Topre rubber dome stiffening

[limmy]

Observations :

HHKB pro2
manufacturing date 2006-08
5 years of use

Left Alt key (used as "Fn" key) - 50g not enough to actuate, 55g enough to actuate and bottom out

[ Attachment Invalid Or Does Not Exist ] 40969[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40970[/ATTACH]

without sliders - 50g not enough, 55g enough

[ Attachment Invalid Or Does Not Exist ] 40975[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40976[/ATTACH]


"a" key - 60g not enough to actuate, 65g enough to actuate and bottom out

[ Attachment Invalid Or Does Not Exist ] 40971[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40972[/ATTACH]

without sliders - 55g not enough, 60g enough

[ Attachment Invalid Or Does Not Exist ] 40977[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40978[/ATTACH]


HHKB pro2
manufacturing date 2009-02
Like new condition

Left Alt key (used as "Fn" key) - 45g not enough to actuate, 50g enough to actuate and bottom out

[ Attachment Invalid Or Does Not Exist ] 40979[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40980[/ATTACH]

without sliders - 40g not enough, 45g enough

[ Attachment Invalid Or Does Not Exist ] 40981[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40982[/ATTACH]


"s" key - 50g not enough to actuate, 55g enough to actuate and bottom out

[ Attachment Invalid Or Does Not Exist ] 40973[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40974[/ATTACH]

without sliders - 45g not enough, 50g enough

[ Attachment Invalid Or Does Not Exist ] 40983[/ATTACH] [ Attachment Invalid Or Does Not Exist ] 40984[/ATTACH]



Interpretation of the above results :

First and foremost, this shows Topre switches' rubber dome changes over time (Edit) under the assumption that my like-new HHKB is very similar to my old HHKB 5 years ago. I also swapped the rubber domes (I swapped whole board) and I get the same measurements. This suggests that rubber dome is main reason for the changed actuation force and that wear and increased friction of the sliders are negligible once they are lubricated. (I used dry teflon to lubricate the sliders after washing them.)

Also I make the following observations and some deduction from them :

1. Frictions from sliders add about 5g more force for keys to actuate. However, it is hard to rule out measurement error. The weights were unstable when I measured on rubber domes without sliders hence I had to hold the weights affecting the accurateness of the measurements.

2. letter keys become harder to actuate compared to other modifier keys. I thought of three possible explanations.
First, fast press and release action stiffens rubber more than constant pressure action.
Second, constant pressure action wears the rubber so much that it is at the stage of loosing elastic property. (Some say very old Topre switches that has seen a lot of usage feels very light and force curve seem to feel like there is very little hump, i.e. little tactile feel of rubber. So if their observations are correct, the rubber dome will soften once aging goes beyond stiffening stage.)
Third, more force is required to depress larger piece of rubber than a single rubber dome because there is more friction in the air duct through which air inside rubber dome passes when it is depressed. (I tested this hypothesis but results are rather inconclusive. However, I wonder if punching holes in the air duct (not the dome itself) will help reducing actuation force)


NOTE1:
Unit of measurements being 5g is not accurate. I am pretty sure that the measurement above are with measurement errors, especially the rubber dome measurements(without sliders).

There were a couple of problems using nickels as weights. Due to its shape, it was hard to position the weights consistently. It was especially hard to do measurements on rubber domes without sliders. The problem also exsits on measurements on key caps. The size of nickels are large enough that it interferes with other keys. For correct measurement, I would recommend taking surrounding keys off. Moreover, the position of weights also matters. If I were to do the measurement again, I would take the caps off and place the weights so that the center of the mass is exactly where the sliders are. However, this measurement would be biased downwards, because in normal use you do not press keys in a straight way or in a way that causes least friction.

NOTE2:
ripster mentioned that my comment(http://geekhack.org/showthread.php?27468-45g-or-55g-87u&p=517630#post517630) was first actual measurement of rubber dome stiffening. If it is the first then I wanted to document what I had done to come to that conclusion and provide proper evidence.

