thinking of taking the model F plunge: need advice

[wellington1869]

okay you model-F'ers.

So I'm thinking of picking up a model F on ebay. two quick questions:

--just how much lighter are F springs (say on the RipOmeter) than M springs?
--is there a feel difference between the XT and the later AT models?


My interest is mainly to use it as a donor board for the springs, I want to stick those springs into my m13.  Any foreseeable issues? (Other than being an incredibly laborious mod).  As I understand it the springs on the F are same diameter as M springs, except longer, but I figure a little snip-snip will take care of it.

[msiegel]

the AT is supposed to be smoother, the XT is supposed to be clackier.

there are subtle differences in the barrels and hammers; i'm not sure about the springs

[wellington1869]

the AT is supposed to be smoother, the XT is supposed to be clackier.

there are subtle differences in the barrels and hammers; i'm not sure about the springs

well my problem is, since i'm mainly interested in it as a donor board for the springs, the XT's are so much cheaper on ebay (I could pick one up for $30, as opposed to $100 for the AT F's), so it might be worth the risk to go with an XT on the assumption that the springs will be the same diameter...

[JBert]

You could try to transplant the springs - IIRC model F's were 55-65g (probably 60) whereas a model M is 60-70g (65 unicomp, 70 IBM).
I think Ripster tried a spring-swap in his Boscom (RIP), but the springs wouldn't buckle. This may have to do with the design change of switching to a membrane. Somehow, the model F patents don't really talk about down-force whereas the model M's even add some drawings.

I have another model F being shipped. If it turns out to be an AT (it appeared to have been mislisted), I'll try to compare the XT and AT in more detail. They are quite similar though, but YMMV due to the age of these keyboards.

[wellington1869]

I think Ripster tried a spring-swap in his Boscom (RIP), but the springs wouldn't buckle.

is that right? Damn I'll have to check that thread.  Did he snip the F springs to make them the same length as M springs? I'll have to read that boscom thread I guess. If he didnt snip the springs then I would think they'd be too long and therefore not buckle.

I have another model F being shipped. If it turns out to be an AT (it appeared to have been mislisted), I'll try to compare the XT and AT in more detail.

ya that would be great if you can review the differences if any. (ie, not only in key feel, but if you're up to it, to physically compare the actual springs by opening both boards up

[msiegel]

i wonder if there's enough detail in these disassemblies to be able to tell:

http://geekhack.org/showwiki.php?title=Island:5492
http://geekhack.org/showwiki.php?title=Island:6580

[wellington1869]

thanks for those links msiegel.

according to this pic (from the links above), XT-F spring on the left, is lighter and longer than IBM A01 spring. (Unfortuantely i dont know how the A01 spring compares with the M spring).



Ch123 here guesses that its the longer spring that results in the lighter spring resistance:
http://geekhack.org/showpost.php?p=98627&postcount=15

other comments from rippy's boscom thread:
ripster says: "IIRC I tested that earlier - stiffer spring makes it click earlier. I think it was as simple as cutting the spring then stretching to length" -- which is basically waht i intend to try out.
http://geekhack.org/showpost.php?p=113808&postcount=93
As rippy points out, its a laborious mod.

Here's the comparison between F and M springs:
http://geekhack.org/showpost.php?p=113651&postcount=79
F on top, M on bottom:

In the pic it doesnt look like they have similar diameter, but rippy says they do. The F has more coils too. F is just a hair shorter than M (no real difference in length?)


And then he tries swapping F springs into an M boscom: "let's see if this is the answer to the puzzle of making the Model M feel like a Model F. "


Rippy's results are pretty clear:

"STILL 70g. Bummer. The keys do feel significantly different. F19 [an F spring] is pretty mushy and you almost have to bottom out to get it to register. The other F keys have an exaggerated Ping sound. Not particularly pleasant.

... it's not a 10g difference by any means though.
So....back to Welly's question of how to lighten BS springs?..."

indeed.

[ironcoder]

Annealing them?

[ironcoder]

Smashing them with a big hammer and then dipping them in liquid nitrogen?

[wellington1869]

I think this problem basically is going to be unsolved until someone who works at a "spring factory" eventually shows up at geekhack. Then the problem will be solved overnight

[wellington1869]

sure but even if its a narrow range, it means the spring can be 'tuned' too, right? I mean if ibm has already produced these springs in a range from about 50g to 70g, then we know at least that range is possible.

[itlnstln]

50g?

I thought the F was around 60-65g.  The lighter ones webwit found in the terminal 'boards were of a different technology (Beam Spring).

[maxlugar]

sure but even if its a narrow range, it means the spring can be 'tuned' too, right? I mean if ibm has already produced these springs in a range from about 50g to 70g, then we know at least that range is possible.

Someone else should try and replicate the results of my key swapping experiment to see if what I observed was just an anomaly.

