[Interest Check] Silencing Rings for Topre RF or HHKB

[nipe]

Cool. I've PMed Hyper about it. We could run the GB any time he or you guys like. The final shipping is what will have to wait until next year.

I'm thinking of letting people order as many rings as they like, instead of specifying kits. Something I learned from the Unicomp GB was how much bulkier sub-packaging can make complex orders! I'd like to put everything streamlined in my luggage, when count them out individually when shipping back at home.

Edit: Sounds like he's cool with teaming up. I'll post an IC for this so we can get a handle on how big it will be.

I'm interested. Would it be to much to ask for a PM when you are ready to do this? Since it so far ahead I'm afraid I might forget to check here.

[Hypersphere]

@koalapear: Elegant board!

Regarding Silencing Rings, yesterday and today I installed them in two RF boards: a RF87UB-EK45 and a RF87UB55. I am very pleased with the results on both boards. Between the two, I prefer the sound and feel of the 55g version.

I've also decorated the 55g in a "pseudo skull squadron" scheme:

RF87UB_BRYW1.jpg (313.24 KiB) Viewed 5937 times

This mix of colors is similar to one of matt30's designs for the Topre keycap set. I really like this board; it is currently my favorite TKL. If it were 60% with a HHKB layout, I think I might like it even more than my silenced HHKB Pro 2.

Overall, I now prefer silenced Topre over stock Topre.

[chedda7]

I thought matt3o's Skull Squadron mockup was the best yet for the new Topre set.

[Hypersphere]

I thought matt3o's Skull Squadron mockup was the best yet for the new Topre set.

Yes, I didn't like it at first, but when I put a similar color scheme on my RF, I found that I sort of like it. The scheme is growing on me. I am typing on it now.

Regarding the red accent keys, I would prefer red blanks in the Esc and Right Ctrl positions. (I remap my keyboards to something close to a HHKB layout; on a standard layout TKL board, I use Right Ctrl as Fn).

For some reason, I like blue alphas. I might try these (Along with blue arrow keys) in place of white on the "pseudo skull squadron" configuration.

[chedda7]

I thought matt3o's Skull Squadron mockup was the best yet for the new Topre set.

For some reason, I like blue alphas. I might try these (Along with blue arrow keys) in place of white on the "pseudo skull squadron" configuration.

Be sure to post pictures if you do decide to experiment!

[Hypersphere]

Per request, here is a pic of my "pseudo skull squadron" with blue alphas from a set for a Japanese 108 RF board:

YB50a.jpg (277.77 KiB) Viewed 5878 times

I like a sky blue background on alpha keys to provide a softer contrast than one gets with white alphas. I also like the arrow keys to match the alphas on a TKL board.

[XMIT]

My rings arrived today. This is not bad considering that I ordered them last week. Installation in a Novatouch took 90 minutes. I had cracked the case before to do a dental band mod. This is better than dental bands in that key travel is not nearly as reduced. If you're not removing dental bands and if you are careful to not disturb the springs when disassembling I believe this can be done in under an hour.

They are a huge improvement from a stock Novatouch, eliminating the rattle entirely. I still prefer a real Topre 45g board to the Novatouch even with PBT key caps.

The rings hardly weigh anything. They cost more than their weight in gold for sure. If they were not so pricey (and/or if I were not so stingy) I would order three more sets. Ah well.

Thanks Hypersphere for everything.

[Hypersphere]

@XMIT: You are most welcome. I am glad that you appreciate what the Silencing Rings can do.

My experience with the Novatouch is similar to yours. The rings made a dramatic improvement, but it is still much more satisfying to type on a Realforce. Moreover, my RF87U55 went from refined to sublime with the addition of the Silencing Rings.

The Novatouch requires a lot of tweaking. I still have not found quite the right keycaps for it. The Granite set works well, but I have trouble typing on spherical tops and the flat DSA profile. I think the NT needs a high profile -- SA or at least OEM, but it also needs to be PBT rather than ABS. I have a dye-sub PBT OEM set on order ....

