Author Topic: In search of the click: Leopold FC750R, IBM Model M and beyond  (Read 2636 times)

0 Members and 1 Guest are viewing this topic.

Offline ikonomov

  • Thread Starter
  • Posts: 42
In search of the click: Leopold FC750R, IBM Model M and beyond
« on: Wed, 25 December 2019, 15:25:08 »
After two years of using two Ducky One keyboards with fading Dye-sub legends on PBT keycaps reported here, about a year ago I decided to give Leopold another try.  This is my second experience with Leopold keyboards, the first being a disappointment that Iíll discuss below.  This time I bought two FC750R TKL keyboards with Double Shot PBT in Two Tone White and the other in Black, both with Cherry MX Blue switches.  Despite having the preconception that Iíll not like Cherry profile keycaps, I felt in love with the keyboards.  The Two Tone White version brought some nostalgia, and about six months ago I decided to push further and find out if thereís something worth for me to rediscover in a good old Buckling Spring IBM Model M, so I bought a full sized and a SSK (TKL) Model M.  The discussion below is about these four keyboards, two from the past, the others available now in the present, and finally some of my thoughts about a possible future for tactile mechanical keyboards.

Case: The compact plastic case of the Leopold keyboard looks well designed and inspires a feeling of using something premium thatís built to last.  There are four good-sized rubber pads on the bottom of the case as well as two rubber pads on the tilt legs.  The Model M keyboards are substantially larger and only have two small rubber pads which despite having a heavier chassis can still allow the keyboard to move.  The Model M keyboards use thicker plastic for the case, supposedly a blend of ABS and PVC, while the Leopold is likely made from ABS.  The Model M feels rigid enough to be able to be used as a baseball bat while the Leopold feels more refined.

Switches: This is the reason why people who like tactile keyboards should get excited about Buckling Spring.  The following image shows the force diagrams of the Buckling Spring used in the Model M and the Cherry MX Blue switches used in the Leopold.  As a key is pressed, tactility on the Model M comes from a sudden drop in resistance as the spring buckles.  The tactility on Cherry MX Blue switches feels like a bump as the key is pressed.  The Model M feels somewhat more tactile because of the vertical drop in force rather than the more gradual slope up and down at the tactile bump on the Cherry MX Blue and also because of the slightly larger drop in force.  While the peak force right before actuation is only 5gr lower at the bump on the Cherry MX Blue, the keys on Model M feel noticeably stiffer as the slope between the up position and actuation is significantly higher on the Model M.  As a result of the tactile bump and the way the click mechanism is reset on a Cherry MX Blue, the keys feel smoother and subjectively more pleasant as they travel up and down on a Model M.  The Model M also does not bottom out harsh like Cherry MX switches.


Another difference is the way the keys reset.  On a Buckling Spring after actuation a key must travel back up past the hysteresis point before it can be pressed again to register another key press.  On a Cherry MX Blue switch the reset point is just below the actuation but before the point where the click mechanism is reset, which means that a second actuation although difficult to execute is possible.  Objectively this makes the mechanism of the Buckling Spring more precise.  The only way a Model M will register a key press is when a spring buckles, while on a Cherry MX Blue an actuation is possible without a click as the key doesnít need to be fully reset.

After using Cherry MX Blue keyboards for years the first thing I noticed after typing on the Model M was the stiffness of the keys.  I was really hoping to get used to the heavier but smoother and more refined key action of the Model M, but after using it for a few weeks every time I switched back to the Leopold the Cherry MX Blue switches gave me joy and felt effortless to type on.  I can only speculate, but my guess is that at least partly the reason why IBM originally designed the Buckling Spring as stiff as they have is because back then there must have been a lot of people still used to typing with two fingers.  Ultimately after many typing tests I went back to the Leopold keyboards as I concluded with disappointment that the Model M is simply too heavy for my fingers.

I hope to be able to come back to the Buckling Spring some day, as I did some tests by using lighter springs and also deforming some of the springs by stretching them, and I was able to get absolutely superb lighter buckle which tells me that there is a huge area still unexplored for tactile keyboards.  In a market that is now saturated by perfectly refined linear Cherry MX style switches, Iím convinced that the time has come to shift the spotlight towards new and improved Buckling Spring keyboards with a superb mechanism that can allow different key weights with only minor modifications.  One of them for example is being able to mount the spring to the hammer with a slightly looser but perfect fit, something that I found to be critical for the consistency of the buckle between different keys.

Stabilizers: While earlier versions of the Leopold keyboards used standard Cherry stabilizers, the ones that I received use modified Cherry stabilizers.  The movement is smooth with no perceivable resistance when bottoming out making the stabilized keys feel exactly the same as the other keys.  Since this keyboard uses thick keycaps, Costar stabilizers would be incompatible.  The buckling spring barrel of the Model M provides a more stable mechanism for all the keys and together with the stabilizers create an absolutely superb key travel action without any wobble.

