Fujitsu Peerless Stabilizer Improvements

drevyek · 14 Aug 2016, 08:27

I have a Epson Q203A with Fujitsu Peerless switches that I have been trying to restore/improve. The board has a lot of personal significance to me (my first keyboard, was my Grandfather's), so I would prefer to do my best on it.

The Q, in its grimy, uncleaned glory

Spoiler:

One of the main sticking points that I have found are the stabilizers and the horridly rough feel to them. Off centre, they bind so hard that they are basically unusable. The normal keys aren't worlds better. It is a really puzzling part of the whole thing. Great construction, with PBT dye-sublimated caps, a curved steel backplate (think the Model M, but pre-bolt-modded!). A really interesting way of using a rubber-dome for tactility, while still being not-awful. But then, so so so rough!
I know that the Peerless switches have some fans out there, so I was wondering if anyone had any experience in improving the stabs, or the roughness of the overall feel.

Slider, cleaned beside a stabilized key (shows the pin-stabilizer, and its barrel)

Spoiler:

Barring that, my plan to get it smooth(er) is mostly by cleaning (mostly done) and lubing. A lot of the friction comes from the stabilizers on the switch, as well as the stabs for the extra-wide keys. If you can add in lube on these contact points, as well as all around the pin-stabs, then it should reduce the friction and rough feeling significantly. I've attached some photos of the contact inside the barrel assembly, which sort of shows the offending areas.

Unpressed

Spoiler:

Pressed on the switch side

Spoiler:

Pressed on the stab side (showing the main binding points, on the slider and the stab)

Spoiler:

I have heard a lot of good about using a silicon grease, but I'm not sure of what type etc might work best. The plastic of the barrel isn't terribly good for the stabs (not unlike the barrel assembly of the Model M). IBM used separate inserts to make it smoother. Fujitsu said "fuck it", and just had the blank barrels. Where IBM used different inserts (or wires) on the numpad, Fujitsu added a separate stabilizer that clips into the assembly, which feels a lot like the Model M horizontal inserts. Works better than the other stabs, but doesn't fully fix the binding issue. This is why I think that adding lubing would make it work, at least better than it does.

Vertical stabs

Spoiler:

I'll post updates here as I try to work through the kinks.

As a reference, the board is from April 1988, and was used for years on an 8080 IBM compatible (the nicotine was quite thick on it). If anyone has any good suggestions on removing that smell, I'm all ears.
The 8080 still works, I'm pretty sure(still has everything, including the printer, and the DOS 3.1 5.25" Floppy), so I might get it booted up and post about that.

seebart · 14 Aug 2016, 16:43

Very nice to see you taking a closer look at Fujitsu Peerless. While not my favorite switch I can appreciate it. I did take my Sanyo apart way back when I looked at it and found very similair problems to what you are describing. I have not tried lubing mine, I just got some Krytox-GPL-205, I would try something similair and lube one switch tentatively one by one part so we can find out what exactly needs lubing. Fujitsu Peerless is a strange switch by construction with the cylinder and the dome. Keep us updated.

wiki/Fujitsu_Peerless

photos-f62/fujitsu-fkb4700-sanyo-t9119.html?hilit=sanyo

Chyros · 14 Aug 2016, 17:28

Stab problems is by far the worst thing about Fujitsu Peerless, it wouldn't've been such a bad switch without this. NOS boards have been reported to not suffer from this issue but it's terribad on used examples, somewhat similar to the right shift with Space Invaders.

seebart · 14 Aug 2016, 17:35

Yes that's why it will be interesting to see how much lubing will help. I may try it on mine but I need every bit of that Krytox for my Alps boards. I'd love to try a NIB FKB4700. Crazy to think that this is the same manufacturer of brilliant swtiches like magnetic reed and leaf spring. And the parent of PFU and the HHKB.

drevyek · 14 Aug 2016, 18:06

I'll give Krytox a go, especially if I can find it locally. The main points of friction are at the 2 feet on the top slider, and the stab peg.

Interestingly, the stab peg has a little cut-out, which I think was intended to absorb some of the down force. It looks like it's doing that inside the vertical stabs.

Chyros wrote: ↑NOS boards have been reported to not suffer from this issue but it's terribad on used examples, somewhat similar to the right shift with Space Invaders.

I hope so. I am making the analogy to Alps, where older switches were lubed by Alps, but later switches uses a different plastic. Similarly, I suspect that the Peerless were lubed to compensate for the high-friction plastic used for the barrels. The vertical stab holes are of a slippery plastic, akin to the lube-less Alps.

