I just got my keyboard in the mail, so I wanted to do a quick review.
The feeling is great. It feels better than my F122.
1- Less force is required to press on the key than my F122.
2- There is less key travel before the click than my F122.
3- It feels more snappy.
The keys feel better than any model m or model f keyboard I've ever used;
The spacebar feels and sounds very weird though;
N-key rollover is fine. Pressing 40+ keys at once is working fine, all are registered as down at the same time.
The keyboard is very heavy and solid, like it was made by IBM. I measured 1.985 kg;
But there are some issues.
Key number 6 was not working for a good 10 minutes when I first plugged it in. I kept pressing it repeatedly... until it finally started to work. Now it seems to be working fine.
The M key also felt very strange. So I tried pressing it slowly and noticed that sometimes it does a kind of double click. When pressing down slowly there's the first regular expected click, but then it seems to get stuck before reaching the bottom, and there's a second click. I pulled the key out, and noticed the black tape there for the space bar. The tape has unravelled itself, and is probably in the way of the key. The only place the tape is still stuck on the metal rod is in the pivot. Everywhere else, it peeled off, and is just being held by that pivot.
I want to try removing that black tape, but I've never removed the spacebar on a model F. Is the space bar easy to remove and put back in so I can remove the black tape?
The hole for the usb on the compact case is a little small, so not all usb-c connectors will fit. The one that comes with the keyboard fits, but the one from my phone won't fit. Also, make sure to push the usb-c cable all the way in. Sometimes, when I plug it in, I think it's all the way in, but it's not. You have to push relatively hard.
The firmware is messed up. When pressing the fn key, any other key you press gets pressed repeatedly, and super fast.
And if a key was programmed to have something on the fn layer, it alternates between the fn layer and the regular key.
So pressing fn-n I get: "n+n+n+n+n+n+n+n+n+n+n" super fast repeatedly, where + is the one from the 101 right side keypad.
Even the modifer keys are repated, so pressing fn-alt gives me "alt alt alt alt alt" repeatedly pressed and released.
I am unable to press any shift+fX, ctrl+fX, alt+fX, or ctrl-alt+fX... combination, which I need for regular operation. So I won't be able to use this keyboard until the firmware is fixed. Is there a firmware programmer that runs on linux?
The screws on the back of the case sticking out are a big annoyance... make sure to put the included rubber feet, because if you don't, and you slide the keyboard while it's on a valuable surface, it will scratch!
The top and bottom plates look and feel different from the middle layer body. They look and feel like they could be plastic. Are the top and bottom plates aluminium or plastic?
In summary, the main problems are the firmware, and the black tape on the spacebar, which should both be fixable; Other than that, it looks like a very good keyboard. Well recommended for anybody who likes model f or m switches and doesn't mind the absent function row;
- Humanoid
Model F Labs, Compact F77 review
-
Ellipse
- Location: United States
- Main keyboard: Brand New Model F Keyboards
- DT Pro Member: -
- Contact:
Thanks for posting this Humanoid - the first in depth review!
Please do keep these reviews coming so that any issues can be fixed.
Some comments:
1. I believe this is due to the barrels and springs being smoother and less oxidized, respectively, while new - as they age and become dirty, key presses often require a bit more force. I recommend 90-99% rubbing alcohol around the barrels when restoring Model F keyboards.
2. Key travel is a function of spring free length, distance inside the key (even 0.1mm differences are important). Also the sound is affected with different amounts of key travel. Having restored dozens of Model F's, I have found that many of them have slightly different key press lengths due to tolerance variations. I prefer the 1984 tolerances on the F122.
Yes the space bar does sound different from the Model F space bar. It is a Model M space bar used in a new Model F Keyboard. I decided early on to maximize keycap compatibility over duplicating the Model F space bar sound exactly.
Sometimes the springs will not work after being in the cold for a while - as noted in the manual, you may need to re-seat keys and/or springs with your key puller and tweezers/toothpick. Some keys will also sound fuzzy and it may be best to replace the spring if repeated efforts do not help (I have been doing this as part of QC but some issues may develop after my testing - as the keyboard is used intensively for the first time).
Regarding the electrical tape to reduce space bar metal on metal contact, I am testing out regular clear tape (Scotch tape) instead of electrical tape. I may also try polyimide tape.
