[Devon Croy] belongs to a hackerspace that works hard to keep hardware from going to the landfill. He found they were in possession of over a hundred Erasable Programmable Read Only Memory chips (EPROM). Not to be confused with EEPROM, which are electronically erasable, these EPROM chips require a strong source of UV light to blank the old data before they can be written again.
Instead of buying a tool to erase two or three chips at a time he built his own bulk EPROM eraser from an old metal toolbox. He used parts from a fluorescent black light and acquired a new bulb that generates light in the UVC spectrum, the band which works as an eraser for the chips. After bolting the parts into the case he added a spring-loaded timer knob and a safety switch that kills the power when the case is opened, similar to the UV exposure box we looked at yesterday.
Of course, if you don’t need a bulk eraser you could shop some garage sales for a UV pacifier cleaner which can also erase EPROM chips.
I applaud his desire to keep e-waste from going to the landfill, but whatever energy savings that results in ( by not having to buy a new equivalent ROM device ) is somewhat countered by the energy it takes to shine the UV light on those things to erase them.
Plus, what’s your time worth ? Most people who know how to work on this stuff probably earn more per hour than to waste their time UV-ing / inserting / removing. But of course it’s his choice what he does with his time.
Excellent, and I applaud resisting the urge to throw away all those windowed parts, as they have plenty of useful life in them.
When I’m testing my hand written machine language, I find I often need to change all 256 bytes of available data in my solid-ceramic case 1702 EPROMS. This looks like just the thing!
I cannot tell you how hard it is to truck my little mistakes down to the dentist to clear them using a goodly dose of X-rays, and then have to wait for hours while they bake in the oven to anneal the damage caused by the X-rays before I can put them in the programmer and have another go.
Am I the only one that misses having to provide -12vdc or +25vdc to a part in order to program it? Why on earth would you want to throw away perfectly good EPROMS when the only alternatives are nearly free eeproms that are drop in replacements? :)
I know the feeling, but it might be time to move on. The days of blow and go are past us – and I don’t miss them very much. However, kudos to the guys out there making it fly. More power to you.
I hope that someone will post an article showing us how to wind core memory and access it with an arduino! 8 x 8 should be easy enough – you could store simple passwords in a way that would resist gamma rays.
And if we could obtain enough mercury, we could build a mercury delay line memory device. And vacuum tubes! So much technology slated for the landfill waits for resurrection. Please, show us how to dig those old DTL and RTL logic cards out of landfills and connect them up.
Nostalgia – it ain’t what it used to be!
@Timmah,
try and find a new eeprom to fit in a Commodore PET or a Rockwell AIM65.
I am off to make one of these erasers so I can use my stock of TMS2532s.
@Timmah,
there are alot of things like arcade and pinball machines that use eproms, even industrial equipment that has been running non stop for the last 30 years… sometimes they go bad but usually they just need to be erased and re written to and will last another 30 years no problem
there really is no need to replace them simply to save a bit of time in programming each chip when there are perfectly good ones already present, its rather wasteful for a one time convenience
@bilbao bob,
ive been fighting off a very strong urge to make core memory for some time now, i need to finish a dual 8 telecine project that has taken two years to finally near completion… i have a ton of old processors (8088, 6502, z80, m6800, m68000…) that i want to play with someday when i have the free time not tied up with a dozen or so projects in juggling around right now
I’d rather see a UV-laser that can erase single bits in a running system. But I guess the refraction of the glass could be an issue.
@bilbao – does your HMO cover the EPROM’s trips to the dentist ?
@other posters: good point, I didn’t think about legacy hw. makes sense, although I’m surprised they don’t have pin-compatible eeprom replacements.
its hard to find parallel eeproms pin compatible for every eprom, the early ones especially
its also hard to find parallel eeproms period, and getting parallel flash chips that are ttl compatible is a serious challenge, most 5v versions only come in tssop packaging requiring even more work
all in all im very happy with how it turned out, a 30 dollar investment that ended in something that would have cost several hundred for a unit with comparable capacity
I am using 27SF512 electronically erasable PROMs in my current project. So much easier!
@everyone
“try and find a new eeprom to fit in a Commodore PET or a Rockwell AIM65.”
2716 (2k x8) and newer all conformed (mostly) to jedec standard. You can generally substitute any JEDEC eeprom for any old school PROM/EPROM. You might have to bend/cut a pin or use a resistor to bring up a select line, but they’re drop-in replacements all the way back.
By the time you’re back to 1702 era, you can burn a pic or avr or other micro and just emulate the damn thing in software. This is a little known way to add features to 70s vintage gear.