[XMIT]

I've got Vortex PBT double shots (OEM profile) on my Novatouch and it's still a little off. I'm waiting on some lasered PBT Cherry profile keys. Let's see what those do.

[Elk]

It should be noted that placing rings on the sliders under the switch housing will cause a significant reduction in the actuation force amount of work necessary to activate switches, the more so with increased thickness of the rings used. I believe a mod should be designed to elevate the switch housing board if one wants to retain the same actuation force amount of work to reach actuation point.

[Muirium]

These rings are paper thin, and do not noticeably reduce travel or actuation force. Hyper did a great job researching them. Maximum damping for minimum thickness.

[XMIT]

It should be noted that placing rings on the sliders under the switch housing will cause a significant reduction in the actuation force necessary to activate switches, the more so with increased thickness of the rings used. I believe a mod should be designed to elevate the switch housing board if one wants to retain the same actuation force.

I measured these rings as 0.50mm thick just now with my cheap digital calipers. The dental bands they replaced are 1.12mm thick.

Recall the Topre force diagram:


Adding a ring does not alter the actuation force (does not move the curve up or down). What it does is slide the release position to the right along the X axis by 0.50mm. The work required for actuation (area under the curve from release point to actuation point) is reduced. This may make the switches feel "lighter".

As for the mod: I believe you could experiment with 0.25mm thick washers between the switch plate and the PCB. You would need 32 of them, one for every screw. I don't have a part number, or really any intention to try it myself, but it is an idea.

[Hypersphere]

@XMIT: Indeed. Thanks for reminding us of the difference between force and work. F = ma, Work = Fx, where F = force, m = mass, a = acceleration, x = distance. We might add Power to this (the rate of doing work).

In practical terms, I have been unable to notice any detectable difference in key travel or effort required to actuate switches between Topre switches with or without Silencing Rings installed.

[XMIT]

One factor here - particularly with the Novatouch - is that the 0.50mm rings serve as shims. They eliminate rattle by removing some amount of play between the key slider and the key switch housing. Exactly what this play is, I do not know, but my guess is that it is on the order of 0.05-0.10mm, which happens to be about the thickness of a typical sheet of paper.

So another experiment we might run is to try paper rings. Archival quality paper could do reasonably well here. I don't think the silencing would be nearly as good as the foam rings but it could eliminate the rattle.

Authentic Topre keyboards seem to have tighter tolerances and typically don't have any play whatsoever at the top of the key stroke. This is what I've observed on the 45g and 55g Topre keyboards in the house.

FWIW, MrsXMIT certainly prefers the silenced Novatouch, but doesn't care enough to have me silence her 45g Topre board.

[Muirium]

She's quite right. The NovaTouch is cured from its wicked curse. The Realforce, well, wasn't so bad to begin with.

[Hypersphere]

@XMIT: I haven't tried paper, but I've tried thinner rings of various polymers and composites at thicknesses of 0.10, 0.15, and 0.30 mm. I was not satisfied with reduction of both rattle and clack until I got to 0.50 mm plus some resiliency.

@Muirium: I agree. The effect on the NT is fairly dramatic. Certainly, the RF is less in need, but I have still been bothered by some clack from my RF boards, and the addition of Silencing Rings elevated them from refined to exquisite. Note that I also lubed the stabilizers and added self-adhesive 1 mm thick foam pads to the top of the case where the spacebar lands. I used SuperLube silicone oil instead of grease for the stabilizers, because I found that the grease dries out after a relatively short time.

I've got Vortex PBT double shots (OEM profile) on my Novatouch and it's still a little off. I'm waiting on some lasered PBT Cherry profile keys. Let's see what those do.

I tried the Vortex PBT double shots as well. These were about the best I've tried thus far on the NT in terms of being high enough to avoid bottoming-out click noise, but I didn't like the fit and finish of the caps. Therefore, for aesthetic reasons, I removed them. I've also tried the Leopold PBT sets, which look great, but the profile is too low. Same with thick dye-sub PBT in Cherry profile -- still a bit too low. I think the NT will need OEM or SA profile (except that the Granite set, which is DSA profile, worked well in terms of clearance, but I had trouble typing on sphericals with a flat profile).