Swapping the one piece key caps on my 83-key Model F with two-piece key caps from a 1989 Model M seemed to lighten the key presses and slightly dampen the clicks.  

I am not positioned to perform ripometer type tests here in the office nor do I have an adequate supply of nickels and Tough As Nails glue on hand.

[wellington1869]

50g?

i read 55-60g on the F; the 50g i'm thinking of might be ibm displaywriter/terminal keyboards which I think have a different switch design so may not count. Still, 55-70 is a decent range, there's no inherent reason why it couldnt be tweaked 5g more downward, AFAI can see. Even if it is a narrow range.

[wellington1869]

Someone else should try and replicate the results of my key swapping experiment to see if what I observed was just an anomaly.

Swapping the one piece key caps on my 83-key Model F with two-piece key caps from a 1989 Model M seemed to lighten the key presses and slightly dampen the clicks.  

I am not positioned to perform ripometer type tests here in the office nor do I have an adequate supply of nickels and Tough As Nails glue on hand.

even if we can get F springs, we could acheive the remaining 5g by changing the sorrounding materials like this. Also dont forget you can put silicon dry spray on the plungers.

[ch_123]

edit: I'm pretty sure that Model F springs don't work in the Model M...

[wellington1869]

edit: I'm pretty sure that Model F springs don't work in the Model M...

i know, i'm trying to figure out why, or if a spring manufacturing company could make something that could work in an M

[wellington1869]

Huh?  They don't work well but they work.

rippy, whats the 'lightest' spring that you think could work in an M? 65g? (ie, is a unicomp as light as a buckling spring is going to go before it feels mushy?

[msiegel]

It's funny you keep repeating that urban legend which Ripster himself, where it came from, corrected multiple times.

that is funny... and a good example of how these things propagate.

btw, i've heard that Model F springs don't work in the Model M.
i kid.

[maxlugar]

I heard the NeXT keyboard is buckling springs, the Model M quality declined, the Model M has doubleshot keys, and Brown Cherries in the Cherry Corp boards are "special edition".  Google is always right.

Google and Wikipedia!

My wife goes ballistic when I quote a Wikipedia "fact"!

[maxlugar]

And I do believe in this nice Scandanavian guy.  He's either giving me a keyboard or a Nobel Prize for Christmas.

Show Image

Hey, that looks remarkably like an 83-key PC/XT Model F brick which is the equivalent of a lump of coal.

Good boys like me are wishing for a NIB 84-key PC AT keyboard...one of these years.

[maxlugar]

Cool, I'd be quoting suspect wikipedia facts while having sex.

Once again webwit, it was a joke that was obviously over your head because you failed to notice the word facts was in quotes.

I bet you really do imagine you are cool quoting suspect wikipedia facts... while having sex...with yourself.

[msiegel]

d'oh!!

[msiegel]

err...

oh baby!

XD

[maxlugar]

d'oh!!

Well, you know it's probably true.  I'll bet those Data Hands have many uses and are mighty handy for a troll.  

[maxlugar]

Either that, or my rater obvious play on your joke was over your head and you are making a complete fool of yourself..

Yeah, yeah, we have absolutely no doubt that you are a genuine intellectual with a validated IQ that is too high for the rest of us to even begin to comprehend.

More power to ya pal if you get your rocks off by finding "clever ways" to make fools of us.  That seems to be your ultimate response to anyone who ends up in a pissing match with you.

Jesus god, what was I thinking when I made a less than nice comment about poor kyamei  

[maxlugar]

Hmm I think I know what happened. I saw a cookie lying around, took it, and ate it, but now I realize it must have been maxlugar's. Don't cry now, here you can have one back. Have two.

Show Image

maxlugar: "webwit is sooo mean!!"

Hey, how did you manage to get a picture of my son?  What special powers do you possess?  I take it all back.  :)

[wellington1869]

I heard the NeXT keyboard is buckling springs, the Model M quality declined, the Model M has doubleshot keys, and Brown Cherries in the Cherry Corp boards are "special edition".  Google is always right.

Now to be fair to Welly even though the RipOmeter doesn't measure it I swear the Unicomp always feels a little lighter.  May be psychological.

It's funny you keep repeating that urban legend which Ripster himself, where it came from, corrected multiple times.

that is funny... and a good example of how these things propagate.

its funny, as recently as 2 days ago mseigel said to me, "it was the unicomps that were around 65 "


Ripster, I know you believe they are the same as the M's. But the sense of different resistance need not be 'psychological'. I dont believe the ripOmeter is necessarily capable, as a test, of catching the 'in use' downstroke force. Its a static test; typing force is kinetic.  IF the different plastics used since the 42H1292 (on which supposedly the unicomps (and my m13) are based, if the plastics used have a different friction coefficient, then the sense of lightness need not be "pscyhological" at all, and wouldnt be caught by the RipOmeter test either.

