jgeorgeLOL what the everloving heck is this, it's the weirdest thing I've ever seen an IBM logo on. Spoilers: it's legit. I have it, it works, it does precisely what it says on the tin, and I love it.
Takeapart thread coming soon.
IBM Audible Random Timer lolwtfbbq thread, part 1 of however many.
This is a legitimate IBM product. Not for sale, clearly an internal - probably Field Engineering - tool. Here's some exterior views of the device. It's a plastic hobbyist project box about 45x70x110mm in size. It was originally plain white plastic, but this one has been brush painted gold. My best guess is Testors gold model paint that used to come in the tiny glass bottles with impossible to remove caps (https://www.dickblick.com/items/testors-enamel-paint-gold-14-oz-bottle).
Tell #1 it's an actual IBM product: Anyone knowing IBM machines from this era (spoiler: mid 80s) knows that stamped serial number tag. Proper tag, proper stamping, proper typeface on the stamp.
IBM Audible Random Timer lolwtfbbq thread, part 2 of however many.
Boring pics of the inside of the plastic case, showing whoever painted it did actually (good for them) test the paint to make sure it wouldn't damage the plastic.
Juicier pics coming, I promise. :)
IBM Audible Random Timer lolwtfbbq thread, part 3 of however many.
GUTS
PCB front and back. Tell #2 it's IBM made are the yellow bodge wires on the back of the PCB to add a couple of diodes in (IBM has a specific bodge wire color scheme, yellow means it's a rework done in manufacturing). Tell #3 are the rectangular tantalum capacitors with IBM part numbers on them.
5615372 and I are old, old friends. That's a standard issue double capacitor part that's been in use since Ramses II was Pharaoh of Egypt. They'd be standard parts for an IBM PCB design.
IBM Audible Random Timer lolwtfbbq thread, part 4 of however many.
How it works - overview.
The slide switch selects off, low volume, or high volume. "Low" volume is quite low. In a quiet room it's noticeable but not jarring. High volume is quite a bit louder, bordering (to me) on annoyingly loud in a quiet room but probably acceptable with any other ambient noise.
When powered on, it beeps 4-5 times to let you know it's working. Then, randomly, it will beep again 4-5 times and stop. You can stop the beeping by pressing the slide switch (it's also a pushbutton). If you leave it alone it'll beep those same 4-5 times then stop.
So far, with only a few (6-8) datapoints, the "random interval" has been between 7 minutes and just over an hour.
Beeper is a little piezoelectric buzzer stickytaped to the inside of the front panel.
Also, IBM-made tell #4 is the front panel. It's the same thickness and material as IBM machine nameplates and is the same screen printed technique.
IBM Audible Random Timer lolwtfbbq thread, part 5 of however many.
How it works - detail.
The right side of the PCB are two 4060 binary counter ships and a 74HC688 8 bit comparator. This is I think where the "randomness" is generated - assuming the two counters run at different rates, the comparator would trigger the alarm when the two counters hit the same number at the same time.
IBM Audible Random Timer lolwtfbbq thread, part 6 of however many.
R1 nearest the buzzer wires is a volume adjustment. I just realized that, and realized it's all the way down. Turned up this thing is SCREAMINGLY LOUD EVEN ON LOW.
R11 is a pitch adjustment for the alarm, and it's EXTREMELY touchy. Only about 5% of the range of the pot runs through audible frequencies, even just putting a finger on the pot will change the pitch significantly.
No idea what the 12-position rotary switch at SW2 does yet. Changing it doesn't seem to make any obvious difference so far. My theory is that switches between various resistor combinations to affect the time scale of the "random" events, but it's hard telling with so few data points so far.
IBM Audible Random Timer lolwtfbbq thread, part 7 of however many.
Non-Audible Random Musings
All the chip dates are 48th week of 1983, meaning this device was probably manufactured early 1984 or so.
There's a belt clip, so this thing is designed to be worn or at least carried on your person. The best any of us have been able to come up with it that this was a way to keep someone from hyperfocusing on something - you get lost in your work, and are off in the weeds looking at something but maybe not realizing how much time has elapsed while you've been hyperfocusing, and this thing might have been around to remind you of the passage of time without being annoying.
IBM Audible Random Timer lolwtfbbq thread, part 8 of apparently 8.
