A look inside the chips that powered the landmark Polaroid SX-70 instant camera | Hacker News Reader

A look inside the chips that powered the landmark Polaroid SX-70 instant camera

righto.com

136 pointspicture46 comments

The film packs included the (flat) battery, and the (non-rechargeable) battery always had plenty of charge left after the film pack ran out. I salvaged those batteries and used them in many electronic projects.

Polaroid even capitalized on this a bit – they produced a radio that ran on the batteries in spent film packs in the 1980s.

The Polaroid 600 Plus Radio!

https://www.flickr.com/photos/52013582@N07/34336986451

I'm impressed that surface mount components existed in the 1970s and in consumer electronics that early. I've taken apart VCRs and game consoles from the 80s and they were DIP mounted and through-hole circuit boards for all of the ones I've come across. I didn't realize the SX-70 was so cutting edge.

Texas Instruments' surface-mount miniDIP chips were "new and risky but cheap", and one of the innovations that made their second design of the Polaroid circuitry successful. Aerospace had been using a lot of surface-mount in the 1960s, e.g. the Apollo Guidance Computer. But consumer electronics didn't really use surface mount until about 1990. I've wondered if there was any particular reason surface mount wasn't popular sooner.

I never knew that aspect of the AGC. I knew the core memory was "sewn" by women at Raytheon but I wasn't aware it wasn't just standard core. The guy who ran software at the Instrumentation lab (and had literally written the book on orbital mechanics) did once give a talk where he related an anecdote of some of the astronauts visiting Raytheon to, as he put it, reinforce that they should really be careful lest these nice young men die.

The AGC is a fascinating device. It used 60% of the ICs produced in the United States in 1963.

If you are interested in the AGC I suggest this excellent video: https://www.youtube.com/watch?v=B1J2RMorJXM&t=454s.

I think he has, in fact, shown a passing interest in the AGC in the past. ;) http://www.righto.com/search/label/Apollo

Haha! I didn't look at the username!

> The AGC is a fascinating device. It used 60% of the ICs produced in the United States in 1963.

As written I interpreted this as saying the AGC used 60% of the kinds of ICs produced in the US in '63.

But it's not that, it used 60% of the production capacity. No, the AGC is not a machine utilizing over half of the IC designs produced at the time. This number says more about the nascent state of the industry the AGC was drawing from than anything else.

> I've wondered if there was any particular reason surface mount wasn't popular sooner.

Soldering, mostly, if I remember correctly.

Discrete components (SMD resistors, capaictors, inductors) really liked to tombstone. Process control was difficult and expensive. Solder wasn't levelled (look at the picture--the solder covers the copper traces) and we didn't have things like ENIG.

This was 1972. Computer-based process control was being invented by these folks.

SMT didn't really get going until the mid-1980s.

I first saw surface mount parts in a camcorder viewfinder - flea market find - I picked up in 1985 or 1986. I don't think I even read about them until the 90s otherwise. I was surprised the first time I found out how far back they went. There must have been a huge chicken-and-egg problem involved in bringing them to the mainstream.

Do you know anything about the manufacturing process? I always connect most of the benefits of surface mount with reflow soldering and pick and place machines. I doubt they used pick and place in 1972 for consumer goods, but maybe you can just have assembly line workers placing by hand before before the line runs through a reflow oven?

There's such a thing as a manual pick and place machine. I agree, though, it would fascinating to know how these were assembled.

I believe they always existed, but were not widely used until the late 1980s when pressure to reduce size and cost became dominant and pin counts on packages rose above the limit for DIP. I worked in the industry at the time and remember the transition. You had to ask assembly houses if they had reflow capability, automatic pick and place, etc, otherwise they couldn't manufacture SMT.

I was surprised by that too. I took apart a lot of electronics as a kid and distinctly remember it being mostly components with leads.

"Instant camera" is a phrase I haven't heard in what feels like a lifetime. Amazing how quickly they lost prevalence.

Hearing it feels like a fever dream, I had to question if I knew the phrase at all.

FWIW they are still a thing: https://duckduckgo.com/?q=instant+camera&ia=shopping&iax=sho...

My subjective, anecdotal sense of trends is that instant cameras were big in the 80s, faded in the 90s, and made a small (but sustained) comeback in the mid-2010s.

The dates will differ but a lot of things (think vinyl albums) probably follow some sort of mainstream, obsolete, retro lifecycle.

They are popular today in the privacy conscious world (including people making nudes)

These were common at Pivotal when I was last there. Everyone gets a snapshot for the pairing board. I bought one and did the same thing at my last company.