Yes humanoid the space bar can be removed on the new Model F's as it is a Model M style space bar. Remember to pull it up a little and then slide out the space bar with stabilizer wire attached, or else the little tabs on the space bar will break and you'll need to replace the space bar. The metal space bar tabs are pushed down a bit to secure the space bar wire stabilizer so please be careful and wiggle it a bit to gently raise the metal tabs before removing the space bar.
Correct, for compact cases the hole in the back fits the narrow design USB cable connectors and will not fit wider connectors.
The plates are all metal - aluminum for the compact cases, steel for the top and bottom plates, and zinc for the classic style cases.
Yes on the screws - I tried countersunk screws in the initial prototypes but the case was not thick enough for the countersunk screws to reliably hold the case together. Also more screws had to be added and the bottom case plate was made thicker compared to the original compact case prototypes.
The holes are for bumpers with screws, similar to the holes in the original 4704's that were used to secure the plastic bar. I offer half inch and one inch bumpers with screw and nut.
Some notes on common xwhatsit issues:
Specifically with your issue of keyboard "spamming" (random key presses/key signal/other related issues) please try the alternative debounce threshold 6 firmware (more details below) and adjust the threshold 2 or 3 higher. For example I saw spamming with threshold 126 that went away at threshold 128 - click store to EEPROM after adjusting the threshold. If set too high, some keys may not register so be sure to test all keys after adjusting the threshold. I am currently typing on a debounce threshold 11 firmware with voltage threshold set to 126 and there are zero issues.
In general for everyone with xwhatsit issues, please do check out the various threads on GH and DT and also please do review the xwhatsit manual (link on ModelFKeyboards.com/code) as a number of the issues you mention have come up before for those using an xwhatsit controller on an original Model F. I will try to summarize some of the issues and potential solutions below.
As noted before, DMA has a working prototype CommonSense Model F controller that will be able to replace the xwhatsit of the new Model F's, so that may be an option to consider down the line (though the xwhatsit has worked well for several hundred Model F fans over the past 5 years).
Wait a while (half hour?) for the keyboard to get to room temperature. Connect the keyboard and wait another half hour for things to stabilize - the xwhatsit is sensitive to grounding and temperature changes.
Your firmware issues were reported years ago as well - sometimes called "xwhatsit spamming of keys." https://geekhack.org/index.php?topic=70247.0
The solution was first to adjust the voltage threshold higher or lower, but not too low as lots of keys will randomly signal - may be best to export your layout and then erase it by importing a blank layout. I recommend leaving it in manual threshold mode so click store in EEPROM after each adjustment. Another working solution was to try out a number of alternative debounce filters and filter levels as updated by Deskthority user joc. I will post a few firmware options here.
I installed debounce threshold 11 firmware on everyone's keyboard so far but for some users level 6 may be better.
Regarding your question about firmware programming, it is easy to do on PC, Mac, and Linux but on Linux you have to compile the GUI software and then review the manual for the procedure.
My layout input mistake on the shift-Fn+X. You just have to add the shift keys and any other needed keys to the "Layer 1" tab of the xwhatsit GUI so that it is recognized. I'll do this going forward for everyone.
Please do keep these reviews coming so that any issues can be fixed.
Some comments:
1. I believe this is due to the barrels and springs being smoother and less oxidized, respectively, while new - as they age and become dirty, key presses often require a bit more force. I recommend 90-99% rubbing alcohol around the barrels when restoring Model F keyboards.
2. Key travel is a function of spring free length, distance inside the key (even 0.1mm differences are important). Also the sound is affected with different amounts of key travel. Having restored dozens of Model F's, I have found that many of them have slightly different key press lengths due to tolerance variations. I prefer the 1984 tolerances on the F122.
Yes the space bar does sound different from the Model F space bar. It is a Model M space bar used in a new Model F Keyboard. I decided early on to maximize keycap compatibility over duplicating the Model F space bar sound exactly.
Sometimes the springs will not work after being in the cold for a while - as noted in the manual, you may need to re-seat keys and/or springs with your key puller and tweezers/toothpick. Some keys will also sound fuzzy and it may be best to replace the spring if repeated efforts do not help (I have been doing this as part of QC but some issues may develop after my testing - as the keyboard is used intensively for the first time).
Regarding the electrical tape to reduce space bar metal on metal contact, I am testing out regular clear tape (Scotch tape) instead of electrical tape. I may also try polyimide tape.