[Elk]

Ah, physics owned me again.

RIght, actuation point of the switch is not modified - unless one uses too thick rings which compress the dome - but your graph shows that at no point the required force to move a slider is null. Help : how would you call the total required 'effort' to move a slider until it reaches its actuation point (from 0% to 100%) and the letter on my key cap shows up on my screen ?

[XMIT]

Help : how would you call the total required 'effort' to move a slider until it reaches its actuation point (from 0% to 100%) and the letter on my key cap shows up on my screen ?

I would call this work - force times distance. (Well, force cross distance if we're dealing with vectors.) This happens to be the area under the curve from the starting point to the actuation point. You can do more work to get beyond the actuation point but it does not result in any further key action.

[Hypersphere]

Ah, physics owned me again.

RIght, actuation point of the switch is not modified - unless one uses too thick rings which compress the dome - but your graph shows that at no point the required force to move a slider is null. Help : how would you call the total required 'effort' to move a slider until it reaches its actuation point (from 0% to 100%) and the letter on my key cap shows up on my screen ?

As XMIT pointed out, this would be "work", which is force x displacement (in the direction of the applied force). The units of force are Newtons; the units of work are Newton-meters = Joules.

[AdamVIP]

Well I ordered the 107 set for my type heaven. Cant wait to get this board to office stealth mode. Now If it was just easier to get cool topre keycaps.

[Elk]

My HyperRings(tm) should be arriving next week, I hope they are as good as they are hyped up to be.

Hypersphere, can you provide the Shore value of the ring material used ?

Thanks.

[Muirium]

All the HyperHype is true… in the NovaTouch. I haven't enough rings yet to try another board, and the NovaTouch needs them so bad that these ones ain't coming out!

[Hypersphere]

@Elk: I don't have a Shore value and I haven't attempted to measure it. The rings have two-ply construction with a resilient layer bonded to a relatively hard supporting layer. I tested various single-ply and two-play materials by installing them side by side in different columns in a HHKB Pro 2 keyboard and assessing attenuation of noise, focusing on the return stroke. Subsequently, I have installed the rings in a HHKB Pro 2, Novatouch, and RF87U. I've been quite pleased with the results. IMO, the HHKB Pro 2 with the HyperRings sounds and feels better than my Type-S HHKB Pro 2, but of course this is a subjective opinion subject to considerable bias!

[RexS]

@AdamVIP I installed a set of these last weekend on a variable weight Type Heaven and was blown away by the improvement. It honestly made me rethink my whole Topre strategy.

Before the silencing rings I always compared the stock 45g Type Heaven I use at home to a 45g Realforce at work. After I installed the rings all plans changed. First, I was reminded why I don't care for the variable weight rubber: just leaning in resting my hands on the keyboard sends w's and a's flying across the screen. Me and 35g just don't play well together. After only a day on the variable weight Type Heaven I swapped the silencing rings to the 45g.

Second, I cancelled a delayed back order for a Realforce to replace the Type Heaven at home. The plan was to swap their domes to make the RF uniform 45g as well. But in the days since installing the rings on the Type Heaven I'd noticed just how clacky my Realforce at work is, and I honestly wanted additional silencing rings more than I wanted a second Realforce (especially one with variable weight and no silencing rings!)

Then something magical happened. I found myself in a position with access to a fixed beam CO2 laser and a few samples of the same 0.5mm poron used for these rings. I missed the first round sale of the rings and literally the day before Hyphersphere announced that the second round was available (ahead of schedule and at a new lower price) I'd made inquiries that I thought were dead ends. As soon as that second round was posted to eBay I pulled the trigger and forgot about all about the inquiries from the day before -- until I came home to a package of uncut material waiting for me the very next day!

