F104+SSK+122+62+77+50+Ergo orders now open! New Kishsaver+Industrial Model F Keyboards

NathanA

23 May 2022, 14:26

Ellipse wrote:
21 May 2022, 02:47
NathanA not sure if you saw my post last year about the foam (linked below) but the original foam used for this project was confirmed not the foam that IBM used, but the foam installed in the most recent batch of keyboards is one type of foam to my knowledge that IBM used. The spare foam I send out is still from that earlier, non-original batch and there is currently no option to order spares of the new material of foam.
I did miss that; thanks. Though from the sounds of it you ended up doing the opposite of what you originally thought you might do (have the factory continue to install the first type of foam into the keyboards but offer the more "authentic" type separately as spares).

I have seen pictures of rotted-out foam in Model F keyboards ranging from the original IBM PC one ("F XT") through the various 4704 models and terminal variants (F122 etc.) to the F AT...so virtually all of them. So at what point exactly did IBM change from this magic non-rotting foam type to the infamous type that seems to be the F's main Achilles' heel? Or did they interchangeably use both types of foam throughout the commercial lifespan of the F (simply because that just happened to be one part they were dual-sourcing or whatever, or perhaps one manufacturing plant used one type while another plant used a different one)?

andrewjoy

23 May 2022, 20:48

Hi everyone, my F77 arrived today and i have to say i am very impressed at the quality and attention to detail. After all the years waiting i think that it was 100% worth it.

I was wondering however, has anyone had any ideas on doing a foot to match that of the originals? My 107 has a plastic bar that clips into the holes on the underside of the case, some way to raise the keyboard up would be nice.

I will use this for a few days and give more detailed impressions, as it stands now this keyboard feels amazing it is similar in feel to the NIB replacement xt internals i have, much lighter key feel than the used Fs i am used to, the sound is a bit more pingy that my other 4704 107 but that's not a bad thing just different.

Overall an amazing projected , now i just need to get used to the HHKB layout and using an F again after about 2 years of using only topre boards.

MrRobot49

24 May 2022, 12:26

It would be extremely helpful to get some troubleshooting videos for how to undo the Via fiasco and how to get the solenoid working if you have dated firmware. ...and is there a definitive location for the latest firmware?

sedevidi

24 May 2022, 13:22

NathanA wrote:
22 May 2022, 11:22
sedevidi wrote:
21 May 2022, 11:15
Debian bullseye 11.3 (QT 5.15.2)
Your binary is attached.
Well, THANK YOU VERY MUCH!

In the meantime, my house was broken into, Thinkpads were stolen then retreived, but the F77 anvil did not even budge one cm...

The utility works wonderfully! The Signal Level Monitor does not reveal anything noteworthy, but the Keypress Monitor shows that some keys (like "I" or some on the numpad) simultaneously press other keys from the same row, not always at the same time. I wonder why this monitor shows such obvious things that I didn't even notice while using the keyboard daily.

This points me to some grounding issues. Time to open the anvil!

(the keyboard stopped responding while typing this, when the Utility was still open; sounds like the previous unresponsiveness when the Vial GUI was open; stopping the Utility then un/re-pluging the keyboard solved the issue; I didn't think about checking if the Utility was still getting things from the keyboard)

headphone_jack

24 May 2022, 14:52

I for one am very not pleased with my F77.

It arrived about a week ago, and frankly I am extremely disappointed. While the build quality is solid and nothing is *overtly* broken, there are so many strange things about it that just make me wish it was done better.

-It took over an hour to assemble. The keycaps that came with it (not Unicomp, Joe's original in-house ones) seem to have some serious trouble sitting on the spring. It took an hour just putting the caps on!
-For all the touting Joe does about his QC, the firmware on mine was buggy and riddled with strange issues. Some keys would only actuate on the upstroke, some would hang briefly, some worked intermittedly. For $400 plus two years is it really that unreasonable to expect a working product on arrival?
-The firmware took me several tries to reflash, on account of the fact neither QMK toolbox nor the capsense util tool recognized the keyboard the first few times around. This did fix the firmware, although it took way too long for the payoff.
-It just straight up doesn't feel amazing. I strongly prefer my FSSK, which is essentially a janked-together Model F, compared to the repros. I've heard people swap out the springs and barrels to make it feel better, but again, for a premium price I would expect a premium feel.