Yes humanoid the space bar can be removed on the new Model F's as it is a Model M style space bar. Remember to pull it up a little and then slide out the space bar with stabilizer wire attached, or else the little tabs on the space bar will break and you'll need to replace the space bar. The metal space bar tabs are pushed down a bit to secure the space bar wire stabilizer so please be careful and wiggle it a bit to gently raise the metal tabs before removing the space bar.
Correct, for compact cases the hole in the back fits the narrow design USB cable connectors and will not fit wider connectors.
The plates are all metal - aluminum for the compact cases, steel for the top and bottom plates, and zinc for the classic style cases.
Yes on the screws - I tried countersunk screws in the initial prototypes but the case was not thick enough for the countersunk screws to reliably hold the case together. Also more screws had to be added and the bottom case plate was made thicker compared to the original compact case prototypes.
The holes are for bumpers with screws, similar to the holes in the original 4704's that were used to secure the plastic bar. I offer half inch and one inch bumpers with screw and nut.
Some notes on common xwhatsit issues:
Specifically with your issue of keyboard "spamming" (random key presses/key signal/other related issues) please try the alternative debounce threshold 6 firmware (more details below) and adjust the threshold 2 or 3 higher. For example I saw spamming with threshold 126 that went away at threshold 128 - click store to EEPROM after adjusting the threshold. If set too high, some keys may not register so be sure to test all keys after adjusting the threshold. I am currently typing on a debounce threshold 11 firmware with voltage threshold set to 126 and there are zero issues.
In general for everyone with xwhatsit issues, please do check out the various threads on GH and DT and also please do review the xwhatsit manual (link on ModelFKeyboards.com/code) as a number of the issues you mention have come up before for those using an xwhatsit controller on an original Model F. I will try to summarize some of the issues and potential solutions below.
As noted before, DMA has a working prototype CommonSense Model F controller that will be able to replace the xwhatsit of the new Model F's, so that may be an option to consider down the line (though the xwhatsit has worked well for several hundred Model F fans over the past 5 years).
Wait a while (half hour?) for the keyboard to get to room temperature. Connect the keyboard and wait another half hour for things to stabilize - the xwhatsit is sensitive to grounding and temperature changes.
Your firmware issues were reported years ago as well - sometimes called "xwhatsit spamming of keys." https://geekhack.org/index.php?topic=70247.0
The solution was first to adjust the voltage threshold higher or lower, but not too low as lots of keys will randomly signal - may be best to export your layout and then erase it by importing a blank layout. I recommend leaving it in manual threshold mode so click store in EEPROM after each adjustment. Another working solution was to try out a number of alternative debounce filters and filter levels as updated by Deskthority user joc. I will post a few firmware options here.
I installed debounce threshold 11 firmware on everyone's keyboard so far but for some users level 6 may be better.
Regarding your question about firmware programming, it is easy to do on PC, Mac, and Linux but on Linux you have to compile the GUI software and then review the manual for the procedure.
My layout input mistake on the shift-Fn+X. You just have to add the shift keys and any other needed keys to the "Layer 1" tab of the xwhatsit GUI so that it is recognized. I'll do this going forward for everyone.
- Attachments
-
- xwhatsit files.zip
- (52.2 KiB) Downloaded 165 times
-
Humanoid
- Location: Canada
- Main keyboard: Matias Tactile Pro for PC
- Main mouse: MX Master 2S
- Favorite switch: IBM Beamspring
- DT Pro Member: -
I successfully removed the tape from the spacebar, and cleaned the leftover glue gunk on the wire stabilizer with rubbing alcohol. It was relatively easy to remove and replace the spacebar.
I almost forgot what it originally sounded and felt like with the tape on... but according to my memory, I think with the tape removed, it sounds louder, but feels much better.
And since feeling is more important than sound to me, I prefer it without the black tape.
I find the F77 spacebar now feels a lot more like my F122 spacebar. And it's much better than the model m spacebar.
As for the sound, it sounds closest to my model F emr keyboard 2, which has a very stiff spacebar compared to the F122. Maybe there's something related going on.
And I almost forgot to mention that the problem with the M key is completely gone after removing the tape. The unravelled tape was indeed sometimes getting in the way.
I almost forgot what it originally sounded and felt like with the tape on... but according to my memory, I think with the tape removed, it sounds louder, but feels much better.
And since feeling is more important than sound to me, I prefer it without the black tape.
I find the F77 spacebar now feels a lot more like my F122 spacebar. And it's much better than the model m spacebar.