Get a Fujifilm Instax and feel the memories come back

I would love to see the SX-70 remade, exact same folding design and size (and leather!) but with a computer printer inside.

This probably isn’t exactly what you mean but still…

https://youtu.be/6CwaT4R5_WA

Author here for your Polaroid camera questions :-)

So the integrator circuit implies it determines the exposure time during the exposure? Meaning, the metering is ongoing during the shutter opening?

Yes. It's a pretty clever circuit. You can think of film as integrating the amount of light that falls on it, so the photodiode integrator is computing the same thing. So the film will get enough light at the same time the integrator gets enough light. By closing the shutter at that time, the film is properly exposed.

The Olympus OM series had some particularly ingenious integrative metering modes, including metering light reflected off the film itself. https://www.dpreview.com/forums/post/24653220

That's very interesting. Just picked up an OM-4 the other week as my first film camera, I knew it was one of the first to have multi-spot metering but just assumed it was standard TTL technology.

Not sure if I'm mis-reading this a surprised reaction, but if so, using light reflected from the film inside the camera body _is_ the standard TTL technology.

https://mir.com.my/rb/photography/hardwares/classics/nikonfe....

Interesting. The parent thread in the linked forum said that other bodies used a different method, so I assumed that was the standard.

> Yep, it applies to almost every AE SLR ever made. Notable exceptions were cameras like the OM-4 that metered off the film during exposure, or others that meter off the closed shutter curtain before exposure. But the vast majority meter from the focus screen so are susceptible to light entering through the eyepiece.

I have an Olympus XA that similarly has a light meter that's active when the shutter's open and closes it when it's "full."

It actually has two light meters, and one of the problems it can have (mine did) is that the light meter that tells you what your exposure will be in the view finder and the light meter inside the body may not agree. The issue was that the internal meter became far slower such that I would see a (correct) value in the viewfinder and then the film would be overexposed by 2 stops. I tested this using slide film and lying to the camera about the ISO by 2 stops and suddenly everything wasn't blown out anymore.

There's a potentiometer hidden inside the body that you can use to adjust the internal sensor's timing (fill rate?) and using audacity to measure the ms between shutter clicks I was able to adjust it to match the viewfinder's reading again.

I have a couple of these cameras. How difficult is it to get an OpenSX70 circuit PCB printed and installed? Is this something achievable by an average tinkerer?

I don't know the details of the OpenSX70 boards; I just looked at the Polaroid chips :-)

Will you be investigating the sonar system developed for later cameras such as the SLR 680?

The sonar system is very cool, but I probably won't be investigating it.

I inherited a Polaroid Supercolor 635 from my dad, who passed away more than twenty years ago. It's just lying in the basement, and I don't know what to do with it. Is it worth anything? Can you still buy film for it?

I have the exact same one. Here's the film that works with it https://us.polaroid.com/collections/film-for-polaroid-600-ca....

Thanks. Maybe I should buy a few films and see if the camera still works.

This is pretty cool. My dad was a Polaroid salesman in the 70’s, so I still have 2 of these and the 1 with the sonar autofocus. Have some of the sales material and his travel case for doing demos.

Loved my old Polaroid in the 80’s too, even when the company was already a shadow of its former self.

It’s sad. They really bet the farm on Polavision[1] and completely ignored video forcing that original vision of instant moving pictures forward. Beta and VHS w/ audio and working on a TV just made the product look dumb.

I used to have one of the projectors but it died to attic heat.

Then I could rant about how their sales practices killed them as well, but that’s likely full of hearsay bias.

[1] https://en.m.wikipedia.org/wiki/Polavision

there's a promotional film made by Charles and Ray Eames about this camera.

https://www.eamesoffice.com/the-work/sx-70-polaroid/

Watch the film:

https://youtu.be/zpv8J8e9gWI

As a programmer, electronics looks like fun. (After dealing with all kinds of thick and dense infrastructure stacks, it’s refreshing to see how things work at the lowest level - where the “stack” below is just Nature herself.)

There is a beauty in "these bits will become those bits because physics says they fuckin' will"

Impressive how back then was possible and "relatively easy" to understand how microchips would work, and also to see what new possibilities they've opened to the general consumers.

500 years from now the technology museums will still have special exhibits about the SX-70, because it's one of the most amazing engineering artifacts of the 20th century. Doubly so because it didn't end World War II or fly to the moon; it was a consumer product with Apollo-grade engineering.

As a kid, I acquired three Kodak instant cameras for free, as they were near worthless. Then Kodak lost a lawsuit from Polaroid and was forced to buy them back at $40 each. I made a huge amount of money for that, as a kid.

Reply on news.ycombinator.com