IBM 3278 review (beam springs)

[Chyros]

A lot of people have been waiting long for this one, but here it finally is: a beam spring review! Hope you enjoy it, this was quite a keyboard .

[derzemel]

Man, that clicker sounds like a jackhammer.

Awesome review

[pansku]

Great review like always

At first I thought that it can't get much louder than the F, but that sound sure is something beyond belief.
Is the speed limitation there even with the xwhatsit contoller and having solenoid turned off?

Too bad I missed the opportunity, in August a 5251 terminal with keyboard went for 150€ IIRC in Finland, so I could've even picked it up and wouldn't have had to pay for shipping

edit: Or was it 150? The page now says the starting bid was 98€ and there is no bidding history, a lot less than ebay prices that's for sure and it was even with a finnish layout

[Chyros]

Man, that clicker sounds like a jackhammer.

Awesome review

Thanks! This one took a while, but I'm very happy with how it turned out ^^ .

In a way, a jackhammer is actually EXACTLY what the clicker is xD .

Great review like always

At first I thought that it can't get much louder than the F, but that sound sure is something beyond belief.
Is the speed limitation there even with the xwhatsit contoller and having solenoid turned off?

Too bad I missed the opportunity, in August a 5251 terminal with keyboard went for 150€ IIRC in Finland, so I could've even picked it up and wouldn't have had to pay for shipping

edit: Or was it 150? The page now says the starting bid was 98€ and there is no bidding history, a lot less than ebay prices that's for sure and it was even with a finnish layout

Thanks . I have no idea how to turn the clicker off, Andy forgot. I tried all Fn combos but they don't do it. I'd be kinda surprised if it had something to do with the clicker though.

Btw "clicker" and "limitation" are two more words it doesn't do. They come out as "cliker" and "limtation" xD .

[alh84001]

Just don't tell me you get this for 50p . Off to watch the video.

Edit:
Wow, this is one of the nicest keyboards I have ever seen. It's basically like new. One thing I always wondered is whether beam spring switches needed to be this tall, or if this was just IBMs intention, and they could actually have been made shorter.

I also can't discern it in the video, but is there a two part to the click (without solenoid), one when key is pressed and the spring collapses, and then another one when the leaf hits the PCB once key is released.

[zslane]

With the clicker turned on it sounds like a teletype machine. Absolutely amazing!

[Chyros]

I also can't discern it in the video, but is there a two part to the click (without solenoid), one when key is pressed and the spring collapses, and then another one when the leaf hits the PCB once key is released.

Yes, it's two-way clicky. This is unavoidable for inversion-type springs like this (Alps' plate spring switches are another example of this). It's hard to say but I don't think the fly plate hitting the PCB actually makes all that much noise.

[supamesican]

Its beautiful, maybe the only one better is the selectric turned keyboards from what i understand. so jelly

[pr0ximity]

Great review as usual Chyros! Very jealous, the 3278 is second only to the 3279 for me I think. Though the 3101 is quite nice as well

Your mention of the solenoid delay is the reason I usually type on mine without the solenoid driver active. It's a lovely sounds but the charge-up time can cause some silly typing mistakes.

[y11971alex]

Are we sure that the 3278's caps are ABS? They never seem to discolour regardless how long they're exposed to sunlight.

[Chyros]

Are we sure that the 3278's caps are ABS? They never seem to discolour regardless how long they're exposed to sunlight.

I'm not going to acetone Andrew's caps to find out :p .

[chzel]

The spacebar of my 3101 has a PEI mark on the underside,

so it's possible they are PEI
I haven't been yet brave enough to acetone test my theory..

[Scarpia]

Wonderful review Chyros -- thanks! And now I'm even more excited for my beamspring keychain to arrive!

[Chyros]

The spacebar of my 3101 has a PEI mark on the underside,

so it's possible they are PEI
I haven't been yet brave enough to acetone test my theory..

PEI is apparently water-soluble :p . Don't think they're that :p .

[chzel]

Apparently it has way better properties than what the wikipedia page would have you believe...
http://www.prlresins.com/polymer-resour ... /pei-resin
http://www.hycompinc.com/PDFs/ULTEMProductBrochure.pdf
https://www.sabic-ip.com/gepapp/Plastic ... rchresults

[y11971alex]

PEI also stands for Prince Edward Island, a province in Canada.