I struggle to comprehend how a product that takes two years to be delivered (sometimes even double or triple that for an unlucky few) supposedly for "rigorous QC" can be riddled with so many flaws. If this is truly a one-man operation, then it shows in the quality and user experience of this keyboard. It also speaks volumes that this thread (which is mainly troubleshooting) is by far one of the longest on this entire forum at over 250 pages. I'm selling my F77, if someone wants to skip the shipping wait.

andrewjoy

24 May 2022, 15:39

I had one or two keys that where a bit of a pain to seat correctly, but once they are on its fine , i have had the odd key like that on a real F but it was a little more annoying than usual, but again only one or two keys.

I have noticed a problem however, sometimes the shift keys ( the left shift in particular ) do not always register if you press them too hard. i have a feeling its the spring not seating properly on the keycap , i will swap in a standard M keycap to test.

Ellipse

24 May 2022, 17:39

MrRobot49 the firmware installation video is on the project web site, and it should cover what to do if the controller is stuck. The only extra step would be to short the PROG pad on the controller by holding the screwdriver before connecting the keyboard's USB cable to the computer.

andrewjoy glad your keyboard arrived safely. I'd imagine a foot bar can be 3d printed. I guess the posts would have to be realigned for the spacing on the new Model F Keyboards compared to the 4704 originals. https://www.shapeways.com/product/HDBG8 ... t-foot-bar

Regarding the shift key, I recommend adjusting the spring by following the QC secrets video in the manual on the project web site. As the last resort noted in the video, switching the spring with another spring or with a spare from the first aid kit can also fix the issue.

As noted before the main two issues why a key does not work is if the spring needs adjustment or if the two controller grounding screws were not screwed tightly enough from the factory.

headphone_jack my apologies you did not have a good experience with the keyboard.

Regarding the install time, did you read the manual and check out the installation videos showing how to hold the keyboard up vertically, space bar end up? It should not take an hour to install the keycaps.
Regarding the QMK firmware, the latest version is not buggy from what I have seen and read. It has been reliably working on the several thousand keyboards that have gone out so far, along with many original F107 and F122 keyboards and beam spring keyboards. If it was buggy I'd imagine there would have been some feedback by now that these keyboards were not usable by most folks. As noted above the main two issues why a key does not work is if the spring needs adjustment or if the two controller grounding screws were not screwed tightly enough from the factory. Unfortunately the keyboards are bounced around in shipping and the springs may shift their position or loosen from the nub and need to be re-seated.
Which capsense utility were you referring to? The original xwhatsit program "ibm_capsense_util"? That program is deprecated and does not work with QMK. If it was the pandrew utility, that one should recognize the keyboard unless it is running Via or the firmware has somehow been erased without being rewritten (this happens occasionally when I flash new firmware with Atmel Flip - it is not a problem as the keyboard just boots in bootloader mode so a new firmware can be flashed without having to reenter bootloader mode - you'd see ATMEGA32 DFU in the device manager in this case).

Jan Pospisil

24 May 2022, 19:25

Got my keyboard today, more impressions when I install keys and try it out in the following days.
Was sad to see a somewhat banana-like spacebar, here's hoping it works at least.

Has anyone tried anything for strain relief for the cable? Since the hole is bigger than the cable's diameter, perhaps some kind of plastic tunnel/spool with a split might work?
(obvs. something specifically designed and 3D printed would, but I was wondering if there was perhaps an existing little bit that'd work for this.)

User avatar
Sheepless

24 May 2022, 19:45

Jan Pospisil wrote:
24 May 2022, 19:25
Got my keyboard today, more impressions when I install keys and try it out in the following days.
Was sad to see a somewhat banana-like spacebar, here's hoping it works at least.