As for the sound, it sounds closest to my model F emr keyboard 2, which has a very stiff spacebar compared to the F122. Maybe there's something related going on.
And I almost forgot to mention that the problem with the M key is completely gone after removing the tape. The unravelled tape was indeed sometimes getting in the way.
-
Wazrach
- Location: Tyne and Wear, United Kingdom
- Main mouse: Razer Viper 8KHz/ Viper Mini
- Favorite switch: Buckling springs
- DT Pro Member: -
The Models F with the large spacebars WILL sound and feel different to the ones with regular-sized spacebars. I am not sure why your F77 spacebar feels stiff, as it should feel as light as the rest of the keys. The large spacebar system of the XT and AT, etc, uses a stabiliser wire inside the assembly which makes the spacebar very stiff. This is not the case on the F122 or any of the keyboards with smaller spacebars.
How is the sound of the spacebar? You said it was loud, but does it sound good? How does the smoothness of the switches compare to your F122?
Really excited to get my hands on mine.
-
Humanoid
- Location: Canada
- Main keyboard: Matias Tactile Pro for PC
- Main mouse: MX Master 2S
- Favorite switch: IBM Beamspring
- DT Pro Member: -
No, it's my ibm emr keyboard 2, which has the stiff spacebar. The F77 spacebar feels really good and is almost as good as my F122 spacebar, now that I removed the black tape.
The sound is loud, but also sounds pretty good. I think the best way to describe it is to say that the regular keys make "click" sounds, but the space bar is a loud "clunk" sound. Sounds very heavy and massive.
For the smoothness, I'm not sure. Some keys feel smoother on the F77, while others feel smoother on the F122. But the difference is so slight, that I don't notice it during regular typing. During regular typing, I mostly notice the difference in force, and the key travel before the click.
-
Bass
- Location: United States
- Main keyboard: IBM Model F104 "Unsaver"
- Main mouse: Logitech G9x
- Favorite switch: Beamspring / Capacitive Buckling Spring
- DT Pro Member: 0206
Care to post some pictures? Since I think you are the first person who has confirmed to receive one of these, I think this would excite a lot of us.
-
Ellipse
- Location: United States
- Main keyboard: Brand New Model F Keyboards
- DT Pro Member: -
- Contact:
The award for first photo in the wild goes to the first keyboard sent out - received by a geekhack member: https://geekhack.org/index.php?topic=79 ... msg2845230
-
Humanoid
- Location: Canada
- Main keyboard: Matias Tactile Pro for PC
- Main mouse: MX Master 2S
- Favorite switch: IBM Beamspring
- DT Pro Member: -
The screws on the back are terrible. They are cheap screws made of very soft metal that damage easily, and some have been screwed on so tight that I've damaged them, taking them out. And there is one long screw left that I can't get out no matter how much I tried. My screwdrivers of course have no damage whatsoever on them, because they are made of steel.
What is the optimal philips screwdriver size for those screws? And how many Nm of torque are they using at the factory??? I need to get that last screw out!
I used a philips #1 driver to get all the other screws out. But this one isn't catching, and isn't moving, so I'm thinking I might be using the wrong size driver.
I will buy replacement screws made of steel, after I get this last one out.
What is the optimal philips screwdriver size for those screws? And how many Nm of torque are they using at the factory??? I need to get that last screw out!
I used a philips #1 driver to get all the other screws out. But this one isn't catching, and isn't moving, so I'm thinking I might be using the wrong size driver.
I will buy replacement screws made of steel, after I get this last one out.
-
Ellipse
- Location: United States
- Main keyboard: Brand New Model F Keyboards
- DT Pro Member: -
- Contact:
Yes I have also damaged screws before by using the wrong size driver, especially with nonstandard slotted screw heads - even a slightly smaller driver will damage these small screws. Sorry I don't know the driver size or torque.
Update - the supplier notes that they use Stainless Steel 304 screws and that non-metric screwdrivers will damage the screws.
Update - the supplier notes that they use Stainless Steel 304 screws and that non-metric screwdrivers will damage the screws.
-
Humanoid
- Location: Canada
- Main keyboard: Matias Tactile Pro for PC
- Main mouse: MX Master 2S
- Favorite switch: IBM Beamspring
- DT Pro Member: -
I don't believe that.
What in the world is a non-metric philips screwdriver? Philips has sizes like 00, 0, 1, 2, 3...