NOTE3:
The old HHKB doesn't feel bad compared to the like-new HHKB. I actually love to use the old HHKB as increased actuation force provides more tactile feel. However, due to the increased force, my left-hand fingers(esp. little finger) gets tired after prolonged use. The feeling of tiredness may partly be due to the fact that I have become used to Poker reds which actuates at about 40g to 45g. I have two HHKB and a Poker red. I have been using the old HHKB for 5 years as a daily driver. I use Poker time to time, but I mostly use the old HHKB.

[mrflow3r]

Thanks for the info =). Can we assume other types of mechanical keyboards also stiffens over time?

[flyball]

Thanks for the info =). Can we assume other types of mechanical keyboards also stiffens over time?

thats a bad assumption

[HaiiYaa]

Thanks for the info =). Can we assume other types of mechanical keyboards also stiffens over time?

I think its the rubber that stiffens in the topre keyboards so I don't think we can assume that

[patrickgeekhack]

This was something that worried me a bit when I was deciding whether to buy a Topre or not. I do think that the keys on my Topre is slightly stiffer now compared to when I got it. It's not to the point where I do feel a big difference.

This was also the reason why I decided to use it as my primary one at home. I figured out that I might as well get as much use out of it considering I paid quite a lot for it compared to my other keyboards.

[seferphier]

that's very disappointing...

thought topre keyboards are suppose to last for decades..

[keyboardlover]

GLWNT.

They don't consider this scientific in the Netherlands.

[shogrran]

Hmmm... well... i was wondering if we can use rubber conditioners like the ones used in table tennis rubber to bring back some if not all of the rubber's freshness. Or how about re-heating rubber... anyone crazy enough to try that yet?

[o2dazone]

Nice write-up, pretty cool to see that after 5 years, it's only changed 5g. I'll take slightly more force over sloppy rebound or gritty switches any day.

[RiGS]

Another reason why Cherry linear switches are superior. Of course it is only true for the vintage ones, the recent ones feel like garbage.

[o2dazone]

Vintage linears like ML's and MY's? I haven't tried those, I heard they felt awful. Is this not true?

[REVENGE]

Vintage linears like ML's and MY's? I haven't tried those, I heard they felt awful. Is this not true?

He's talking about vintage Cherry Blacks.

[RiGS]

Indeed, though I think thick PBT keys, some lubricant, stickers and somewhat lighter custom springs make the feel even better.

[limmy]

Nice post.

Added to RipOmeter wikis.

Too bad nickel pics are banned at other Keyboard Forums like Deskthority.net.

In your RipOmeter wiki, you used my results that are most prone to measurement error. As I mentioned in the OP, I had to touch the weight so the accuracy of the measurement is somewhat questionable. I included the results to support that indeed the main problem is rubber domes, but I don't consider that as the main evidence of rubber domes stiffening.

The increased actuation force is pretty significant. If it was designed to actuate at 45g, actuation force at 60g is 33% more, 65g is 44% more, 70g is 55% more force than the designed actuation force. Although I didn't mention in the OP, the letter keys require 60g - 70g to actuate.

I think unit of measurement being 5g is not accurate enough. Let's say actual force required is 50.5g. Using the method, it would show as 55g as actuation force (if everything was done right and measured several times). This is 9% error that cannot be reduced. This in my opinion is huge. Maybe for the purpose 5g unit is fine, but it surely can be improved(by using pennies at the last stage) and the possible errors using the method should be recognized rather than ignored away.

[RiGS]

See my method for more accuracy.

[jkercado]

Vintage linears like ML's and MY's? I haven't tried those, I heard they felt awful. Is this not true?

I dunno, I love ML's. I have four Cherry G84-4100 units, and one of them is my daily coding keyboard at the office. I love it!

[RiGS]

If you were a man of Keyboard SCIENCE for real, then you would consider evaluating other member's observations.
Sadly that's rarely or never the case.
In fact it is very common that a few different members notice something irregular about a their keyboard parts, switches, etc, so they share their thoughts about it,
but you simply choose to ignore their findings without even bothering to try or evaluate it by labeling it with usually something nonsense like the McRip Effect.
Then you keep annoying those members by repeating your boring terms like a parrot.