Chyros · 14 Aug 2016, 20:30

seebart wrote: ↑Yes that's why it will be interesting to see how much lubing will help. I may try it on mine but I need every bit of that Krytox for my Alps boards. I'd love to try a NIB FKB4700. Crazy to think that this is the same manufacturer of brilliant swtiches like magnetic reed and leaf spring. And the parent of PFU and the HHKB.

I tried lubing the stab on one of my Peerless boards but that actually made it worse D: . Didn't use Krytox though, just PTFE lube.

seebart · 14 Aug 2016, 20:44

Chyros wrote: ↑
seebart wrote: ↑Yes that's why it will be interesting to see how much lubing will help. I may try it on mine but I need every bit of that Krytox for my Alps boards. I'd love to try a NIB FKB4700. Crazy to think that this is the same manufacturer of brilliant swtiches like magnetic reed and leaf spring. And the parent of PFU and the HHKB.
I tried lubing the stab on one of my Peerless boards but that actually made it worse D: . Didn't use Krytox though, just PTFE lube.

I think I remember you talking about that but if we know that NOS boards have been reported to not suffer from this issue it must be either some missing lubricant or worse just abrasion. Fujitsu really designed and built a strange and mediorce switch with really bad aspects to it when they came out with Peerless. Makes BTC dome with slider look brilliant.

drevyek · 14 Aug 2016, 21:39

seebart wrote: ↑either some missing lubricant or worse just abrasion

If it's abrasion, there's nothing to be done. I have a feeling that it isn't, though. If it were, then you would find the numpad keys much smoother than the normal alphas. However, on this board, they are just as bad. The barrels also don't look any more worn down than any of the others.

seebart wrote: ↑ Fujitsu really designed and built a strange and mediorce switch with really bad aspects to it when they came out with Peerless

Honestly. They were clearly trying to copy the M as much as possible, except use rubber dome instead. There are some great aspects to the board, but it feels like you had two teams working, and they came together at the last moment to put them together.
I wish they had uses the spring-loaded stab wire like they did on the spacebar. That one works flawlessly.

seebart · 14 Aug 2016, 21:47

Well I'm looking foreward to your progress, please keep us updated on what you find.

Chyros · 14 Aug 2016, 22:20

drevyek wrote: ↑ I wish they had uses the spring-loaded stab wire like they did on the spacebar. That one works flawlessly.

Yeah, that's definitely one of the best spacebar stabilisers I've used yet, ironic isn't it? xD

dokyun · 15 Aug 2016, 08:35

Definitely gonna pay attention to where this goes. I have an FKB that I've held onto and would love to use...at all, really.

drevyek · 15 Aug 2016, 18:45

If you look at the photo I have labeled "Pressed", you can see something I just noticed. Looking at the holes beside the barrels for the main slider's feet to go through the plate, the holes on the stabilized keys are square, while the normal holes have a notched hole that better matches the sliders themselves. I'll have to look through the sliders to see if there is any variation between them. I'll update when I have wifi

Edit: Nope, doesn't seem to be a different between the sliders. Not sure it would be for then.

drevyek · 18 Aug 2016, 06:09

UPDATE: Moderate success!

I bought lithium grease (only one handy, I'll re-try with Krytox when I can get it), and applied to it the stabs and the legs of the sliders. Applying it to the pins and the barrels of the stabs made it actually slide down properly, sorta. Applying to the sliders of the main switches doesn't seem to work well, however. Makes the switch even gummier than normal. As a result, I only used it on the modifiers, to aid in the stability.

What I mean by sorta is that it isn't perfect. There is still pressure from the pin on the barrel, but the keys don't bind under normal use. Slowly, however, they may (the worst one was the left shift). You can feel that it is much heavier when pressing on the pin-side of the keys, but it isn't unusable.

What is next for this is to re-do the lube-job, and see if I can lube the switch sliders without making them gum up. I then will most likely need to get a hold of a new/NOS dome sheet, as I think that my domes are a bit worn out.

seebart · 18 Aug 2016, 15:29

Hmm interesting, how much lithium grease did you apply? Or rather should I ask how little?

XMIT · 18 Aug 2016, 17:01

I do not generally recommend lithium grease for keyboard work. It doesn't hold up long term. It dries and can be difficult to remove. If someone else has a more positive long term experience with the stuff please let me know.

seebart · 18 Aug 2016, 17:08

XMIT wrote: ↑I do not generally recommend lithium grease for keyboard work. It doesn't hold up long term. It dries and can be difficult to remove. If someone else has a more positive long term experience with the stuff please let me know.

