Can anyone confirm ABS has fixed the nkey issue?

[Special K]

I have recently purchased one and I tried the link to a keyboard testing program here in the forums but I didn't really know what to look for?

How many keys can you press simultaneously and have them all register with the test program?  The manufacturer claims 6 simultaneous keypresses are possible, so that is what you should be testing for.

As an example, try pressing QWAS at the same time in one of the test programs.  Are all 4 keys registered?

[1cewolf]

I just received the ABS M1 I ordered from Chiefvalue for $40 shipped. I figured it was a good deal at the time. I just tried the rollover test and it doesn't register anything when I try QWA. So much for the rollover issue being fixed!

The feel is quite nice, though; this is my first experience typing on a mechanical keyboard and I have to say I like it a lot. These fake black alps seem to be at a good level of pressure.

[Special K]

See the wiki entry - linky.

From a reviewer at Newegg:

If 6 keypresses was the max across 2 rows, then combinations such as QWAS should work, correct?  It has been said by several people in this thread that QWAS isn't recognized.

[Special K]

I sent a follow-up post to the ABS CSR asking them to try the simultaneous key press link posted in this thread.  Here is the reply I just received:

Dear Valued Customer,
 
Thank you for your feedback.
 
We truly apologize for our delayed response and the inconvenience this may have caused you.
 
As to the 6 simultaneous keypresses you have mentioned, we have tested it via the test link that you have provided, M1 could reach maximum 6simultaneous keypresses. However, M1 is not with N-key rollover function, we could not ensure that there will not have any ghosting keys in the game.  

I just sent them a reply asking which 6-key combinations they were able to register with their M1.  Maybe I can get to the bottom of this.

[Special K]

Ah, now I get it.

Keyboard manufacturers have decided it's cheaper to use the MAXIMUM number of keys registered in their specs.

Well, THAT's convenient.

The way I read their message is that they were able to get some unspecified 6-key combination of keys to register, but the keyboard most definitely doesn't support NKRO.  Of course we already knew that, given that it's a USB keyboard so I'm not sure why they mentioned it.  They never claimed it was NKRO.

[Special K]

But they did claim this.

Show Image

Defining "simultaneous keypress" should mean more than just pressing the keys and then nothing happening.

So do you think they lied to me in their response, or is there some 6-key combination on the keyboard that actually is registered?

[Special K]

Wow, check out the latest reply I received:

Dear Valued Customer,
 
Thank you for your feedback.
 
We truly apologize for the confusion. Our M1 does not have n-key rollover or anti-ghosting function. If you are player, we will not suggest you to purchase this model. The simultaneously 6-key press advertisement is incorrect. M1 could reach maximum 6simultaneous key presses. We have contacted our related department to modify it urgently.
 
Once again, we truly apologize for the unfavorable circumstance.  If you have already purchased it, please reply to this email with your Newegg/Chiefvalue purchase invoice number, we will help you to refund it.

I think they meant to say "M1 could NOT reach maximum 6 simultaneous key pressses", given the rest of the message.  I suspect English is not the first language of their customer service department.

In any case, it looks like they acknowledged that their marketing for this keyboard is incorrect.

Maybe they will fix the keyboard to match the marketing instead of the other way around :wink:

[1cewolf]

Wow, check out the latest reply I received:



I think they meant to say "M1 could NOT reach maximum 6 simultaneous key pressses", given the rest of the message.  I suspect English is not the first language of their customer service department.

In any case, it looks like they acknowledged that their marketing for this keyboard is incorrect.

Maybe they will fix the keyboard to match the marketing instead of the other way around :wink:

I doubt it. They would probably have said so if that was the case. I wish I could believe otherwise, but I'm a pessimist of the highest order when it comes to companies "doing the right thing" and I always expect the worst.

Honestly, aside from the rollover issues, this is the finest keyboard I've ever had. The build quality is phenomenal and it's a joy to type on. The small footprint makes it easy to fit in my cramped keyboard tray. There's even something almost endearing about the metallic ringing the keys make. I've tried using it in Call of Duty 4 and it does work well most of the time, but I've noticed occasional hiccups that concern me.

[Special K]

There's even something almost endearing about the metallic ringing the keys make.

Are the switches mounted to a metal backplate?  I have an AT101W and I also like the metallic sound the keys make when they hit the metal backplate the switches are mounted to.

I may still end up buying an M1 despite its known issues.  I've gamed on an AT101W with no issues, and AFAIK its rollover capabilities are no better than the M1.