Thats all for now until I can put a meter or logic probe on this thing to figure out more in detail how the counters operate or what that giant rotary switch does.
I'll answer any questions with my best guesses, or show more detail of the PCB for anyone that wants it.
I'm here all week. Tip your server. Try the Ceasar Salad!
IBM Audible Random Timer lolwtfbbq thread, part 9 of what is now 9.
I’ve left this thing on for almost 2 days solid. I haven’t been timing the “random” beeping, just kind of getting a feel for its scale. It will beep around 10 times in what I’d consider an “8 hour shift”.
Several people said it might have been used for task sampling, log what you are doing when it beeped, but those same people also said they’d be completely irritated if that was actually the use case, so I don’t think that’s it.
One of the most plausible explanations I’ve seen so far is that it might have been used as a “random sampler” for a manufacturing line - when it beeps, you pull the next thing off the line and send it off for validation testing. Given its wide range of randomness (minutes to hour+) and that the timing averages out to about 10-12 times a shift, that seems like it’d be a possible use.
Scope and logic analyzer (and sound samples) coming.
Marsh Ray
geekpower
awight@jgeorge thanks for the dissection notes, and for the new life goal to build a board stamped „A.R.T.“
xxx@jgeorge no sound / videos to listen to?
leyrer
tecleandor@jgeorge I was reading your toots and I was thinking... "I don't know what did they use this for, but I think I could build one for me for EXACTLY THE SAME REASON you said here."
Eli the Bearded@jgeorge
Evoking "Harrison Bergeron" by Vonnegut to me.
Evoking "Harrison Bergeron" by Vonnegut to me.
Pyxaron@jgeorge My guess for what it might be for is for QA purposes, where sampling needs to be done at a random interval to avoid predictability of testing.
@Pyxaron This is the most plausible explanation so far. Unlike other comments about task sampling, someone suggested that if you were working a manufacturing line or something, when this thing beeped you took the next machine off the line for validation testing. It would be reasonably random and (spoiler alert) based on its timing samples so far it would result in about 10 samples over an 8 hour shift.
O! ver 
@jgeorge Kudos for the explanation of what this timer is used for. There should have been a setting that makes the timer never signal an end, like a Schrodinger timer.
My name is Gordo@jgeorge That's such an elegantly stupid simple mode of operation. I love it.
Phil M0OFX@jgeorge Very interesting thing - and unusual. This is the first time I've heard of a "mod wire colour scheme" though. If yellow is a factory mod, what other colours did IBM use, and for what?
@philpem The ones I remember (I’m sure there were more) are:
Yellow: Factory EC (Engineering change). Applied to a board to indicate that the board was upgraded to a new EC level at the factory.
Black: Factory Configuration. Applied to a board for necessary modifications to support a specific combination of features. These were config specific, so you knew what wires to remove if you changed the machine configuration that required it).
Blue: Field EC. Engineering Change applied to a board in the field by an FE.
Pink: Pluggable jumper wires used in the field for temporary modifications or diagnostics. “Temporary” was a highly variable time span.
If you opened a failing machine and saw yellow and black wires, you didn’t really worry at all. Blue wires, you raised an eyebrow because you might be tracing out some other FE’s work. Pink wires everywhere? Call the spouse and tell ‘em you’re gonna be late.
@jgeorge There metallic parts to drive the screws, I wish I could find some of them nowadays.
@olireiv Those are called “heat set inserts” and are pretty popular for 3D printing You use a small soldering iron to heat them up enough to melt the plastic, stick ‘em in a hole, and let the plastic cool around them.
Stewart Russell@jgeorge you were probably supposed to note down what you were doing it went off, preferably involving therbligs
my log would be a series of "screaming because this thing beeped at me" comments
Mekki@jgeorge following because I work at IBM and we have an internal IBM history community that loves unpacking where these weird objects came from. Please share more soon!
Benjohn@benjohn @mekki Your guess is as good as mine or anyone else's. Random sampling is a good idea. My thinking would be more of a "hyper focus interruptor" to keep from going off into the weeds on a problem, but then again I have a problem with hyperfocus so that would be MY use case for something like this.
@mekki ...you do know that the majority of my hobbies involve restoring and documenting old IBM midrange systems, right?
