost well over 70 dollars.

[Daniel Beardsmore]

ASX still only registers as AS... what I expected from a $30 dollar keyboard

Y U NO UNDERSTAND ROLLOVER?

[ruhtraeel]

ASX still only registers as AS... what I expected from a $30 dollar keyboard

Y U NO UNDERSTAND ROLLOVER?

I'm not sure I understand you, more than rollover itself

[Daniel Beardsmore]

I'm not sure I understand you, more than rollover itself

Plenty of keyboards recognise A, S and X pressed together. They just cost more money, that's all. Mine has no issue with ASX, works perfectly. It's just a bog standard Filco.

Simultaneous keypresses is called "rollover" — yours is 2-key rollover, or 2KRO: it will only permit two keys to be pressed at once, not counting modifiers (shift, ctrl etc). Most keyboards are 2KRO — you'd have to study the internal keyboard circuitry to understand where this limit comes from. (I've since realised that my diagrams on the wiki don't make any sense at all. They're correct, just clear as mud. The whole page needs a complete rewrite.)

Good keyboards let you press far more keys at once, often as much as every key on the keyboard. A keyboard with no limit on simultaneous keypresses is called N-key rollover, or NKRO.

The way USB is generally used limits it to six keys plus modifiers, or 6KRO.