Then again, if Filco would make one of their standard 104-key boards with NKRO and black alps, I would just buy it instead and that would be the end of it.

[1cewolf]

The guy who wrote the Tech Report review of the M1 states...

"I assume that sound is caused by the springs inside each key switch resonating and possibly interacting with the (likely metallic) back plate."

It sounds to me like the most noticeable ringing occurs when the keys come back up from being pressed.

I've gamed with non-NKRO boards my whole life without issue up until this point. Heck, I used a Microsoft Natural Ergonomic 4000 just fine for years before two drops of water killed it. What surprises me about this M1 is that I've seen it experience rollover issues in situations I'm sure my old 4000 wouldn't have. Then again, it could just be that I'm being overly sensitive since I'm admittedly a little miffed at ABS.

[Viett]

Edit: Accidentally posted twice.. can't figure out how to delete :/.

[Viett]

Anybody that has bought the board and returned it should really TRASH IT in the reviews.  I'd do it in the Newegg support forum but it's not really appropriate there.  Only way to keep the keyboard manufacturers honest.

Pros: Great clicky feel, heavy, and durable.

Cons: Very poor for gaming. The claim that it "Supports the use of maximum 6-key simultaneously" is very misleading, as it does not even support a minimum of three! Pressing W and two keys like AS or D causes the keyboard to lock up and not output anything at all. This is very disappointing for a gaming keyboard, which should be able to support pressing WASD (for gaming) simultaneously.

I guess I was pretty kind. I was pretty ticked, though. Either way, I'd take my Dell AT101W over an ABS M1 any day.

I've gamed on an AT101W with no issues, and AFAIK its rollover capabilities are no better than the M1.

Not true. Hitting some 3 keys all at the same time on the ABS would output absolutely nothing. The AT101W gives you at least 3 keys without blocking.

[Special K]

Not true. Hitting some 3 keys all at the same time on the ABS would output absolutely nothing. The AT101W gives you at least 3 keys without blocking.

I think you're partially correct.  If you press any 3 of QWAS on the AT101W, you will get 2 of the 3 keys.  It will not give you 3 simultaneous key presses in all instances.

Of course if the ABS M1 would output absolutely nothing in that case, then it is obviously worse than the AT101W.

Then again, wasn't the lockup/no output error with the M1 a separate issue that was already fixed?

[1cewolf]

My M1 registers nothing at all when I press QWA. Clearly, this is another ongoing issue with the M1. I think you're probably better off just sticking with your AT101w...and I'm probably better off returning this M1 to Newegg and picking up an AT101w of my own. There's a guy who's selling them on eBay NIB for $20 BIN/MAO with free shipping, so I could probably get one for $15-16 or so.

http://cgi.ebay.com/Dell-104-Key-PS2-6P-USA-Keyboard-NEW-227KN_W0QQitemZ370211755944QQcmdZViewItemQQptZPCA_Mice_Trackballs?hash=item56325717a8&_trksid=p3286.m20.l1116

[Special K]

My M1 registers nothing at all when I press QWA. Clearly, this is another ongoing issue with the M1. I think you're probably better off just sticking with your AT101w...and I'm probably better off returning this M1 to Newegg and picking up an AT101w of my own. There's a guy who's selling them on eBay NIB for $20 BIN/MAO with free shipping, so I could probably get one for $15-16 or so.

http://cgi.ebay.com/Dell-104-Key-PS2-6P-USA-Keyboard-NEW-227KN_W0QQitemZ370211755944QQcmdZViewItemQQptZPCA_Mice_Trackballs?hash=item56325717a8&_trksid=p3286.m20.l1116

Wait a minute - the description says 227KN, not AT101W.  The keyboards in the auction look identical to AT101Ws, however.  Are they really exactly the same?

I did a search for 227KN on here, but didn't find too much info.

[kyamei]

227KN is the part number for AT101Ws

[Special K]

227KN is the part number for AT101Ws

Does it say that anywhere on the keyboard?  Do the AT101Ws use any other part numbers?  I checked mine and didn't see 227KN anywhere on the board.  All I found was DP/N 077EUG.  I assume DP/N is Dell Part Number.

[kyamei]

I don't know if there are other part numbers, but here's a pic of my AT101W.

[itlnstln]

Too many posts in here that I missed, but FWIW, the ABS will allow some 6-key combos to work.  At least my 'boards do it.