So last night, after hours spent adjusting step motor speeds, attenuating laser power, and varying the number of passes we finally found a combination that made its way through the material without cooking it. My laser buddy is a Cherry MX guy; he had an interesting idea of adding an additional 5.5mm cut to the slug left from cutting the Topre ring leaving smaller rings for him to trial as an alternative to the o-rings on his Cherry key caps. The three-pass configuration we settled on took a little over an hour to cut (including the extra cut for the Cherry). The final tally, after a slow tedious process of hand separating the rings and slugs, was about 350 rings.

For me, this exercise has only reinforced the value in Hypersphere's rings. I had fun learning how to replicate the process, but producing these rings requires expensive, hard-to-source material, laser beams, plenty of time, and patience. I also believe the craftsmanship of Hypersphere's rings is superior. Under a microscope my rings showed more thermal damage than his (that polyester support layer is super strong but it also melts really easily).

I hope all you folks outside the US (where shipping/fees are prohibitive) are soon able to acquire a set of Hypersphere's silencing rings. They really are transformative. I spent the morning installing them onto my Realforce for work and already I'm looking forward to Monday. I've got my 104-key teardown/rebuild technique down to about two hours.

But now that the clack's been curtailed I should silence those squeaky stabilizers. It never ends.

[AdamVIP]

Great to hear that the type heaven took well to them. I avoided the the vari-weight myself. In all honesty I'm not a great typist, I just have to do it a lot for work so I justified the purchase. I still do the majority of my typing with my index and and middle fingers. I didn't like the black on black caps for the realforce so got a Type Heaven with the intention to swap out to some PBT caps. Sadly I didn't really look up what that entails (availability wise) so I just have the ABS for now.

At home I have a WASD Code with cherry clears. Its nice but even louder than the topre and the stiffness doesn't have a nice break point like the topre's do. I have plans and the springs/extra LEDs on hand to make it an ergo clear board but I don't look forward to unsoldering and resoldering all those LEDs.

Side note to those still reading. I ordered my rings and got a shipping notice that day from ebay so customer service so far is AAA+.

[Hypersphere]

Glad to know that the Silencing Rings continue to work well for various Topre-switch keyboards. They certainly worked wonders on my Novatouch, and now that I have OEM profile PBT caps on it as well, it is almost as nice as my silenced RF87UB55.

Lubing the stabilizers also helps, but I am still searching for the right lubricant. I've noticed when I open keyboards that have been factory lubed that the white grease on the stabilizers has dried and started to crumble. In view of this, the last time I applied lube, I used SuperLube oil instead of grease, but presumably this would not stay put very well. I've ordered some Krytox GPL 205, which I may try next, although after ordering this, I saw some posts recommending 206 or 207.

To help silence the spacebar, I put small squares of 1 mm thick self-adhesive foam rubber on the top case under the central spacebar struts. This yields a dull thud instead of a loud click when the spacebar is pressed.

The only stabilizer noise that I now notice on the Novatouch is from the Backspace (which after my remapping is the Backslash key -- I seldom use it) and Right Shift (which unfortunately I use quite a lot).

[vinzbe]

Why not the EK MechLube 2 instead of Kryptox GPL 20x?

[Hypersphere]

Why not the EK MechLube 2 instead of Kryptox GPL 20x?

I had been using EK MechLube 2 for lubing stabilizers, but I noticed that this lube, like whatever factory lube I have found in various keyboards, dries up and crumbles after some time. I haven't yet had the opportunity to try Krytox GPL 20x to see how well it works and how well it holds up over time. It would be good if someone could do a systematic test of various teflon, lithium, molybdenum, and silicone greases or oils, evaluating them for efficacy and longevity in helping to silence stabilizers with continuous long-term use.

[vinzbe]

I ordered some from techkeys. When adding the silencers I would also apply the lube.

[XMIT]

I agree. Too much of what we hear about lubrication is based on pseudo-science. This is an issue in other fields as well.

My love of Finish Line Extreme Fluoro (a PFPAE based grease, a much more affordable alternative to Krytox as is it nothing more than rebadged Krytox) comes from this article:

http://www.sandsmachine.com/grease_t.htm

Now, this is for metal-on-metal contact, but as you can see, nothing comes close.