Has anyone tried anything for strain relief for the cable? Since the hole is bigger than the cable's diameter, perhaps some kind of plastic tunnel/spool with a split might work?
(obvs. something specifically designed and 3D printed would, but I was wondering if there was perhaps an existing little bit that'd work for this.)
Couple of ghetto suggestions (assuming you're talking about the classic case, where the cable clip doesn't do a good job of gripping the cable):

1. Cut some PVC sleeve from a scrap cable and wrap it around the USB cable where it passes through the clip.
2. Just tighten a cable tie around the USB cable behind the case hole.

Daniel_CH

24 May 2022, 21:27

headphone_jack wrote:
24 May 2022, 14:52
-For all the touting Joe does about his QC, the firmware on mine was buggy and riddled with strange issues. Some keys would only actuate on the upstroke, some would hang briefly, some worked intermittedly.
I've had that issue as well during the first couple hours of usage. These steps fixed it entirely (now over a week of usage with no recurrence):

- unplugging the keyboard for at least 15 minutes (while doing the next steps).
- removing and reinstalling the springs on the problematic keys (I recommend watching the video for correct procedure).
- retightening the controller screws (they were already somewhat tight, but I loosened them slightly before retightening them as hard as possible).

Not sure if one or all of these steps helped, but that was it.

Jan Pospisil

24 May 2022, 21:32

Sheepless wrote:
24 May 2022, 19:45
Jan Pospisil wrote:
24 May 2022, 19:25
Got my keyboard today, more impressions when I install keys and try it out in the following days.
Was sad to see a somewhat banana-like spacebar, here's hoping it works at least.

Has anyone tried anything for strain relief for the cable? Since the hole is bigger than the cable's diameter, perhaps some kind of plastic tunnel/spool with a split might work?
(obvs. something specifically designed and 3D printed would, but I was wondering if there was perhaps an existing little bit that'd work for this.)
Couple of ghetto suggestions (assuming you're talking about the classic case, where the cable clip doesn't do a good job of gripping the cable):

1. Cut some PVC sleeve from a scrap cable and wrap it around the USB cable where it passes through the clip.
2. Just tighten a cable tie around the USB cable behind the case hole.
I'm mostly annoyed by how it's rattling in the bigger hole.
So perhaps something like this? So it also looked neater?
https://www.thingiverse.com/thing:2939210

NathanA

24 May 2022, 23:26

sedevidi wrote:
24 May 2022, 13:22
In the meantime, my house was broken into, Thinkpads were stolen then retreived,
Oh my goodness, so sorry to hear that, but glad to hear you got (at least some of?) your stuff back. I imagine that's been a pain to deal with though! :cry:
sedevidi wrote:
24 May 2022, 13:22
The Signal Level Monitor does not reveal anything noteworthy, but the Keypress Monitor shows that some keys (like "I" or some on the numpad) simultaneously press other keys from the same row, not always at the same time. I wonder why this monitor shows such obvious things that I didn't even notice while using the keyboard daily.
I should have mentioned/warned you about this, but that seems to be "normal" behavior for the Keypress Monitor. I don't know why, but it does the same thing (occasionally and randomly show you pressing keys that you aren't pressing) on 2 perfectly-working F77s I've tried it on. I think this is a red herring. Whether it is a bug in the utility or what actually causes this, is unclear to me.

The more interesting things (IMO) to observe are whether the high (pressed) and low (unpressed) values for your Ctrl key in the Signal Level Monitor seem to differ notably on average from what the other keys show, and what the computed threshold value is for that Ctrl key is vs. the other keys on the Keypress Monitor (which is what that number is: the threshold that pandrew QMK capsense driver computed for that particular key).

andrewjoy

25 May 2022, 00:09

Ellipse wrote:
24 May 2022, 17:39
Regarding the shift key, I recommend adjusting the spring by following the QC secrets video in the manual on the project web site. As the last resort noted in the video, switching the spring with another spring or with a spare from the first aid kit can also fix the issue.
Ignore me i am an idiot , i forgot to install the stabiliser inserts!!!