Keycaps: The Leopold has thick double shot PBT with Cherry profile keycaps.  These are by far the most beautiful and well made keycaps I have ever seen.  The whole keycap including the legends is PBT.  The font is in bold, quite simple and very legible.  The surface texture is perfect, not too smooth and not too rough.  I could not see any seams or difference on the surface of the keycap between the legends and the rest of the keycap.  The texture at the top of the keycap is one completely uniform surface.  This is the pinnacle of quality.  The PrtSc and Pause buttons have Sys Rq and Break likely pad printed on the side, which is fine as theyíll never wear out being on the side of the keycaps, and the print is absolutely indistinguishable from the double shot legends on top of the keycaps which attests to the quality of the double shot legends.  The quality of the molding on these keycaps and their legends is in my opinion on par with the very best double shot ABS keycaps I have ever seen.  Similarly the keycaps and the legends of the Model M are just as nice, in taller profile with an arguably more elegant font, if only slightly less sharp because of the dye sublimation.  A clear advantage of the double shot keycap is that it allows lighter legends on darker colored keycaps, as is the case with the black Leopold.

I have been reluctant to buy a keyboard or keycaps with Cherry profile because they canít be used with regular o-rings as the key travel is reduced too much.  My doubts about Cherry profile was further reinforced almost a year before buying these two Leopold keyboards when I bought my first two Leopold keyboards which then werenít available in my preferred TKL form factor with Cherry MX Blue switches.  My intention then was to simply use the keycaps and install them on my Ducky One keyboards.  When I did I discovered that the switches did not click properly.  Thanks to a comment from a fellow geekhack member, I later found out that the problem was that Cherry profile keycaps have to be used on switches mounted with LED side facing south in order to bottom out properly.  This makes these Leopold keyboards quite unique, as they are one of the few that can accept thick keycaps with Cherry profile having both Cherry stabilizers and properly mounted switches.

Sound: Prior to buying the Leopold keyboards I was using thick Ducky PBT OEM keycaps with o-rings.  I like my typing to be as quiet as possible and as an additional measure to lower the noise I use an extra large mouse pad underneath my keyboard.  One of my major concerns was that the Leopold will be louder since I canít use o-rings with it.  To my surprise Iím happy to report that after doing many typing tests both on the Ducky and Leopold I did not notice significant change in the overall sound level.  With o-rings the Ducky was nearly silent when bottoming out, but significantly louder on key release, while the Leopoldís overall sound is slightly lower in pitch and on average I would say perceived as having about the same loudness.  The loud key release on the Ducky is likely caused by a number of factors including slightly thinner keycaps, 1.4mm vs 1.5mm and larger and taller inner chamber of the OEM profile keycaps.  The thin sound absorbing pad inside the Leopold might also contribute to dampening or in the very least changing some of the sound characteristics.  I believe that even the sound of the click from the Cherry MX Blue switches themselves is more pleasant, slightly altered with lower pitch.

The Model M produces a very loud and distinct ping from the Buckling Spring which is one of its defining characteristics.  Unlike the loud clack from bottoming out on a Cherry MX keyboard, the Buckling Spring bottoms out softer and is almost inaudible.  Overall the Model M is definitely louder than the Leopold, but I donít really have a problem with the sound of either keyboard although I donít really enjoy the click or the ping either.

Software: None.  I donít see how it can be useful for most people on a non-lit keyboard.

Backlighting: There is no backlighting on the Leopold except the LED indicators on the Caps Lock and Scroll Lock keys.  I prefer the lock indicators in their standard placement on the case itself as on a Filco, but surprisingly they are bright enough to be easily visible, illuminating the underside of the keycaps.  I like to be able to easily see the legends of all keycaps both during the day and at night.  Unfortunately I have never seen backlit PBT keycaps that have enough contrast in their legends to be easily legible during daylight as well.  The best solution that I have found is to use white or very light keycaps with black legends, which reflect very well any available light at night and are surprisingly almost always visible and readable.

The Model M doesnít have backlighting, and the SSK version doesnít even have LED indicators for the lock keys, which I find quite inconvenient.  As a result I wrote an AutoHotKey script which allows lock key indication to be shown on Windows Taskbar.  The full sized Model M has LED indicators for the lock keys that are pretty much perfect.

Layout:  The ANSI layout first introduced by IBM with their Model M keyboards has now become a standard and is widely used by most modern keyboards including the FC750R.  The Leopold has a standard bottom row with standard stabilizer mounts spacing on the spacebar.

A surprising discovery for me while using the Model M keyboards was in my opinion the superior placement of the f-row keys on the Leopold keyboards.  While the curvature of the metal plate on the Model M seems to work great for the first 5 rows of keys, the top f-row is moved too far back which results in those keys having their axis of actuation tilted forward too much.  The natural movement of the fingers while typing and depressing the keys is mostly vertical up and down regardless of whether the fingers are bent or fully extended.  Pressing the f-row keys on the Model M felt both unnatural and not easily accessible.  Between the two types of keyboards my preference is definitely for the ergonomic profile of the Cherry keycaps using flat mounted switches.  If I was to design a new Buckling Spring keyboard, I would definitely use a flat plate with a keycap profile similar to Cherry.