[wellington1869]

More power to ya pal if you get your rocks off by finding "clever ways" to make fools of us.  That seems to be your ultimate response to anyone who ends up in a pissing match with you.

qft

we have absolutely no doubt that you are a genuine intellectual with a validated IQ that is too high for the rest of us to even begin to comprehend.

I have doubts that thats what webwit is!

[wellington1869]

The question remains I guess, what is the minimum resistance springs under which the buckling springs are unusably mushy. If F springs (with their 55-60g range) are too much in a M.  And why would they be mushy in an M but not in an F? I know people have speculated about the rocker assembly making the difference (again showing just how much the sorrounding construction - and not the spring - contributes to overall feel), but it seems like all we've got on this question still is a lot of speculation.

[msiegel]

The question remains I guess, what is the minimum resistance springs under which the buckling springs are unusably mushy. If F springs (with their 55-60g range) are too much in a M.  And why would they be mushy in an M but not in an F? I know people have speculated about the rocker assembly making the difference (again showing just how much the sorrounding construction - and not the spring - contributes to overall feel), but it seems like all we've got on this question still is a lot of speculation.

i *can* tell you than when model f springs don't buckle, they feel mushy.

it's the combination of their natural lightness and a lack of buckling that made for mush in the test i tried.

[msiegel]

i was testing the at&t patented model m ping silencer mod using a model f switch module.

inserting the foam core (which might have been too tall) caused the buckling to fail, and the switch felt like mush

[wellington1869]

Oh - yeah, it doesn't take much to change the dynamics of a buckling spring.  Anyway Welly you seem skeptical of everybody else's tests.  Why not buy a Model F and try to mod it?  If you can beat the IBM engineers I'd be impressed.

well, first, I'm not sceptical, what i'm pointing out is that friction and other factors make a difference, and what all of your tests show is the spring cant account for everything.
second, as  far as the spring goes,  I'd like nothing more than to buy an F and get to work on hacking it into an M (which is why I posted this thread to begin with). This is a long term interest for me though since I dont have decent workspace (or time) to do it in the timeframe i'd like while i'm still in student housing  BUt yes, absolutely, this is something i'm going to do at some point cuz I'm very curious about the question of how to lighten a BS.

I doubt I can 'beat ibm engineers' (my dad was one, and believe me, there are some things, like engineering stuff, on which I'd rather not compete with him); however, ibm engineers in the '80s were, I think, shooting for a relatively heavy switch compared to the switches we use today, so it may be that they never seriously tried to 'lighten' it. (They would have had no need to do so). So its worth tweaking and experimenting with the basic BS design today to see if that can be done. I dont consider that competing with ibm engineers, I guess, since I dont believe they ever were shooting for that same goal. Since unicomp wont step up to the plate with keeping up with the times, I guess its up to hackers to see whats possible.

[wellington1869]

I'm skeptical that you're not skeptical.  LOL - being skeptical is OK since I'm the biggest skeptic around here - it's just some of these things you'll probably have to try for yourself.  Anyway, the Model F ATs don't seem that expensive and would be easy to sell.

lol, really i'm not sceptical, i'm just trying to get as much info as possible before I take the hacking plunge here, so I dont repeat your efforts.  Its in that sense that i'm not sceptical. I'm not trying to "re-test" stuff you guys have already done.

If my goal is 'how to lighten the BS' then I need to know what part the spring plays and what part other factors play. Thats what i'm trying to figure out.

I do think your interpretations of your own results will maybe be different from mine. For instance, I do think you put too much emphasis on the spring itself.  One of the revelations I've had recently (after I got my TVS) is just how much everything-other-than-spring matters in overall feel/sound.

But that doesnt mean i'm going to redo your test. I take you at your word that the F spring is mushy in an M.  I just want to know why - and at what resistance it wouldnt be mushy.  Cuz getting an F spring to work in an M would go a long way towards getting a lighter BS, which is ultimately the goal here.

I believe you when you say it doesnt work. My question is: What would it take to get it to work?

Thats not being sceptical; thats taking you at your word and then asking the next question.

If i was simply sceptical believe me I have no qualms about saying so. For instance, I'm sceptical that webwit is "a genuine intellectual with a validated IQ that is too high for the rest of us to even begin to comprehend." There, i've said it. See? I have no qualms about saying so when i'm sceptical about something. xD

[ch_123]

the Model M quality declined.

So a Unicomp is of the same quality as a 1391401?

[wellington1869]

So a Unicomp is of the same quality as a 1391401?

this is one of those debates that I think is more in the eye of the beholder (despite ripster's certainty on it . Personally I think it 'declined' tho not so much that i'd avoid unicomps or anything (my daily driver is a unicomp).

[wellington1869]

See...  Skeptical again.

nah, just sayin we have a different 'interpretation' xD

[msiegel]

we have a different 'interpretation' xD

heyyy, what is that supposed to mean?