I have reconditioned many model Fs in my time and just forgot as most of my model Fs use stabiliser bars on the wide keys no M style inserts, what a numpty.

NathanA

25 May 2022, 00:36

headphone_jack wrote:
24 May 2022, 14:52
-It took over an hour to assemble. The keycaps that came with it (not Unicomp, Joe's original in-house ones) seem to have some serious trouble sitting on the spring.
I can speak to at least the fact that this is not a keycaps issue. Though I have some Ellipse caps, I have been using Unicomp caps on my F77 for the time being while I wait for the rest of my custom caps from him to arrive. The Unicomps are just as fiddly to put on these keyboards, sometimes requiring a couple of tries (pull cap & try again) before they buckle correctly.

I don't have much experience with Model F boards outside of this F77 repro, but I can tell you that if what you are used to is installing caps on a Model M, it's a completely different experience, and I believe this comes down to both the springs themselves (likely related both to the different materials & perhaps the tighter wind) as well as the heavier flippers they are attached to, vs. both components on the M. On an M, the springs pretty much center themselves while the keyboard is laying flat on a desk. This is absolutely not true of the F (or at least these repros?), where the spring tends to want to fall towards the bottom side of the barrel when a key isn't in place (I think because the larger, heavier flipper is pulling it down!!). This is precisely why Ellipse's instructions tell you to pivot the keyboard 90 degrees with the back (cable-side) down on your desk while you install the caps, as this helps keep the springs centered in the barrel while you're putting the caps on. Even so, I found it can still take a couple of tries, while I had nearly 100% failure rate if I was not pivoting the keyboard back.
headphone_jack wrote:
24 May 2022, 14:52
-For all the touting Joe does about his QC, the firmware on mine was buggy and riddled with strange issues.
I've become pretty intimately familiar with the firmware at this point (or at least as familiar as someone who isn't actively working on its development can probably hope to be), and though I will be the first to tell you that the firmware is not without its faults, and that it's hard to argue against your experience where you re-flashed the keyboard firmware and it seemed to fix some of your problems, these really don't sound like firmware problems...or at least like any firmware issues I've ever either heard of or experienced. So I have a hard time not thinking that the blame is being falsely assigned here. And it's not like you got some unique copy or version of the firmware flashed to your keyboard "from the factory" compared to everyone else, either. It also appeared that you had already somehow decided that this was a "firmware problem" before you had even managed to fix it, and I'm wondering how you arrived at that original conclusion.

One potential theory that comes to mind is related to your next point where you describe your experiences with reflashing. You don't explicitly say it, but I wonder if at some point you opened up your keyboard to try to get access to the PROG pads on the controller board in order to kick it into bootloader mode? If so, that would have required you to unscrew the controller and flip it over, since the PROG pads are on the side opposite the one exposed when you take the bottom cover off of the keyboard. Which also means that you would have had to screw it back down again after you were done. Perhaps, as others have theorized, the controller simply wasn't screwed down properly to begin with, which in turn led to poor ground connection (controller is grounded to chassis through the exposed ground plane surrounding the screw holes), which would have caused matrix sensing problems. And you inadvertently fixed this when you went on your reflashing adventure.
headphone_jack wrote:
24 May 2022, 14:52
-The firmware took me several tries to reflash, on account of the fact neither QMK toolbox nor the capsense util tool recognized the keyboard the first few times around. This did fix the firmware, although it took way too long for the payoff.
I do agree that the documentation specifically surrounding the software and firmware is the least polished aspect of the project...I recognize that I have a tendency to excuse and/or overlook this at times, both likely because the reason(s) why things are in the state that they are is fairly clear to me (as I have been closely observing this project since the beginning and so have some historical context for the "whys" / how things organically developed in this area) and also since I am generally comfortable with spelunking in software & troubleshooting issues. I've been trying to help out where I can, though it is sometimes a bit demoralizing answering the same questions dozens of times either because people seemingly can't read or just can't be bothered to. (This is not a jab at anyone specifically / I have nobody particular in mind. Just an audible...*sigh* moment on my part.) Again, though, I acknowledge it's difficult when the information is so spread out in a hundred different unofficial places.