[1cewolf]

I e-mailed Chief Value and told them that they, Newegg, and ABS should take down the misleading information on the M1 and refund my money. Here's what I got back:

Thank you for your Email, we appreciate your feedback.

We truly apologize for the troubles you have encountered. We at Chief Value take all criticism as constructive so that we can use any useful suggestions to better our business practices and ourselves.

We have forwarded your thoughts to our related department to review. Then we have found that the simultaneously 6-key press advertisement is incorrect.  M1 does not have n-key rollover or anti-ghosting function.  Actually, it is not a gaming keyboard. It only could reach maximum 6simultaneous key presses.  

We have contacted Newegg and ABS to modify this info on their website, it will be updated soon.

 As you are our valued customer, after confirming with our upper management, we informed that we would like to issue an RMA for refund and the restocking fee will be waived. And we will make an exception to cover the return shipping fee, however it is up to $8.00, so we suggest you ship it back with UPS Ground. Please kindly reply to us with the shipping receipt once you ship it out. We will issue credit to your original payment account.

We have created RMA # for replacement for you. Please have the RMA # [snippy!] as your reference. Please kindly note all original equipment, components, manuals, cables, documents, packaging must be returned with your item in order for ChiefValue.com to process your RMA. Missing items will incur further charges or less of a refund. Please clearly mark the RMA number on the outside of your shipping box and send all the RMA items to the below location.

Send your RMA's to:
[Snippy!]

Further more please kindly note that RMA number will be voided after 15 days.

Please let us know if you have any further concerns.

It looks like they're well aware of the problem now and taking steps to correct it. And so am I; I'll be dropping off my M1 at the UPS store when I go out today.

[Special K]

I e-mailed Chief Value and told them that they, Newegg, and ABS should take down the misleading information on the M1 and refund my money. Here's what I got back:



It looks like they're well aware of the problem now and taking steps to correct it. And so am I; I'll be dropping off my M1 at the UPS store when I go out today.

Their message seems to contradict itself:

Then we have found that the simultaneously 6-key press advertisement is incorrect. M1 does not have n-key rollover or anti-ghosting function. Actually, it is not a gaming keyboard. It only could reach maximum 6simultaneous key presses.

First they say the 6-key rollover advertisement is incorrect, then they say it could reach a maximum of 6 simultaneous key presses?

The overall wording of the message suggests they are aware of the problem, but their choice of words makes the message confusing.

[alpslover]

Keyboard manufacturers have decided it's cheaper to use the MAXIMUM number of keys registered in their specs.

yes, apparently when they say 'up to 6 simultaneous keypresses', they mean there exists at least one combination of 6 keys that the keyboard registers.  but when we hear 'up to 6 simultaneous keypresses', we interpret that to mean all combinations of keys up to 6.  (and in a way, that is also a maximum, because a keyboard that can register all combinations of 6 keypresses must inherently register all combinations of 5, 4, 3, and 2 keypresses.)

i think it's unlikely the marketing departments are going to change their tune, though.

i liked ibi's idea of mapping out the matrices of the more widely available keyboards and perhaps creating an app that will allow people to see exactly what key combinations will and won't work, but i don't know how feasible that is.

[alpslover]

First they say the 6-key rollover advertisement is incorrect, then they say it could reach a maximum of 6 simultaneous key presses?

The overall wording of the message suggests they are aware of the problem, but their choice of words makes the message confusing.

what they mean is, they're aware that not all combinations of 6 simultaneously depressed keys will work, but they have found some combinations that will.

[alpslover]

I've gamed with non-NKRO boards my whole life without issue up until this point. Heck, I used a Microsoft Natural Ergonomic 4000 just fine for years before two drops of water killed it. What surprises me about this M1 is that I've seen it experience rollover issues in situations I'm sure my old 4000 wouldn't have. Then again, it could just be that I'm being overly sensitive since I'm admittedly a little miffed at ABS.

no, it's probably not you.  which key combinations will and won't work will vary between different keyboards because they use different keyboard matrices.  so a particular key combination that works fine on your ms keyboard may very well fail on a different keyboard.

the way to avoid this problem entirely is to get a keyboard with n-key rollover - it's a guarantee that any combination of keys will be registered correctly.  but if you were misled by abs's marketing into thinking that the m1 could register any combination of up to 6 keys, then you have every right to be pissed.  they should offer refunds for this sort of crap.

[Special K]

I don't know if there are other part numbers, but here's a pic of my AT101W.

Show Image

My black AT101W says 077EUG where your white one says 0227KN.  I guess that's just the part number.