[Parak]

The issues with missing characters while typing are likely the fault of the replacement controller rather than something inherent to beam springs. Given that a lot of these keyboards were used for data entry by ex-selectric users, any fault along these lines would have been completely unacceptable. I've noticed similar issues with the xwhatsit controller regardless of threshold settings when I used it for a friend's keyboard, and that was an F - until I turned off the beeper/solenoid header. After that, no more missed characters. No idea if that's the controller being overloaded or something, or a bug of some other sort.

[y11971alex]

The issues with missing characters while typing are likely the fault of the replacement controller rather than something inherent to beam springs. Given that a lot of these keyboards were used for data entry by ex-selectric users, any fault along these lines would have been completely unacceptable. I've noticed similar issues with the xwhatsit controller regardless of threshold settings when I used it for a friend's keyboard, and that was an F - until I turned off the beeper/solenoid header. After that, no more missed characters. No idea if that's the controller being overloaded or something, or a bug of some other sort.

My experience is that my 3278 suffered from it. My 3101 didn't, or at least not to a noticeable degree.

[Ratfink]

The spacebar of my 3101 has a PEI mark on the underside,

so it's possible they are PEI
I haven't been yet brave enough to acetone test my theory..

Apparently it has way better properties than what the wikipedia page would have you believe...
http://www.prlresins.com/polymer-resour ... /pei-resin
http://www.hycompinc.com/PDFs/ULTEMProductBrochure.pdf
https://www.sabic-ip.com/gepapp/Plastic ... rchresults

You two are talking about different PEIs! But I'm pretty sure it's not either. According to UL, polyetherimide (the PEI that chzel was referring to) was introduced in 1982, so it's definitely not made of that PEI, which would be an awfully expensive plastic to make keycaps out of anyway. The other one, being soluble in hot water, would be an awful material for keycaps.

Most likely, they're ABS. It's flame retardants in ABS that cause it to yellow, not the plastic itself.

[Ratfink]

I also can't discern it in the video, but is there a two part to the click (without solenoid), one when key is pressed and the spring collapses, and then another one when the leaf hits the PCB once key is released.

Yes, it's two-way clicky. This is unavoidable for inversion-type springs like this (Alps' plate spring switches are another example of this). It's hard to say but I don't think the fly plate hitting the PCB actually makes all that much noise.

It's quite possible to say, actually! Remove a switch from the keyboard, press it while against something, and keep it in the pressed position. Release it against something soft and see if it clicks. It's been a while since I've had my Displaywriter apart, but my recollection is that they don't click in that test.

[Chyros]

I also can't discern it in the video, but is there a two part to the click (without solenoid), one when key is pressed and the spring collapses, and then another one when the leaf hits the PCB once key is released.

Yes, it's two-way clicky. This is unavoidable for inversion-type springs like this (Alps' plate spring switches are another example of this). It's hard to say but I don't think the fly plate hitting the PCB actually makes all that much noise.

It's quite possible to say, actually! Remove a switch from the keyboard, press it while against something, and keep it in the pressed position. Release it against something soft and see if it clicks. It's been a while since I've had my Displaywriter apart, but my recollection is that they don't click in that test.

Yeah, or maybe, you know, not start tearing someone else's keyboards apart .

It also doesn't account for acoustic dampening anyway.

[hypkx]

Nice review, I am curious how is the mx stem from the keycap modified to fit on a beamspring stem?

[jacobolus]

A slight click on the upstroke is a positive feature. Compared to the solenoid clicker sound, you’ll never even hear it though.

[Scottex]

Nice review, I am curious how is the mx stem from the keycap modified to fit on a beamspring stem?

wondering this as well

[andrewjoy]

The caps are ABS , but don't forget that all ABS is not made equal.

Just beacuse you have the 3 chemicals and the fire retardants does not mean that you will make the same quality plastics with them.

Its like saying that you have a plate of ingredients and a untrained average joe would use them to make a cake that is as good as a top chef.