In my book this is trolling, and definitely not Keyboard SCIENCE.

Just a quick reminder in case you forgot why you are not an expert, and why you never will be. Period.

[pitashen]

nice post and good work. But it shouldn't come as a surprise that rubber eventually degrades and lost its original properties.

[limmy]

Hmmm... well... i was wondering if we can use rubber conditioners like the ones used in table tennis rubber to bring back some if not all of the rubber's freshness. Or how about re-heating rubber... anyone crazy enough to try that yet?

I like the idea of rubber conditioners. Can you recommend one? If rubber conditioners fail, I might try punching small holes. I see there is a hole puncher that makes 1.2mm holes.

About heating rubber, I think it would damage or further degrade rubber. It may help, but it somehow feels dangerous to me.

[F u r u y á]

Great investigative work. Thanks for sharing it.

It would be nice if Realforce supplied replacement rubber domes.

[patrickgeekhack]

nice post and good work. But it shouldn't come as a surprise that rubber eventually degrades and lost its original properties.

That's my feeling too. Rubber will degrade over time, no question about it. Nice article though.

[limmy]

EK is offering a sale now where I can buy a 103U for $230, however in the back of my head I really want an 86U.

This thread sort of worries me and makes me wonder if I should buy "as fresh as possible."

Should I be able to buy a preowned 86U and not experience any issues?

Before you answer and say that KeyboardCo is getting more in: as far as I know they are all going to be black 86's: I prefer to stick with a white board.

I think the rubber stiffening is less of an issue than friction build up on the sliders. I say this because I don't really feel discomfort due to increased force required to actuate keys after lubing the sliders with high-end oil. I suspect friction plays very important role.

If you can get a fresh batch, it would be always better.

BTW, ripster. You said you would check the plastic of realforce switch housing. I personally think this is important matter because ABS is bad for places where friction is important. Did you get to the to do list yet?
( http://geekhack.org/showthread.php?26646-Is-key-switch-housing-of-Topre-Realforce-ABS )

[seferphier]

EK is offering a sale now where I can buy a 103U for $230, however in the back of my head I really want an 86U.

This thread sort of worries me and makes me wonder if I should buy "as fresh as possible."

Should I be able to buy a preowned 86U and not experience any issues?

Before you answer and say that KeyboardCo is getting more in: as far as I know they are all going to be black 86's: I prefer to stick with a white board.

they are going to sell more 86u?!

[elitekeyboards]

First and foremost, this shows Topre switches' rubber dome changes over time.

Wow, I missed this thread. Sorry to say, but this is rather hokey science!

Yes, I distribute these products, so while I can be accused of bias; as an individual, I would be embarrassed to put my name on this bit of deductive logic.

Unfortunately, if you didn't test your 2006 HHKB weightings when you first bought it, there is no record that they've actually changed over the last 6 years. So, the data required to deduce a conclusion of the sort you seek to make - that the rubber changes over time - is missing.

I'm not saying that isn't a worthwhile theory to test, but you actually have to test it first!

At the most, all you have demonstrated is that keys from two keyboards from the same factory vary in weighting, and in this case, both are within the specification set by Topre, which is +/-15 grams. In case you're wondering; this tolerance is relatively wide because multiple variables contribute to the performance of the keyswitch; not limited to: small variations in material composition, molding pressure, and environmental temperature.

NOTE3:
The old HHKB doesn't feel bad compared to the like-new HHKB. I actually love to use the old HHKB as increased actuation force provides more tactile feel. However, due to the increased force, my left-hand fingers(esp. little finger) gets tired after prolonged use. The feeling of tiredness may partly be due to the fact that I have become used to Poker reds which actuates at about 40g to 45g. I have two HHKB and a Poker red. I have been using the old HHKB for 5 years as a daily driver. I use Poker time to time, but I mostly use the old HHKB.

This is a common experience for All-55g Realforce users, but one that is overcome with a week or two of normal usage; your fingers do get stronger.