Crossfire wrote: ↑I've used krytox gpl 205 and with combination of bending the 4x teeth on the leaf, the dampened cream alps feel amazing! Smooth, clicky switch. I've only lubed the contact patches between the slider and the housing and it improved the feel drastically. Completely other switch tbh.

keyboards-f2/smooth-alps-t13534.html?hilit=krytox

Of course Crossfire is talking about Alps SKCM and not Fujitsu Peerless.

drevyek · 18 Aug 2016, 19:32

XMIT wrote: ↑I do not generally recommend lithium grease for keyboard work. It doesn't hold up long term. It dries and can be difficult to remove. If someone else has a more positive long term experience with the stuff please let me know.

Doing a bit more searching makes me think that some of the gumminess is not due entirely due to the poor domes, but also to the goop. I had only applied it to the modifiers and a few alphas before reassembling. Testing it, the unlubed alphas are much more responsive and snappy, though not fully. I'm going to clean it all out, and retry with much much less lube, and only on the stab pins this time.

seebart wrote: ↑Hmm interesting, how much lithium grease did you apply? Or rather should I ask how little?

I'm going to say too much. I started small, but the stabs kept catching. So I added a bit more, and they started working much much better. After a few dozen presses, the feel got even smoother (though gummy). I'll bet that using either a lot less, or using different lube will be the correct course of action. When I get back home in September, my buddy has a tube of Krytox, so I can try that out.

Engicoder · 19 Aug 2016, 20:42

I have good condition FKB4700. It is essentially unused, but was stored for a long period of time without a box, so it has accumulated the commensurate amount of dust. The stabilized keys, while much better than heavily used examples, are still uneven in feel, with the stabilized end feeling mushy and a bit gritty. I thought I would follow drevek's lead and have a try at lubing.

I experimented on the right shift as it is the worst offender. If first cleaned the switch and stabilizer to remove any dust using compressed air and a brush.

Lubing the stabilizer
I applied a thin layer of mkawa's "thick" (Krytox GPL 206/1506) mix to the stabilizer mechanism, both the exterior of the pin and interior of the barrel. After working it for several minutes, it was noticeably smoother, but the mushy. uneven feel persisted.

Lubing the slider
I then lubed the switch slider and barrel of the switch with a "thin" Krytox mix. As with other switches, this resulted in less rattling and *may* have improved the uneven feel slightly, it did not resolve the issue.

Conjecture
The design of the Peerless simply has a higher, inherent level of binding between the various sliding parts as compared to the Model M. While cleaning and lubing reduces the effect binding and may help restore a used example to the original feel, the binding can't be eliminated. The devil is somewhere in the details: tapered stabilizer pin and barrel vs straight, square slider vs round, specific plastics used, etc.

Chyros · 19 Aug 2016, 22:35

It's just a fucking shit design, not thought out as well as the Model M, I reckon xD .

seebart · 19 Aug 2016, 22:46

Chyros wrote: ↑It's just a fucking shit design, not thought out as well as the Model M, I reckon xD .

That's what I fear also, and that's also what I found the most shocking in all of this. I mean it's Fujitsu, it's not like Peerless was their first switch. But they really fucked this one up good. We should send them an email and complain.

I want plenty brand new Leaf Spring swtiches for compensation, Magnetic Reed ones for Chyros.

Thanks for the feedback Engicoder!

Chyros · 19 Aug 2016, 23:19

seebart wrote: ↑
Chyros wrote: ↑It's just a fucking shit design, not thought out as well as the Model M, I reckon xD .
That's what I fear also, and that's also what I found the most shocking in all of this. I mean it's Fujitsu, it's not like Peerless was their first switch. But they really fucked this one up good. We should send them an email and complain. I want plenty brand new Leaf Spring swtiches for compensation, Magnetic Reed ones for Chyros. Thanks for the feedback Engicoder!

Indeed, that's the most sour thing of all. We know Fujitsu aren't amateurs, they made some of the best, and well-designed switches out there; absolutely top-tier stuff. You can tell that even for this switch, they tried to come up with something refined — it's one of the most complicated rubber dome designs out there for sure. And it wouldn't even have been a particularly bad one if not for this one issue with the stabiliser they skimped upon, tbh I think they just overlooked it. But if I'm honest, a fantastic rubber dome design doesn't have to be as complicated as this.

seebart · 19 Aug 2016, 23:36

Fujitsu being an old Japanese electronic company they most definitely had their engineer teams look at the IBM Model M before coming to the conclusion that they needed something more complicated to supercede the M which was they're first mistake. Their other swtiches are also quite exotic in design and functionality. Peerless seems "over-engineered" something the Germans also like to do occasionally. The brilliant switches are simple in design 100% serviceable and feel great, production cost being another factor.

emdude · 19 Aug 2016, 23:49

The IBM buckling spring is definitely a great example of wonderfully simple and elegant keyboard switch design; KISS works!