That said, it still kind of boggles my mind that it's not immediately "obvious" to almost everyone here (again, my myriad biases and blindspots shining through, almost surely) that "ibm_capsense" (original xwhatsit-designed-and-written) and QMK + pandrew's capsense support for QMK are wholly separate software projects that both just happen to support the same underlying hardware (much like you can freely choose to install either Windows or Linux on your PC, and Microsoft ain't responsible for making your Ubuntu Desktop work), that all keyboards that have been shipping after at least spring 2021 have come with QMK preloaded on them and not ibm_capsense/xwhatsit, and that "ibm_capsense_usb_util" is only applicable to the xwhatsit firmware and not QMK. So every time somebody receives a keyboard and posts "I tried to run the ibm_capsense util and it won't detect my keyboard", I die a little inside...
headphone_jack wrote:
24 May 2022, 14:52
I'm selling my F77, if someone wants to skip the shipping wait.
Well, you're really selling its virtues to your potential buyers. ;)

headphone_jack

25 May 2022, 01:35

NathanA wrote:
25 May 2022, 00:36
I don't have much experience with Model F boards outside of this F77 repro, but I can tell you that if what you are used to is installing caps on a Model M, it's a completely different experience
I've restored probably about a dozen Model Fs, way more than I've restored Model Ms. Never have I had a cap reseating take this long, or be quite so frustrating. Not mentioned was that twice removing a keycap somehow took the spring with it, which I originally attributed to error on my part but now that I know my reseating troubles were (most likely) not a cap-specific issue I blame poor spring tolerances. I ended up opening the assembly to replace the spring properly as well as to check for any other issues, which is actually one of the few advantages this model has to original Fs. The assembly was incredibly easy to open up, however this probably contributed to how much less "tight" these repros feel.
NathanA wrote:
25 May 2022, 00:36
I wonder if at some point you opened up your keyboard to try to get access to the PROG pads on the controller board in order to kick it into bootloader mode?
I never touched the screws on the controller, I already checked if what I was experiencing was a grounding issue and the PCB appears to be grounded properly. I just used the F77 util to put it in bootloader mode. It took many, many, many tries for the keyboard to be recognized in the first place. I have no idea why this was, but it seems to have been resolved my reflash.
NathanA wrote:
25 May 2022, 00:36
these really don't sound like firmware problems...or at least like any firmware issues I've ever either heard of or experienced. So I have a hard time not thinking that the blame is being falsely assigned here.
A reflash fixed every issue I was having, there is zero chance it was any kind of hardware problem. I'm not a noob, I've been working with model Fs for quite a few years now and I've never seen this kind of issue before. It was also fixed with a reflash, which very very strongly indicates a problem with the firmware.
NathanA wrote:
25 May 2022, 00:36
That said, it still kind of boggles my mind that it's not immediately "obvious" to almost everyone here (again, my myriad biases and blindspots shining through, almost surely) that "ibm_capsense" (original xwhatsit-designed-and-written) and QMK + pandrew's capsense support for QMK are wholly separate software projects
I'm well aware of this, I've used both tools plenty of times. My keyboard specifically was produced late enough that it comes with QMK installed, which is also what I flashed onto it.
NathanA wrote:
25 May 2022, 00:36
Well, you're really selling its virtues to your potential buyers. ;)
Sure beats the hell out of Karening to Joe for my money back, doesn't it?