[limmy]

Wow, I missed this thread. Sorry to say, but this is rather hokey science!

Yes, I distribute these products, so while I can be accused of bias; as an individual, I would be embarrassed to put my name on this bit of deductive logic.

Unfortunately, if you didn't test your 2006 HHKB weightings when you first bought it, there is no record that they've actually changed over the last 6 years. So, the data required to deduce a conclusion of the sort you seek to make - that the rubber changes over time - is missing.

I'm not saying that isn't a worthwhile theory to test, but you actually have to test it first!

At the most, all you have demonstrated is that keys from two keyboards from the same factory vary in weighting, and in this case, both are within the specification set by Topre, which is +/-15 grams. In case you're wondering; this tolerance is relatively wide because multiple variables contribute to the performance of the keyswitch; not limited to: small variations in material composition, molding pressure, and environmental temperature.



This is a common experience for All-55g Realforce users, but one that is overcome with a week or two of normal usage; your fingers do get stronger.

Should I be embarrassed!! Oh my..

All rubbers degrade and harden over time. Here is how EPT(same as EPDM) degrade over time in vitro environment. http://www.tiger-poly.com/product/sheets/r08.html In aging tests, most of EPT hardens and loose elasticity. You may argue that it is not the same rubber used in Topre, but I would argue that this is general enough result of rubber aging. In case you didn't know, the rubber used in EPT as mentioned in Topre's patent.

Your claim that I had to measure the force in 2006 to conclude that the feel changes is not very persuasive. Also +/- 15 grams error isn't error if every measurement is pointing towards +15g. 15g difference by the way is huge when base is 45g. (it is what 30% upward bias)

[keyboardlover]

Stiffening is due to an increase in friction.

For more info see ripster and I's newfound friendship

[elitekeyboards]

Should I be embarrassed!! Oh my..

All rubbers degrade and harden over time. Here is how EPT(same as EPDM) degrade over time in vitro environment. http://www.tiger-poly.com/product/sheets/r08.html In aging tests, most of EPT hardens and loose elasticity. You may argue that it is not the same rubber used in Topre, but I would argue that this is general enough result of rubber aging. In case you didn't know, the rubber used in EPT as mentioned in Topre's patent.

Your claim that I had to measure the force in 2006 to conclude that the feel changes is not very persuasive. Also +/- 15 grams error isn't error if every measurement is pointing towards +15g. 15g difference by the way is huge when base is 45g. (it is what 30% upward bias)

I wasn't trying to be persuasive. I'm just telling you that you're doing it wrong.

Regardless of how much empirical data you have on the particular material used in Topre domes; you cannot conclude that the domes have changed even 1% from 2006 to 2012 if you never measured them in 2006!

And while anyone would probably agree that +/-15g is a huge tolerance relative to 45g - your audience isn't going to be very amenable to your theory while the measured difference still falls within this range.

About all you might have to hang your theory on is your EPT data. However, the data doesn't express how "Heat-Resistant Aging @ 100°C×72h" measurements correlate to changes that will occur in a normal operating environment; i.e. 25C at 50% humidity for 1 year, 10 years, or 50 years. Which is why this type of test data is typically only useful as a reference in comparing one material to another; it's otherwise of limited value in determining long term performance.

[jrockroll]

yea stiffness occurs due to the some env temperature i heard

[Soarer]

Your claim that I had to measure the force in 2006 to conclude that the feel changes is not very persuasive.

No, EK has a valid point there. Even if you measured many HHKB from 2006, and many from 2011, and found a statistically significant difference, you still could not conclude the cause of the difference.

[Jim66]

No, EK has a valid point there. Even if you measured many HHKB from 2006, and many from 2011, and found a statistically significant difference, you still could not conclude the cause of the difference.

Correct, too much unsystematic variance present in: a) actual actuation tolerance and b) the ripometer.

+ unless you collected data from a normally distributed population of NIB 2006 and 2011 HHKBs (i.e., the control condition) you findings don't really tell us anything.