[Special K]

That's a pretty neat chart, ripster.  All keyboards marketed toward gamers should include one, assuming they don't support true NKRO.

Here's another bit of information I just received from ABS support:

Dear Valued Customer,
 
Thank you for your feedback.
 
We truly apologize for the inconvenience. It cannot support n-key rollover or anti-ghosting function. The simultaneously 6-key press description is incorrect. If key is in the same row, it could reach maximum 6simultaneous key presses.
 
Thank you so much for your patience and understanding. If you have any further questions or concerns, please feel free to contact us.

I bolded the key piece of information there.  Has anyone with an M1 verified whether you can do 6 simultaneous keypresses in the same row?

[kyamei]

I just got my M1 recently, so if there is a new revision of it, I'd have it.  

No, 6 keys from the same row doesn't work.  I get 4 keys max, the same as my non-rollover Filco w/ browns.  That doesn't really bother me though.  For $45 I didn't expect a Filco FKBN with black alps and a different logo.

[Special K]

Oooo...  This is getting better.  Now ABS is admitting the "error".  Some poor marketing guy is being called into his boss's office about now.  He's saying, "CURSE YOU GEEKHACK!"

Yeah, but now they are backtracking and saying that 6 keypresses in the same row will work.  Earlier they told me that 6 simultaneous keypresses wouldn't work, period.

[Binge]

I just recieved my ABS today and I am actually pretty happy with the construction and overall build.  I game pretty regularly and n-key is really nice.  After a bunch of testing I found a bunch of rollover issues.

The problems I found:

-The "Q" key when used with any number of keys from other rows beyond a total of 2 keys will produce nothing.

-The "W" key when used with any number of keys from other rows will produce the same results as the "Q" key.

-The "E" key when used with the "D" key in any arrangement of 3 or more keys will produce nothing.

-Most 4+ button combinations that use keys from separate rows will produce nothing.  The key combinations that do work for 4 keys are asdv qwef zxc  (that is zxc(SPACE)) 123r 234y 789p 890\ q890 q789 asdv asdb asdn... I'm not sure if you guys can see the pattern from this but for some reason keys directly beneath any given set of keys (from multiple rows) pressed simultaniously do not work.  If you were to map the keyboard into a grid then it's easy to see that for some reason there is a pattern.

-The three button combinations 345, 456, 567, and 678 return nothing when pressed simultaniously.  Oddly some combinations like 346 work while others like 457 do NOT work.  The number row is the only spot on the keyboard that exhibits this behavior.  The number pad on the otherhand is spotless.

I think I've covered everything I've found in the past 20 minutes.    Feel free to use this information for your benefit, geekhack.  I'm happy to contribute.

[Special K]

I just recieved my ABS today and I am actually pretty happy with the construction and overall build.  I game pretty regularly and n-key is really nice.  After a bunch of testing I found a bunch of rollover issues.

The problems I found:

-The "Q" key when used with any number of keys from other rows beyond a total of 2 keys will produce nothing.

-The "W" key when used with any number of keys from other rows will produce the same results as the "Q" key.

-The "E" key when used with the "D" key in any arrangement of 3 or more keys will produce nothing.

-Most 4+ button combinations that use keys from separate rows will produce nothing.  The key combinations that do work for 4 keys are asdv qwef zxc  (that is zxc(SPACE)) 123r 234y 789p 890\ q890 q789 asdv asdb asdn... I'm not sure if you guys can see the pattern from this but for some reason keys directly beneath any given set of keys (from multiple rows) pressed simultaniously do not work.  If you were to map the keyboard into a grid then it's easy to see that for some reason there is a pattern.

-The three button combinations 345, 456, 567, and 678 return nothing when pressed simultaniously.

I think I've covered everything I've found in the past 20 minutes.    Feel free to use this information for your benefit, geekhack.  I'm happy to contribute.

Did you try any 6-key combinations using keys from the same row?  If so, did they work?  The ABS CSR I spoke with claimed that there is apparently at least one 6-key combination that would work, assuming all 6 keys are from the same row.

I am awaiting the CSR's reply to find out exactly which 6 keys those are,  but if you uncover any through your testing, that would help too.

[Binge]

The following all work asdjkl asdkl; zxcm,. zxc,./ qweuio qweiop 789123 123890.

No other row wide combinations worked for me.

[outofideas]

I had M1 for about a week.  As far as I remember, it's keyboard matrix was nearly identical to a Model M.