NathanA

25 May 2022, 02:31

headphone_jack wrote:
25 May 2022, 01:35
I've restored probably about a dozen Model Fs, way more than I've restored Model Ms.
Fair enough & I shall defer to your experience in this department.
headphone_jack wrote:
25 May 2022, 01:35
I just used the F77 util to put it in bootloader mode. It took many, many, many tries for the keyboard to be recognized in the first place. I have no idea why this was, but it seems to have been resolved my reflash. [...] A reflash fixed every issue I was having, there is zero chance it was any kind of hardware problem.
Then I don't know what to tell you. Best theory I can come up with in light of this is that maybe there was some subtle corruption that occurred during the initial flash.
headphone_jack wrote:
25 May 2022, 01:35
NathanA wrote:
25 May 2022, 00:36
That said, it still kind of boggles my mind that it's not immediately "obvious" to almost everyone here [...] that "ibm_capsense" (original xwhatsit-designed-and-written) and QMK + pandrew's capsense support for QMK are wholly separate software projects
I'm well aware of this, I've used both tools plenty of times. My keyboard specifically was produced late enough that it comes with QMK installed, which is also what I flashed onto it.
Okay, so then did I misunderstand you when you said that you tried both QMK toolbox as well as "the capsense util tool"? Did you mean something by that other than the ibm_capsense util by xwhatsit? (Maybe you were trying it "just in case" lack of detection by QMK toolbox was an indication that somehow your board had been accidentally flashed with xwhatsit instead?)

headphone_jack

25 May 2022, 02:46

Andrei's firmware has a diagnostic tool that detects the capacitance of each pad, keypresses, allows you to put the board into bootloader mode, etc etc. Fairly sure it works on everything that uses QMK capsense.

NathanA

25 May 2022, 02:48

headphone_jack wrote:
25 May 2022, 02:46
Andrei's firmware has a diagnostic tool that detects the capacitance of each pad, keypresses, allows you to put the board into bootloader mode, etc etc. Fairly sure it works on everything that uses QMK capsense.
Yes. It's just not commonly called the "capsense util" but rather the "pandrew util", thus my confusion.

sedevidi

25 May 2022, 12:54

NathanA wrote:
24 May 2022, 23:26
The more interesting things (IMO) to observe are whether the high (pressed) and low (unpressed) values for your Ctrl key in the Signal Level Monitor seem to differ notably on average from what the other keys show, and what the computed threshold value is for that Ctrl key is vs. the other keys on the Keypress Monitor (which is what that number is: the threshold that pandrew QMK capsense driver computed for that particular key).
The Signal Levels were uniform, with low values on the left part of the split backspace, right control, and some keys in this area, which all worked fine. The Keypress Monitor did show uniform values too (with lower values on the right part of the main section of the keyboard). No funny things.

I openned the anvil, added a layer of isolation between the controller and the bottom of the case, tightenned one controller screw, tested without the case, and everything was seemingly OK. After closing the case with everything in place, all keys were OK too.
So: grounding issue... solved...

Left Control seems to be fully OK now, and the whole keyboard generally feels better. Maybe it's just that I openned it, found something, and I'm now happy, or it's that the keyboard is effectively better now: better responsiveness of the keys, or something really subtle.
So: a bad feeling may just be a matter of tightening a screw inside the beast.

headphone_jack

25 May 2022, 15:45

Ellipse wrote:
24 May 2022, 17:39
Regarding the install time, did you read the manual and check out the installation videos showing how to hold the keyboard up vertically, space bar end up? It should not take an hour to install the keycaps.
I've installed the keycaps on about a dozen original model Fs and Ms, exactly the way you described, even before I got my F77. Only on this keyboard did it take about three reseats each on average to have the switches click properly. Regarding what NathanA said, I believe this is caused by poor spring tolerances as opposed to the caps themselves, but I would need further testing to confirm.
Ellipse wrote:
24 May 2022, 17:39
Regarding the QMK firmware, the latest version is not buggy from what I have seen and read. It has been reliably working on the several thousand keyboards that have gone out so far, along with many original F107 and F122 keyboards and beam spring keyboards. If it was buggy I'd imagine there would have been some feedback by now that these keyboards were not usable by most folks.
I don't know what to tell you man. The firmware was very buggy when it first got to me, a reflash (although it took many many tries) eventually fixed the issue. I already had to open the assembly to check if some of the flippers were misaligned from shipping and I didn't see any kind of movement of the springs from their nubs.
Ellipse wrote:
24 May 2022, 17:39
If it was the pandrew utility, that one should recognize the keyboard unless it is running Via or the firmware has somehow been erased without being rewritten
Yes, I was using the Andrei utility. And again, I really don't know what to tell you. It took me many, many tries to get either QMK toolbox or the Andrei util to recognize the keyboard, and even more to flash it. It finally worked, for some reason, but I sincerely hope I never have to flash it again.

vpap

25 May 2022, 17:33

Hi everyone, my model f77 keyboard finally arrived! The initial impressions are good. The quality of the materials is indeed very high. Some minor issues regarding the keycaps were resolved easily by following the manual. Didn't hurt that I had already some experience with the model m.