PCB: The Leopold keyboard has two-layer PCB which should ensure durability and secure mounting for the soldered switches.  There are dip switches on the back of the keyboard to change the location of Windows and Caps Lock keys.

Fn functionality: I find this functionality unnecessary and I wish there was a way to remap the Fn key as a second Windows key using a dip switch on the back of the Leopold keyboard.  The repeat rate functionality in particular I find absolutely useless and I really wish it wasnít there so that it would not be possible to change it by mistake from the default x1.  I think the multimedia and volume Fn keys would have been much more useful if the side of those keys were pad printed (same as Sys Rq and Break keys) with icons representing their functions in a color different (say gray) from the main legends.  The Model M doesnít have an Fn layer, although the SSK does have an extra layer for Num Lock functionality which the FC750R TKL doesnít have.  The Fn functions on the Leopold are as follows:

Home: NKRO (USB only)
End: 6KRO (USB only)
F1: repeat rate x1 (PS/2 only)
F2: repeat rate x2 (PS/2 only)
F3: repeat rate x3 (PS/2 only)
F4: repeat rate x4 (PS/2 only)
F6: multimedia previous
F7: multimedia play/pause
F8: multimedia next
F9: multimedia stop
F10: mute
F11: volume down
F12: volume up

USB Controller: The Leopold uses low speed USB controller with NKRO available both under USB and PS/2.  This means non-standard USB NKRO implementation and also necessitates 6KRO in its default state.  Since there is no memory function, if the computer looses power or the keyboard is unplugged it will reset to 6KRO.  I would like to have seen a full speed USB controller used with NKRO and a dip switch on the back for 6KRO compatibility.  Something interesting that Iíve discovered worth mentioning is that unlike most of my previous Cherry MX keyboards, this keyboard has been programmed with the right amount of debounce delay for Cherry MX switches.  All my previous Cherry MX keyboards except the Leopolds and a Filco were registering double clicks, while for the one year that I have used the Leopolds and the few years I have used a Filco I have not noticed this happening even once.  After watching this video and confirming that certain keyboards have different latency regardless of the USB polling rate or interface, Iím convinced that these Leopold keyboards and my previous Filco use more conservative longer debounce delays.

The Model M has 2KRO over PS/2 which before buying the keyboards I thought would be the biggest problem for me.  After doing various tests, I can confirm that I was able to consistently get between 3 and 5 keys to register simultaneously in addition to the modifier keys.  I found this enough to play all my games, and I only see a problem if a game requires the use of both hands on the keyboard.  Still a full speed USB controller with NKRO would have definitely been preferred, especially using a Model F capacitive technology.

Connection: The PVC sheathed cable on the Leopold is detachable with a safe mini-USB connection port underside the case with channels to be routed straight, left or right.  There is a USB to PS/2 adapter included.  Using the adapter to plug in straight to the PS/2 port on a computer doesnít seem very secure, however, as it sticks out quite a bit.  The PS/2 port is not really plug-and-play safe, so I bought a PS/2 extension cable that I have plugged directly to the computer with a much more secure plug.

The two Model M keyboards that I received have PS/2 coiled cables.  They look quite thick and durable and use a secure but proprietary port to connect to the keyboard.  It is possible to connect the keyboard to a USB port using a PS/2 to USB converter.

This would have to be it for now.  Sorry for the long read, but you guys know how hard it is at times to resist hitting those beautiful keys...  ;D
« Last Edit: Mon, 06 January 2020, 09:33:34 by ikonomov »

Offline leech

  • Posts: 79
  • Location: gemani
  • Öalways on the lookout for higher qualityÖ
Re: In search of the click: Leopold FC750R, IBM Model M and beyond
« Reply #1 on: Mon, 30 December 2019, 04:42:45 »
very thoughtful, insightful, careful, informative, truthful, detailed write-up!

actually the best I've ever seen on dis geeks forum

Thank you and Happy New Year man!!

Offline HungerMechanic

  • Posts: 245
Re: In search of the click: Leopold FC750R, IBM Model M and beyond
« Reply #2 on: Mon, 30 December 2019, 09:43:28 »
I agree with your opinion of the Leopold, and especially the FN key. It would be useful to have a DIP switch to transfer the FN key into a second Windows key. I grew accustomed to having 2 OS keys from my Filco.

Offline ikonomov

  • Thread Starter
  • Posts: 42
Re: In search of the click: Leopold FC750R, IBM Model M and beyond
« Reply #3 on: Mon, 30 December 2019, 15:03:28 »
Thanks guys.  Happy New Year!