Nice review, I am curious how is the mx stem from the keycap modified to fit on a beamspring stem?

wondering this as well

Its modified by the jam it on method , some MX caps fit , some don't. I think beacuse the tolerances are so wide for MX caps. I would not replace a whole board with them. But for that one odd cap its fine ( like my FN key) , the default beamsring cap is just too low , no way to use it effectively.

[Hypersphere]

Spectacular! With the clicker on, it sounds like a rapid telegraph key to me.

You've given me additional incentive to pull my beam spring out of the box and set to work on refurbishing and converting it to USB. I have an older-generation Xwhatsit converter that requires soldering all the connections.

[Elrick]

Suspect this keyboard can only handle 60wpm?

The Mrs couldn't use it at all but I love the sound of it, so "Mechanical" like some factory producing parts. The noise reminds me of a few places I worked at when I was young and I actually missed it because it lulls you into sleep as the sound is never ending or stops.

Thanks for finally doing the Beamspring, very unique piece of Keyboard history here and I wonder if anyone will produce a Keyboard using these type of switches, in a future keyboard.

Have to wait and see.

[lot_lizard]

You make such wonderful videos. Couldn't be said enough... I love the comparison to the F, and couldn't agree more... this is such a different animal (in an EXPONENTIALLY better way). And @andrewjoy, I haven't come across a DisplayWrite beamspring version. This is most definitely for that application (and the late timing of the dates you found would certainly confirm). All of the green front pad printing would be for that exact application. This might actually be more rare than the APL's we drool over (even though the APL's are actual triple shots).

Suspect this keyboard can only handle 60wpm?

The Mrs couldn't use it at all but I love the sound of it, so "Mechanical" like some factory producing parts. The noise reminds me of a few places I worked at when I was young and I actually missed it because it lulls you into sleep as the sound is never ending or stops.

I really like the novelty of the solenoid, but can never have it on. In almost all my beamspring setups, the accuracy with and without that solenoid varies greatly. Something about that little devil makes the accuracy hiccup. I have never looked at the code backing all of this, but the solenoid vibrations (or need to charge the driver) ALWAYS influence the xWhatsit.


With the controller HIGHLY tuned (think upwards of 15-30 minutes per board), I can get 90-100 wpm on these boards without any hiccups in missed keys (all my beamsprings). As soon as I turn on the solenoid, it falls to maybe 70?!? wpm before issues occur. That coupled with the solenoid driver needing to recharge at a slow rate (perceived missed keys on the audible, even if the keystroke shows up), it slows me down to the point that I just can't operate. If you are hearing clacks for most keys, but not some... it REALLY messes with you.

It took a lot of time, but I use my small collection of beamsprings as my daily drivers 30-50% of the time. I LOVE them. but... I can't ever use with the solenoid on . Just frustrates too much. We type so much faster than an ancestors (and I am by no stretch BLAZING compared to others). I can consistently hit 95 wpm as an actual average, and 120 in short bursts for common collections of words. With all of the beamsprings I have encountered though, they can keep up using the xWhatsit (lots of tuning though).

Again... @Chyros. LOVED the video, and so happy to see you do this one. I've been waiting for it.

[Chyros]

Cool glad you liked the video .

Tbh I'm really stoked to get a beam spring of my own now. I don't mind one of the really really big ones; big is good, after all. But a working one, especially with the solenoid, that'd be amazing .

[micrex22]

Thanks for the review. I almost bought a 3270 beamspring board, the one with the colours (but really wished the seller kept the whole terminal then it'd be worth it). The WPM cap is a huge limitation... of course if one had the time the controller board itself could be replaced but that would take a lot of time and work.

There are two things that utterly distraught me about beamspring keyboards if they were used as a main with everything else aside:
#1 The sound. It doesn't sing because there is no buckling happening. Sure you can turn on the solenoid to compensate, but then... you can do the same for a Model F and have an even louder 'orchestra'.
#2 No PBT keycaps.

What I find curious is that there isn't really a clear path to cost cutting in certain situations. For instance, buckling spring keyboards upgraded the keycap plastic to PBT (although the fact PBT wasn't used could have been about something else and not cost). Model M 'innovated' by introducing a more expensive detachable cable rather than a fixed cable on most of the beamspring and Model F keyboards. And of course many Model Fs use a weird brittle plastic instead of PVC.