But after I used it for a bit, I noticed some erratic behaviour regarding key presses. For example alt + tab wouldn't work every time or some key presses would not get registered. I had installed all the keycaps correctly without having to force anything into place. By following some of the instructions that I found so I proceeded to disassemble the outer casing and then screwing down the daughter board tightly. The problems still persisted. So I opened up the utility pointed out in the manual. Here are some screenshots.
before_pressing_keys.png
before_pressing_keys.png (41.45 KiB) Viewed 30587 times
after_pressing_all_keys.png
after_pressing_all_keys.png (52.65 KiB) Viewed 30587 times
pressing_only_alt_tab.png
pressing_only_alt_tab.png (29.7 KiB) Viewed 30587 times
Similar behaviour to the last image I got by pressing all the keys on the left most column several times which resulted on most of the keys of the first 3 rows to get green without pressing them.

NathanA

25 May 2022, 21:14

In your signal level monitor test, I note that your keys in the Y, U, F, V, and B positions seem to show very weak capacitance level differences when being pressed vs. unpressed. (Either that or you forgot to press them before taking the after_pressing_all_keys screenshot.) And the keypress monitor shows computed threshold values for those keys (in the mid-to-upper 140s) that are higher than the values they show after they've allegedly been pressed (between 120-132), so I would expect some trouble while using those keys unless the root problem there is identified and addressed.
vpap wrote:
25 May 2022, 17:33
Similar behaviour to the last image I got by pressing all the keys on the left most column several times which resulted on most of the keys of the first 3 rows to get green without pressing them.
This is the same behavior that sedevidi and I were discussing just a couple posts above yours, where multiple random keypresses show as being simultaneously registered in the utility when you press an unrelated key. I see the exact same kind of thing happen on my 2 perfectly working keyboards, which is why I called it a possible "red herring"...it doesn't seem to be diagnostically relevant.

It's almost like the keypress monitor in the utility is overly sensitive to a row getting "strobed", registering very rapid rise and falls in capacitance as a phantom keypress even though the actual matrix decoding logic in the firmware itself either ignores or filters those out. (The other keys showing phantom keypresses do not continue to show as being held down even if you continue to hold down the one key you pressed. I have tested it repeatedly, and the phantom keystrokes do always seem to happen only to keys on the same row as the key that you actually pressed. Select the raw "Matrix" layout at the bottom of the layout drop-down list when testing to see if it is the same in your case.) It would be nice if the keypress monitor reflected how the firmware would ultimately interpret what it saw rather than give you a seemingly more "raw" view of things, which I think is potentially leading people down the wrong paths, troubleshooting-wise.
Last edited by NathanA on 25 May 2022, 21:32, edited 1 time in total.

vpap

25 May 2022, 21:32

I see, thanks for the heads up NathanA. From what I understand, the 3rd note on the keypress monitor states exactly what you are saying. I just tried to reflow a bit the wire connections to the daughter board, which are easily accessible, but I still get the exact same behaviour. I am not 100% sure about which other keypresses are not working as they should but I can reliably replicate the unreliable alt + tab behaviour :P

NathanA

25 May 2022, 21:35

vpap wrote:
25 May 2022, 21:32
From what I understand, the 3rd note on the keypress monitor states exactly what you are saying.
Oh, good catch! Yes, I believe you are correct.
vpap wrote:
25 May 2022, 21:32
I am not 100% sure about which other keypresses are not working as they should but I can reliably replicate the unreliable alt + tab behaviour :P
Do either Alt or Tab show weird behavior when used individually?

vpap

25 May 2022, 22:17

I tried them individually on normal day to day work and I couldn't find something wrong. But by using the Keypress Monitor I found out that when I was pressing down some keys, some where a bit more responsive than others on quick successive presses. I tried after some presses to press down the key fully for a second or two. And then I saw that the alt key when completely pressed would not always be shown as pressed down.