OH HELLO NEW BEST FRIEND
@jgeorge <Waves hello> You can reach me via email if you wish too: mekkiATibm.com
@gordoooo_z @mekki It’s hard not to be an IBM nerd if you work there. There’s SO MUCH history, and IBM has made so many many different things over its lifetime, there’s always going to be something that you can take a historical interest in. And honestly, IBM has a fair bit of secrecy to it that it has a bit of cachet as like AT&T or the Bell companies did back before the breakup, not even intentional secrecy but they do so much in so many places that there’s always something to discover.
@jgeorge @mekki Oh for sure. 100% on board. I would love to just take a walk through that building (well, buildings), let alone get to do so regularly. I guess I just wonder how many of those IBMers were already IBM nerds, vs. how many became IBM nerds once they were hired and were suddenly surrounded by all of it.
I have to assume many of them. I'm sure AT&T has its share of Ma Bell geeks as well.
By the way... y'all hiring? :p
@gordoooo_z @jgeorge hiring is a complicated question with broad swaths of layoffs across the tech sector. IBM is always "hiring" but whether or not it's worth your time to throw your hat in the ring is an entirely different question. Find me on LinkedIn if you want to discuss. I'm deliberately easily discoverable.
Jazz Rabbit 🐇🏳️⚧️@jgeorge i have seen a claim that these were used in speed shooting competitions; the semi random timer was the starter alarm, so that the competitors could be gauged from an unpredictable start time.
Plausible(?) but unverified.
@madrabbit I saw that on reddit. Given the time-scale of this thing's randomness (10 min to 60+ min so far) it seems like it would be a completely unfeasible random start timer for any kind of competition.
Chris@jgeorge @madrabbit maybe the Barkley Marathons but that's pretty niche!
Maier Amsden@jgeorge @madrabbit At the same time, 7 minutes to an hour is a pretty broad time span for preventing excess tunnel vision. 7 minutes seems way too short. An hour seems a bit long. And why not have a standard timer set to the perfect interval for what you typically do? I hate annoying beeps, so I'm calling this a torture device.
Stu Duerson, machina exo deusI came on to say the same thing, however in Single Action Shooting, once the timer goes off, it also senses the times between shots, for a performance indicator. I've got one floating around here, somewhere...not IBM.
This one may be to signify time to take measurements, in a study that requires a random time selection.
Irenes (many)@jgeorge audible random timer :D fair enough lol
Rufus J. Cooter@jgeorge Instant follow, if only because I suddenly have an urgent need to get updates on this as soon as they're available
Tormod HalvorsenI’m sure that the printed manual for this thing will have several numbered pages with the only text being:
THIS PAGE INTENTIONALLY LEFT BLANK
🙂
Simon Frankau@jgeorge So cool!
https://www.reddit.com/r/whatisthisthing/comments/1e960ti/ibm_audible_random_timer_vintage_electronic/ has some good ideas.
Have you tried running it? The length of the random intervals might give you an idea of its intended use.
I love the way they put some perfboard into the PCB design and used it with the bodge wires! Maybe it's a prototype, if they were so confident they'd need to bodge it? :)
Anyway, such a neat find, and some very cool archaeology by you.
Dan Neuman 🇨🇦@jgeorge The first retry with random wait protocol
Trammell Hudson@jgeorge that's definitely weirder than the spray can of IBM oil that was included with my 129 card punch.
@jgeorge although I think the calculating cheese slicer is the weirdest official product from IBM.
arclight
Badscrew@th I saw a restoration of one on YT, and I agree its up there in its weirdness. But I feel the ART still wins out because although both products tell you what they DO ("I calculate cheese slices" "I randomly make noise") at least you can tell WHY the Calculating Cheese Slicer does what it does. Nobody so far seems to have any idea why the ART exists or what problem it was made to solve. :)
Its less annoying with the pitch of the beep potted down like half an octave though. :)
SuperIlu
Lysander il breve@jgeorge
Omg this is beautiful!
Omg this is beautiful!
Fiona Ebner@jgeorge Irregularly Beeping Machines
JamesB@jgeorge Didn't Bob Wildlar build something similar years before?
Richard W. Woodley ELBOWS UP 🇨🇦🌹🚴♂️📷 🗺️Well for those times the recipe says "cook for a random amount of time", I guess ?