Now Alps.. One wonders why they bothered with complicated switchplates or lube or whatnot, they should have went straight to the simplified design!

seebart · 19 Aug 2016, 23:55

I know you're kidding about Alps SKCM, that switch went from brilliant to mediocre, presumably all for the cost cutting.

emdude · 20 Aug 2016, 00:00

Sadly, that seems to have been the trend for keyboards in general, including buckling springs..

Hm, rubber domes too are relatively simple in design, but that's probably the only good thing you could say about them..

seebart · 20 Aug 2016, 00:05

If done right domes can be OK, but not pure RD like today.

drevyek · 20 Aug 2016, 08:07

Engicoder wrote: ↑I have good condition FKB4700. It is essentially unused, but was stored for a long period of time without a box, so it has accumulated the commensurate amount of dust. The stabilized keys, while much better than heavily used examples, are still uneven in feel, with the stabilized end feeling mushy and a bit gritty. I thought I would follow drevek's lead and have a try at lubing.

I experimented on the right shift as it is the worst offender. If first cleaned the switch and stabilizer to remove any dust using compressed air and a brush.

Lubing the stabilizer
I applied a thin layer of mkawa's "thick" (Krytox GPL 206/1506) mix to the stabilizer mechanism, both the exterior of the pin and interior of the barrel. After working it for several minutes, it was noticeably smoother, but the mushy. uneven feel persisted.

Lubing the slider
I then lubed the switch slider and barrel of the switch with a "thin" Krytox mix. As with other switches, this resulted in less rattling and *may* have improved the uneven feel slightly, it did not resolve the issue.

Conjecture
The design of the Peerless simply has a higher, inherent level of binding between the various sliding parts as compared to the Model M. While cleaning and lubing reduces the effect binding and may help restore a used example to the original feel, the binding can't be eliminated. The devil is somewhere in the details: tapered stabilizer pin and barrel vs straight, square slider vs round, specific plastics used, etc.

Thanks for the testing! I feel like that is about what I found. The binding persists, if at a low level. The stabs pegs have the little slit in them, which I think was intended to reduce the friction on the keys by bending, but the PBT is too tough, comically.
The actual switch is very similar to the M, actually. Especially when you consider how the M actuates: it is not the pressure on the spring that causes it to buckle, but the little tough from the slider on the hammer that pushes it to click over. On these, I suspect that the idea was for the main slider to touch the secondary slider. That then gives the spring the oompf to actuate the rubber dome, giving the "same" tactility as the M, in theory.
In practice, however, the domes actuate irregularly- they aren't touch enough to hold the pressure of the spring until the slider gives the push. And, as with many dome boards, as the domes are used, they get gooey.
That, coupled with the sliders just being garbage just kills these boards for me. I'd love to have seen these done better, with hindsight. The mechanism is complicated, but for good reason. It would have made a really great switch, had they pulled it off.

JCkingp1n · 03 Jun 2018, 03:23

Engicoder wrote: ↑I have good condition FKB4700. It is essentially unused, but was stored for a long period of time without a box, so it has accumulated the commensurate amount of dust. The stabilized keys, while much better than heavily used examples, are still uneven in feel, with the stabilized end feeling mushy and a bit gritty. I thought I would follow drevek's lead and have a try at lubing.

I experimented on the right shift as it is the worst offender. If first cleaned the switch and stabilizer to remove any dust using compressed air and a brush.

Lubing the stabilizer
I applied a thin layer of mkawa's "thick" (Krytox GPL 206/1506) mix to the stabilizer mechanism, both the exterior of the pin and interior of the barrel. After working it for several minutes, it was noticeably smoother, but the mushy. uneven feel persisted.

Lubing the slider
I then lubed the switch slider and barrel of the switch with a "thin" Krytox mix. As with other switches, this resulted in less rattling and *may* have improved the uneven feel slightly, it did not resolve the issue.

Conjecture
The design of the Peerless simply has a higher, inherent level of binding between the various sliding parts as compared to the Model M. While cleaning and lubing reduces the effect binding and may help restore a used example to the original feel, the binding can't be eliminated. The devil is somewhere in the details: tapered stabilizer pin and barrel vs straight, square slider vs round, specific plastics used, etc.

Do you know any reliable sources that can explain how to lube/take apart the switch/slider? I can't seem to find much information on it, and the switch seems to be complicated and I don't want to break it