[o2dazone]

Not to fuel any fires, but finger rolling on my heavily used (8 hours a day programming) HHKB for 2.5 years doesn't feel any different than a brand NIB HHKB. Keys are a bit shinier though Yes even the PBT gets shiny.

[limmy]

I posted some evidence. It may not be accurate but still an evidence. If you want to refute my point that rubber domes harden over time, I will readily accept the counter argument if you would provide some measurement and evidence.

I didn't know I would have to be super scientific to prove one point that is so natural. Things deteriorate when they are used 5-years! Two of the deterioration process of rubbers are hardening and loss of elasticity. One assumption I made in order to reach my conclusion was that my new HHKB was similar to my old HHKB back in 2006. I think it is a fair assumption. If you trying to argue against my conclusion, your argument will be much more persuasive if you provide something to prove your point rather than nit picking on my methodologies. You can argue against any study all day long saying something was not done right, but that is just lame way of refuting.

I referred to aging test results used in industry and you even question that. It is standard procedure to heat in order to accelerate aging process. This is so that you don't have to wait years to find out how rubber materials age. It was to show that EPT age and when they age what properties are changed. You are right that the amount of deterioration may or may not mean much in normal conditions, but it is nearly impossible to get that result.

Since it hardens and at the same time loose elastic property, the rubber domes would later be mushy if significant amount elastic property is lost. My HHKB in my opinion has not reached that point yet.

[Soarer]

I posted some evidence. It may not be accurate but still an evidence. If you want to refute my point that rubber domes harden over time, I will readily accept the counter argument if you would provide some measurement and evidence.

What you have there is a finding, not evidence. No-one is saying that rubber doesn't harden over time, just that your test does not prove that it does. Too many variables are unaccounted for, since it is only one test. To counter your conclusion therefore does not require measurement and evidence.

Don't get me wrong - the finding is interesting by itself. Had you found no difference, there wouldn't be much point in going any further.

But the finding needs to be replicated to become evidence, and even then it would only be evidence of a difference. The differences would need to be analysed to determine possible causes, and the hypothetical conclusions would need to be tested to determine their validity.

[Typhaeon]

All this skepticism assumes this one case of materials testing is being done in a complete vacuum independent of any other results regarding the tensile properties of rubber.  Which is silly.

[limmy]

What you have there is a finding, not evidence. No-one is saying that rubber doesn't harden over time, just that your test does not prove that it does. Too many variables are unaccounted for, since it is only one test. To counter your conclusion therefore does not require measurement and evidence.

Don't get me wrong - the finding is interesting by itself. Had you found no difference, there wouldn't be much point in going any further.

But the finding needs to be replicated to become evidence, and even then it would only be evidence of a difference. The differences would need to be analysed to determine possible causes, and the hypothetical conclusions would need to be tested to determine their validity.

Agreed. It is a finding, and it may or may not be generalized. (I think one finding could be evidence of some phenomenon, but your terminology of evidence may be different from mine)

Although it would be hard to draw hard conclusion from my observations, I think it is at least indicative. That is what one consumer can do at most in most cases, not having power and resources to test it rigorously. Moreover, the consumer has very little incentive to publicize some findings, when it is likely that resale value is going to be negatively affected. So lack of negative information doesn't necessarily mean it doesn't exist. (There is whole literature on incentives to report incorrectly or not to report)

Now that many questions my methodologies, I will document how my new 2009 HHKB ages over time in normal conditions. I hardly feel the necessity, but I will do it to convince you guys. I will come back with more data for you in following years.

[Soarer]

Always a tough crowd here - just ask ripster :-)

Manufacturing variation is the other main suspect, I guess. Some idea of whether Topre's +/- 15g is random, or biased at times (e.g. some years different to others) would be useful.

[Surly73]

I posted some evidence. It may not be accurate but still an evidence. If you want to refute my point that rubber domes harden over time, I will readily accept the counter argument if you would provide some measurement and evidence.