00 and 0 are clearly too small and wiggle too much. 1 fits like a glove, while 2 is clearly too big.
I used a philips #1 to get all the screws out, except for that one screw, which I think might've already been damaged before I got the keyboard due to over tightening at the factory. They clearly did not use a specific torque as some screws were very easy to take out, while others were difficult, and this last one is impossible with a regular screwdriver.
The only way I got the wrong screwdriver is if it's not supposed to be a philips. I've never heard of a metric vs non-metric philips screwdriver.
-
Humanoid
- Location: Canada
- Main keyboard: Matias Tactile Pro for PC
- Main mouse: MX Master 2S
- Favorite switch: IBM Beamspring
- DT Pro Member: -
-
Ellipse
- Location: United States
- Main keyboard: Brand New Model F Keyboards
- DT Pro Member: -
- Contact:
Thanks Humanoid for the feedback on the screws. The factory reports they are #1 phillips head (PH1) screws. Because of your feedback the compact case screws are expected to change to Torx screws for the final round (more resistant to slipping). I expect to include the appropriate size torx screwdriver with each keyboard so people don't have to buy it.
-
Humanoid
- Location: Canada
- Main keyboard: Matias Tactile Pro for PC
- Main mouse: MX Master 2S
- Favorite switch: IBM Beamspring
- DT Pro Member: -
I'm glad to report that all remaining issues have been solved!
I compiled and installed xwhatsit's capsense utility. The problem with the repeated keys was not due to the threshold value. I read the manual, and it says that when you set a particular key to be Fn1 on the base layer, you must also set it on the other layer as well.
So what happened, was the Fn1 key was set on the base layer, but on layer 1, it was set to ignored. So when you press Fn1, it switched to layer 1, where the key is now ignored, which makes it release the key, switching back to the base layer, where the key is now considered Fn1, so Fn1 is pressed again, and this would repeat forever. I find it odd that this error wasn't detected during QA.
So all I had to do was set the Fn1 key on layer 1, and it fixed the problem.
I also got a pair of vampliers in the mail. Using the vampliers I was immediately able to take that last screw out! When squeezing the vampliers, the screw would dent so easily, making it easy to catch and twist out. I still don't believe those are steel screws... I think somebody in the supply chain is lying.
Now I want to buy my own steel screws. Ellipse, do you have the specifications for the screws so I can buy the right ones?
I compiled and installed xwhatsit's capsense utility. The problem with the repeated keys was not due to the threshold value. I read the manual, and it says that when you set a particular key to be Fn1 on the base layer, you must also set it on the other layer as well.
So what happened, was the Fn1 key was set on the base layer, but on layer 1, it was set to ignored. So when you press Fn1, it switched to layer 1, where the key is now ignored, which makes it release the key, switching back to the base layer, where the key is now considered Fn1, so Fn1 is pressed again, and this would repeat forever. I find it odd that this error wasn't detected during QA.
So all I had to do was set the Fn1 key on layer 1, and it fixed the problem.
I also got a pair of vampliers in the mail. Using the vampliers I was immediately able to take that last screw out! When squeezing the vampliers, the screw would dent so easily, making it easy to catch and twist out. I still don't believe those are steel screws... I think somebody in the supply chain is lying.
Now I want to buy my own steel screws. Ellipse, do you have the specifications for the screws so I can buy the right ones?
-
Ellipse
- Location: United States
- Main keyboard: Brand New Model F Keyboards
- DT Pro Member: -
- Contact:
Great Humanoid, glad everything is working now!
Noted on the Fn1 - the layouts all have fn1 on both layouts - not sure why yours did not. Sorry about that error.
M2.5 x 20 Pan Head Machine Screw to attach the case pieces together and M2.5 x 4 Pan Head Machine Screw for the 4 screws connecting the bottom case plate to the bottom inner assembly plate.
The classic cases use 6-32 screws, 1 in 3/16 length (strain relief P clip) and 4 case screws in 1/2 inch length.
Noted on the Fn1 - the layouts all have fn1 on both layouts - not sure why yours did not. Sorry about that error.
M2.5 x 20 Pan Head Machine Screw to attach the case pieces together and M2.5 x 4 Pan Head Machine Screw for the 4 screws connecting the bottom case plate to the bottom inner assembly plate.
The classic cases use 6-32 screws, 1 in 3/16 length (strain relief P clip) and 4 case screws in 1/2 inch length.