[Special K]

Alright, here is yet another reply from the CSR:

Dear Valued Customer,
 
Thank you for your feedback.
 
It could reach maximum 6 simultaneous key presses that mean 6 is the maximum. We have tested it via http://random.xem.us/rollover.html that you have mentioned in your email. However, M1 does not have n-key rollover or anti-ghosting. We cannot press QWAS simultaneously. If press ASDFGH, most of the time, it works. M1 is not a gaming keyboard, we are sorry for this.
 
Thank you so much for your patience and understanding. If you have any further questions or concerns, please feel free to contact us.

Can anyone with an M1 get ASDFGH to work?  Why would it only work most of the time?

The more I learn, the further away I want to run from this keyboard.

[Binge]

M1 is not a gaming keyboard?  Let me check the box again, lol.  Special K, this is NOT a bad keyboard for $44.  I'm thinking of taking it apart and seeing if I can't fix these issues, that is unless there are blockers set up in the firmware.

[kyamei]

No, ASDFGH does not work.  

Yes, the M1 is full of marketing failures, but it's still a nice board for its price.

[Special K]

M1 is not a gaming keyboard?  Let me check the box again, lol.  Special K, this is NOT a bad keyboard for $44.  I'm thinking of taking it apart and seeing if I can't fix these issues, that is unless there are blockers set up in the firmware.

Yeah, but you can still get a NIB AT101W for $20 :wink:

[kyamei]

I don't know about everyone else's AT101W, but my M1 feels nothing like my AT101W.  

The feel of the Dell is inconsistent across the board.  For most of the keys, I cannot for the life of me feel the tactile bump.  All I feel is a linear action with friction, no matter how slowly I press the key.  However, there are a few keys which do have a noticeable tactile bump and have a bit less friction.  (My AT101W was purchased NIB)

The ABS M1 on the other hand has a much more consistent key feel, little to no friction, and a very noticeable tactile bump.  It's not very surprising that the M1 feels better since I thought white alps copies felt better than the real thing as well, but the difference here was HUGE.  The AT101W is my least favorite board in terms of feel, yet I find the M1 to be quite nice.

[o2dazone]

which is $30 after shipping. I'd pay $10 for usb and a little more desk real-estate. imo

[Binge]

Yeah, by no means am I saying it's the best board for the buck.  I just mean it's solid, a true black alps keyboard, and for typing I find it absolutely stellar.  I have gamed on it too and I don't find it horrible.  I'll have a board with cherry browns in a few weeks for my gaming rig, but I am going to keep my ABS for better or worse just because it's a good asthetically pleasing  standard layout keyboard with mechanical switches that responds well to the weight of my keystrokes.

[itlnstln]

I don't know about everyone else's AT101W, but my M1 feels nothing like my AT101W.
 
The feel of the Dell is inconsistent across the board. For most of the keys, I cannot for the life of me feel the tactile bump. All I feel is a linear action with friction, no matter how slowly I press the key. However, there are a few keys which do have a noticeable tactile bump and have a bit less friction. (My AT101W was purchased NIB)
 
The ABS M1 on the other hand has a much more consistent key feel, little to no friction, and a very noticeable tactile bump. It's not very surprising that the M1 feels better since I thought white alps copies felt better than the real thing as well, but the difference here was HUGE. The AT101W is my least favorite board in terms of feel, yet I find the M1 to be quite nice.

I haven't had these consistency problems with my Dells, but I do agree that the Dell and the ABS feel nothing alike.  The ABS feels more tactile, and, IMO, more satisfying.  Because of the added tactility, the ABSs keys feel a little heavier than the Dells' as well.  Overall, I really like the ABS.

[IBI]

Does it say that anywhere on the keyboard?  Do the AT101Ws use any other part numbers?  I checked mine and didn't see 227KN anywhere on the board.  All I found was DP/N 077EUG.  I assume DP/N is Dell Part Number.

227KN is the part number for the white AT101W, the other varients (without windows keys, different colour, different layout) will have other part numbers.

Has anyone with an M1 verified whether you can do 6 simultaneous keypresses in the same row?

From everything I've seen on here it looks like a standard matrix so you should be able to find some combinations of 6 keys in a row that work but which keys work together depends on where the traces run on the PCB. If anyone can take a clear photograph of the PCB we should be able to put together a picture without the more distracting details.

M1 is not a gaming keyboard?  Let me check the box again, lol.  Special K, this is NOT a bad keyboard for $44.  I'm thinking of taking it apart and seeing if I can't fix these issues, that is unless there are blockers set up in the firmware.