I have no real investment in this debate either way, but wanted to inquire about how the "hardening" of the material due to its properties is countered by the softening due to repeated flexing if we assume that the keyboard is actually being used over those years.  There are threads upon threads here with people talking about how their RD keyboards get softer and mushier as they wear in/out.  Why wouldn't this apply to Topre?  I'd expect the softening to be a larger effect than the hardening, but admit that's just an expectation with no observations.

[limmy]

I have no real investment in this debate either way, but wanted to inquire about how the "hardening" of the material due to its properties is countered by the softening due to repeated flexing if we assume that the keyboard is actually being used over those years.  There are threads upon threads here with people talking about how their RD keyboards get softer and mushier as they wear in/out.  Why wouldn't this apply to Topre?  I'd expect the softening to be a larger effect than the hardening, but admit that's just an expectation with no observations.

They do get mushy according to 002 of Deskthority source But the keyboard is very old. I heard also from Korean community members that HHKB displayed in a store also felt mushy.

When rubbers age, it hardens and at the same time loose elasticity. One easy example of it would be very old rubber band. (The surface will harden and it doesn't pull back as hard when stretched.) If large enough elasticity is lost due to repeated use or age, I would say the switch is going to feel mushy. So far my HHKB doesn't feel mushy after 5 years of daily use. So, I am guessing there is some life left of my HHKB after 5 years.

[limmy]

RipOmeter my neglected stored in box 87U indicates it's 5g stiffer than when I tested it last time.  In other words measuring 50g on the F key.

I'll need to plug it in tomorrow, warm it up for a day or two, and retest.  I bet it'll drop down to 45g levels.

Thanks for the input.

If you could, for the sake of completeness, please provide more information on the keyboard such as
1. when it was purchased and
2. how long has it been used

[Typhaeon]

I'll get some nickels and run tests on my RF when it arrives to actually contribute to the thread.

[Soarer]

I'll get some nickels and run tests on my RF when it arrives to actually contribute to the thread.

And please repeat the test after a couple of years hard use. This could be a long thread! :-)

[Aranair]

wait, so does it actuate at 52g? (Mine is still 60+ :smow:)

I also have a feeling all those "burn it in for a week" perceptions are around because their fingers are slowly getting used to the weight and not the domes softening back.

Anyhow, my topre switches needs to soften up or I get used to it, before I get RSI lol.

[Aranair]

I'm using wikipedia for my country's coin weights lol. No other way to find out unless I buy a small weighing machine heh.

However, seeing my cherry browns are using 45-50g to actuate, I'll say its relatively accurate.

Oh well, thanks anyways. I'll wait till monday to decide if I should offload it lol. (/wails at losing the layout)

[limmy]

So Girlz now that you little spat is hopefully over.... these are the numbers:

50g
(Attachment Link) 44144[/ATTACH]

or around 52g using 1Yen 1.00g pieces if you want Webwit false precision (see Elitekeyboards post earlier in this thread before all the bickering).
(Attachment Link) 44143[/ATTACH]

I'm typing on it now after a week on the Cherry Reds.  It does feel a bit stiffer but if it's anything like last time in a week my fingers will adjust.

I AM curious to see if the RipOmeter numbers come down though.  They were 45g when I measured it here:

The Invention Of The RipOmeter - A MOMENTOUS Occasion In Keyboard Science

http://geekhack.org/showwiki.php?title=Island:5697&do=comments&page=6
(Attachment Link) 44146[/ATTACH]

Thanks for the update ripster!

To say it in other words, 50g was not but 52g was enough to actuate switches that were designed and were actually actuated at around 45g about three years ago.

These measurement is pretty similar to what I measured with HHKB that was produced in 2009.

"s" key - 50g not enough to actuate, 55g enough to actuate and bottom out

(Attachment Link) 40973[/ATTACH] (Attachment Link) 40974[/ATTACH]

I don't know if it would generalize, but it seems that the rubbers become slightly stiff even if you don't use it.