Neightro

26 May 2022, 01:38

Does anyone know if there are any firmware configs for the alternate numpad layouts, e.g. Layout 5? I'm curious to see if it's possible to get it working properly with Num Lock, thought it's my understanding that doing so is complicated because of how scancodes work. Part of my reason for asking is to learn by example, so configs for QMK or Via[l] would be fine, but I would like to see both for comparison.

Thank you in advance. :)

NathanA

26 May 2022, 03:18

headphone_jack wrote:
25 May 2022, 15:45
It took me many, many tries to get either QMK toolbox or the Andrei util to recognize the keyboard, and even more to flash it. It finally worked, for some reason, but I sincerely hope I never have to flash it again.
Out of curiosity, do you find it is still as hard to get the keyboard to properly go into bootloader mode after having finally successfully reflashed it once? I'm wondering if in addition to solving the quirky functionality issues you faced, the reflash also fixed whatever issue was preventing it from going into bootloader mode in the first place?

NathanA

26 May 2022, 03:21

vpap wrote:
25 May 2022, 22:17
And then I saw that the alt key when completely pressed would not always be shown as pressed down.
Sounds like between the two of them, it's the Alt key that's the actual problem, then. If you sometimes see that keypress monitor doesn't show the key as being pressed down when it is, do you also similarly see signal levels not always changing as expected for that key on the signal level monitor when you press it?

I still think it's weird that Y, U, F, V, and B all show as low 130s at best on your signal level monitor screenshot after you said that you pressed them...are you also having problems with those keys?

headphone_jack

26 May 2022, 03:23

NathanA wrote:
26 May 2022, 03:18
headphone_jack wrote:
25 May 2022, 15:45
It took me many, many tries to get either QMK toolbox or the Andrei util to recognize the keyboard, and even more to flash it. It finally worked, for some reason, but I sincerely hope I never have to flash it again.
Out of curiosity, do you find it is still as hard to get the keyboard to properly go into bootloader mode after having finally successfully reflashed it once? I'm wondering if in addition to solving the quirky functionality issues you faced, the reflash also fixed whatever issue was preventing it from going into bootloader mode in the first place?
I haven't actually tried. I was so frustrated by the end of the whole endeavor I boxed up the keyboard and haven't opened it since, except to take pictures for listing.

NathanA

26 May 2022, 03:31

Neightro wrote:
26 May 2022, 01:38
Does anyone know if there are any firmware configs for the alternate numpad layouts, e.g. Layout 5? I'm curious to see if it's possible to get it working properly with Num Lock,
There is no way to get this working with a proper Num Lock as far as I am aware, at least not without some extensive re-working of QMK under-the-hood (like making it possible for the keyboard to change layers in response to the system's current Num Lock status, which I think would actually be a pretty cool feature...).

But I would be happy to put together a F77 Layout 5 config for you, much like I did a Layout 4 config for HatefulLight. The disclaimers & qualifications are that it would be applicable to Vial, so you'd have to flash to that, and it would be a similar kind of "pseudo" Num Lock: I can make the key actions switch between outputting numbers and nav commands by having you press a key that is physically labeled "Num Lock" or that you think of as "Num Lock", but actual Num Lock status on the computer won't be changing when you press it. If that works for your needs, I'll throw it together...

(Also, which key should toggle "pseudo-NumLock" on and off? Unlike Layout 4, there is no dedicated Num Lock key on the 3x5 pad. So you'll have to pick some other key on the 'board to use as a toggle. By default I think Ellipse mapped right Ctrl to the actual Num Lock. I can either re-map that to the so-called pseudo-NumLock, or you can pick a different key and use right Ctrl as an actual Ctrl key.)

Post Reply

Return to “Group buys”