Let’s say you have a ranging head off an old IR tracking… uh, lawn dart… and you find that crap 1960’s tech just isn’t cutting the mustard and you want to make it able to locate the smiling face of hello kitty, assuming the face of hello kitty is created using a modulated laser pointer. Well, you might need someway to replace the chopper mechanism with something a little less archaic. Electronics has come a long way since 1965, so …maybe – I’m sorry, I seem to have drifted back off topic again.
You see, in those days, we hung an onion from our belts because that was the style of the times, and it had to be a white onion. Red onions were for commies.
@bilbao bob
Why are you so adamant about getting new parts? He found a bunch of old EPROM chips, and is actually making them work. There’s no sense in purchasing new chips when he already has a perfectly good solution. While he probably spent way more making the eraser, it’s a good tribute to the older technology, and a good thing to use to teach just how our technology has grown.
Hey, Abbott… :)
It’s all good. It’s just that when you add up the cost of waiting for them to burn clean… I spent my time smelling ozone (it should be obvious) while chips baked, and I applaud anything someone chooses to build.
But it’s like putting a hand crank starter on your car – one day you decide that hand cranking died out for a good reason. I was the same way about toggle switches on computers, but one grows older and starts to think about time.
My current project is a 1937 something or other, so I’m far from being able to criticize him. I’m just saying the burn/test/edit cycle is old school but ineffective.
I think you miss my point – we can make it better/faster/bigger/stronger and have fun doing it. It’s fun to go one step past reuse and aim for recreation. Seriously, there’s a lot of room for people who build functional replacements for old electronic parts like EPROMs and NLA integrated circuits.
I have boxes of formerly expensive parts. I will never burn another eprom or design a noiw defunct intel supercomputer into another system. I hate to admit it – given the choice between an arduino and a Rockwell AIM65, I’d now use the arduino.
I have to agree with taking the easy route for some projects, i went with a springwound timer simply because i didn’t feel like messing around with a micro controller and relay to turn the bulb on in timed intervals
we have a dirt packer that weights at least 4 tons beside the hackerspace that is missing a controller and i can guarantee i won’t be using a single eprom to get it running
but building the tool so that i can use them at some point in time did seem like a good idea, maybe for some horrible project ill implement an 8 bit cpu in eproms or something challenging along those lines to keep my more dangerous ideas on the back burner (autonomous dirt packer anyone?)
Is the box grounded? It looks like the ground pin of the power cord socket is simply capped off with shrinktube, which seems silly given there are several perfectly good grounding points right next to it.
Also, where did you get the UVC tube? I’ve only seen them selling for ridiculous prices.
no i didn’t ground the box, not yet anyways… i have a few minor adjustments to do with it still
i got the bulb from a friend who ordered some off ebay, a g15t8 bulb will run below 15 dollars plus shipping usually
@dcroy
“So easy an arduino could do it”.
Hang an arduino off that dirt packer using a couple of relays and have at it. There is nothing on any packaging equipment old enough to lack a controller that won’t be easy to get running.
People make decent money retrofitting old controls, and it’s damnably easy if you watch out for safety issues (mechanical interference, operator safety, burning the place down) while doing the retrofit.
Publish it as an article on HAD with VIDEO of the packer operating, and I’ll send you some lovely french pate that goes well with cheese and crackers… if you get it running by 1 JAN 2011.
We should add a section called “So easy an arduino could do it” and fill that baby up.
unfortunately we are too busy getting our building in shape for the winter to get to the packer for awhile, and for reliability none of the electronics can be on the packer itself, originally all of it was in the controller, which we don’t even have
the vibrations from the engine just running are enough to not even think of having any critical electronics on board, we might end up getting a broken controller and going from there
right now our concerns are with our new building, a 14 by 70 foot trailer that currently needs a ton of work (electrical, plumbing, insulation, windows) it would be nice to work on the packer but nicer not for freeze our asses off working on it (already started snowing here)
@Rachel
Grounding isn’t needed here, as the UV tube and timer mechanism are completely isolated from the case. It just happens to be a metal box.
@dcroy
Good luck beating the weather.
Grounding is definitely needed, for the same reason metal electrical boxes must be grounded: if a wire happens to come loose, it could touch the case and electrify it. It would need to be double insulated to be up to code if ungrounded. It only takes five minutes and a scrap of wire, so the only excuse is laziness (which is perfectly acceptable if you’re the one taking the risk).
im going to use the transformer casing as a ground point, ill scrape some paint off so it will ground the box at the same time… and add a fuse to it when i find a spare fuse holder in my pile of stuff
on another note i used some really flexible wire rated at 600v 110c from an ancient crt to prevent fraying at the hinge, im not entirely sure how all the plastic coatings will hold up to that much uv exposure… time will tell i guess
just found a few 27c801 eproms in a piece of telecom equipment, i think ill build a snes devcart next
Well, i really lol’:d at Hank’s coment on that one=) BUT, just use a filesystem that doesnt need defragging hehe