It's an inherant part of the design rather than something easily fixed, but you might be able to install diodes between every key which would solve the problem, details on here the matrix works and how you can fix it with diodes here.

The ABS M1 on the other hand has a much more consistent key feel, little to no friction, and a very noticeable tactile bump.  It's not very surprising that the M1 feels better since I thought white alps copies felt better than the real thing as well, but the difference here was HUGE.  The AT101W is my least favorite board in terms of feel, yet I find the M1 to be quite nice.

Which type of simplified alps does the M1 use? Has anyone taken a switch apart yet? I'd also be interested in knowing which spring it uses.

I've got a keyboard with clicky Type 1 simplified and I much prefer the complicated blacks in my AT102W but the M1 and Filco appear to be Costar rather than Strongman so maybe their type of simplfiied alps are nicer.

[kyamei]

Which type of simplified alps does the M1 use? Has anyone taken a switch apart yet? I'd also be interested in knowing which spring it uses.

I've got a keyboard with clicky Type 1 simplified and I much prefer the complicated blacks in my AT102W but the M1 and Filco appear to be Costar rather than Strongman so maybe their type of simplfiied alps are nicer.

I don't know if these are the kind of pictures you had in mind, but here's the internals of the M1's black alps

[keyb_gr]

It's an inherant part of the design rather than something easily fixed, but you might be able to install diodes between every key which would solve the problem, details on here the matrix works and how you can fix it with diodes here.

Hmm. I'd think that they already have the diodes in and the problems occurred due to buggy ghosting suppression or somesuch. Looks like that's a bit more involved with NKRO boards. I know that my G80-1xxxHAD/17 (3KRO, diode-equipped switches) can act a little funny with more than 3 keys at once. In fact, ASDFG also eventually results in in the rollover test! SF + GJ gives either SG or SJ.

We tend to forget the amount of know-how and fine-tuning that goes into these things until someone tries to reinvent the wheel and promptly comes up with something that doesn't work very well. Happens all the time.

[Special K]

But they did claim this.

Show Image

Defining "simultaneous 6-key press" should mean more than just pressing six keys and then some not registering.

Looks like they took that claim off of their M1 ad:

[o2dazone]

you twit bro?
I twote this morning

[itlnstln]

you twit bro?

I twa... wait...

[IBI]

I don't know if these are the kind of pictures you had in mind, but here's the internals of the M1's black alps

Show Image

Show Image

Great pictures, and very interesting. I wasn't expecting type 1 Alps.  They're the same type I've got in my new left-handed keyboard and my my playing around them seem fully compatible with complicated alps internally (except maybe the rubber dampers) so assuming there aren't multiple varients of type 1 there are some neat customisation opportunities availible for those who don't know how to solder. I wonder if the Filco Zero has the same Type 1 as the M1 is supposed to be from Costar as well.

Hmm. I'd think that they already have the diodes in and the problems occurred due to buggy ghosting suppression or somesuch. Looks like that's a bit more involved with NKRO boards.

I was assuming it was just the fault of the marketing department, has anyone said that it actually does have diodes? And do the Alps come with built in diodes like the MXs or are they attached to the PCB?

Does anyone have a Filco Zero with 'White XMs' and an ABS M1 that they could check for differences, as we know the Filco blocks on 7 keys?

[Special K]

And the price of the ABS M1 continues to drop - now only $39.99 with free shipping at Newegg:

http://www.newegg.com/Product/Product.aspx?Item=N82E16823224001&Tpk=abs%20m1

If it drops any more, I might have to just buy one regardless.

[Special K]

It's an inherant part of the design rather than something easily fixed, but you might be able to install diodes between every key which would solve the problem, details on here the matrix works and how you can fix it with diodes here.

Interesting article.  I'm tempted to try that mod on a cheap Dell rubber dome keyboard I have here.  If that works, maybe I'll try it on an M1.  It could be a fun project.  It seems straightforward in theory, but I'm not sure how much of NKRO is handled by circuit components vs. code in a microcontroller.  If it's the latter, then I'm pretty much SOL.

[keyb_gr]

The keyboard controller for a regular 2KRO board has to perform ghosting suppression. A key pattern that is invalid on these may well be valid on a board with diodes. Pressing all 4 keys in a 2x2 matrix segment would still be treated as a fault condition. For 3 it might work.

So you can try it, but if you're unlucky the board will have no better rollover capability than before. Best try it with a small part of the matrix only.