About feeling of stiffness, I think the feel has a lot to do with friction build up. Effects of friction are greatest if you trying to type fast. If friction is built up on switches then lightly tapping on the switch will not depress it and it will only depress if enough force is applied, which is increased due to the friction. This is for a brief moment like tapping on a hard surface which will give you fatigue. The increased force required to actuate didn't really bother me once I lubed the switches. However, it was rather expensive journey for me to find the just right lubricant that would perform well. I am currently trying out Krytox oils, and I like them so far. I will see how long the effect of the oil last.

So far I tried KG8 and silicone oil on my 5-year old HHKB. KG8 worked pretty well in reducing friction, but the effect wore off quite quickly. Silicone oil worked well for a while but it gave me sticky feeling after a while. I am currently trying out Krytox for two weeks and it is working well so far. The Krytox is very expensive though.

[Surly73]

About feeling of stiffness, I think the feel has a lot to do with friction build up. Effects of friction are greatest if you trying to type fast. If friction is built up on switches then lightly tapping on the switch will not depress it and it will only depress if enough force is applied, which is increased due to the friction. This is for a brief moment like tapping on a hard surface which will give you fatigue. The increased force required to actuate didn't really bother me once I lubed the switches. However, it was rather expensive journey for me to find the just right lubricant that would perform well. I am currently trying out Krytox oils, and I like them so far. I will see how long the effect of the oil last.

So far I tried KG8 and silicone oil on my 5-year old HHKB. KG8 worked pretty well in reducing friction, but the effect wore off quite quickly. Silicone oil worked well for a while but it gave me sticky feeling after a while. I am currently trying out Krytox for two weeks and it is working well so far. The Krytox is very expensive though.

I agree - friction has a totally different feel than spring/dome weight.  Low friction and no off axis binding are two of the things I love about Realforce.  Have you ever tried a PTFE "dry" lubricant?

[seferphier]

no wonder why i feel like topre's feels different, every time i press it. sometimes i love it, sometimes it is just too stiff...

[limmy]

Here are my measurements of my 5-year old HHKB which was used daily since I bought them. I measured only the letter part of the keyboard. That is..
QWERTYUIOP[]
ASDFGHJKL;'
ZXCVBNM,./

Most of my keys support 60g without actuating.
When I put 62.5g, only 6 keys actuate. The keys that actuated at 62.5g were mostly non-alphabet keys.
Putting on 65g actuate 19 keys out of 33 keys that I test.
67.5g actuate all of the 33 tested keys.

I built custom weight system similar to webwit's(50g calibration weight + attached container for additional weights) but I use coins (nickels and pennies which respectively weigh 5g and 2.5g) instead of some sort of metal balls webwit use. I verified the weight using jeweler's scale sold on ebay for 5 bucks that measures in unit of 0.01g. The above measurements are off at most by 0.1 gram, but I think it is negligible inaccuracy for the purpose.

I don't tap my desk as ripster did because it would only lower the measurement (it would only decrease friction on the sliders by agitating. static friction vs dynamic friction). Instead I took the initiative of being thorough and measured all my keys for each time I change weights by 2.5g.

I lubricated the keys with best lubricant I could get that is safe for plastics and rubber -- Dupont Krytox. Before the lubrication some keys could support 70g without actuating.

[limmy]

I agree - friction has a totally different feel than spring/dome weight.  Low friction and no off axis binding are two of the things I love about Realforce.  Have you ever tried a PTFE "dry" lubricant?

I consder KG8 as dry PTFE lubricant as it is PTFE powder + volatile solvent. I put the disassembled the sliders in a plastic bag and poured some KG8 so that the powder is applied. As noted, the lubrication effect didn't last long although the performance was pretty impressive. Please note that the solvent used in KG8 melts ABS plastic which is used in housing of HHKB.

[Aranair]

I wish someone could tell me why the hell my new hhkb feels like a 5 year old hhkb D:
But man 70g without actuating will feel so stiff D:

Also I'm pretty positive that the 37 keys are harder to depress than the alt on the bottom left for some weird reason.

(I think i'll be most happy if all my keys felt like the left alt...)

[fossala]

Topres are trainer switches for BS, get you used the weightings gradualy.