You have to be of a certain vintage to remember doing research on microfilm and microfiche. Before the age of mass digitization of public records, periodicals, and other obscure bits of history, dead-tree records were optically condensed onto fine-grain film, either in roll form or as flat sheets, which were later enlarged and displayed on a specialized reader. This greatly reduced the storage space needed for documents, but it ended up being a technological dead-end once the computer age rolled around.
This was the problem [CuriousMarc] recently bumped into — a treasure trove of Hewlett-Packard component information on microfiche, but no reader for the diminutive datasheets. So naturally, he built his own microfiche reader. In a stroke of good luck, he had been gifted a low-cost digital microscope that seemed perfect for the job. The scope, with an HD camera and 5″ LCD screen, was geared more toward reflective than transmissive use, though, so [Marc] had trouble getting a decent picture of the microfiches, even with a white paper backing.
Version 2.0 used a cast-off backlight, harvested from a defunct DVD player screen, as a sort of light box for the stage of the microscope. It was just about the perfect size for the microfiches, and the microscope was able to blow up the tiny characters as well as any dedicated microfiche reader could, at a fraction of the price. Check out the video below for details on the build, as well as what [Marc] learned from the data sheets about his jackpot of HP parts.
With the wealth of data stored on microforms, we’re surprised that we haven’t seen any readers like this before. We have talked about microscopic wartime mail, though.
One of the major factors for [Michael] was the great feel of the keys on these classic units. Wanting to experiment with different layouts without a lot of rewiring, the idea of keys with individual displays became attractive. Existing parts on the market were prohibitively expensive, however. Instead, [Michael] used a single touchscreen with a switch mounted underneath to provide tactile feedback with a nifty scissor-arm guide mechanism. Combined with individual see-through plastic overlays, the MP-29 has a fully reconfigurable pad of 30 keys with dynamically updatable labels.
It’s a creative choice, and one that looks highly satisfying to use. It has all the tactile benefits of individual keys, both in the keypresses and being able to navigate the keypad without looking. Combined with the benefit of reconfigurable keys thanks to the touch screen underneath, it’s a great way to build a user-interface.
The rest of the calculator design closely mimics the HP-29, though [Michael] is also experimenting with alternative layouts too. There are plenty of religious wars in the calculator community over usability, after all – mostly over which side of the pad has the arithmetic functions.
See if you can talk your local school district into buying a computer that costs about $5,000 and weighs 40 pounds. That was HP’s proposition to schools back in 1968 so really it is more like $35,000 today. The calculator had a CRT display for the RPN stack that you could mirror on a big screen. You could also get a printer or plotter add-on. Pretty hot stuff for the ’60s.
The 1970 videos promoting the HP 9100, posted by the [Computer History Archive Project], shows something we’d think of as a clunky calculator, although by the standards of the day it was a pretty good one with trig functions and a crude programming capability.
One of the best parts about Hackaday is how much you learn from the projects that people tackle, especially when they are repairs on old gear with unknown failure modes and potentially multiple problems. By the same token, the worst part about Hackaday is seeing what other people are capable of and knowing that you’ve got a long way to go to catch up to them.
A case in point is [Curious Marc]’s recent repair of an old pulse generator. The instrument in question is an H-P 8082A, a device from a time when H-P was a place where “good engineers managed by even better engineers [wanted] to help other engineers,” as [Marc] so eloquently puts it. The instrument was capable of 250 MHz output with complete control over the amplitude, frequency, duty cycle, and rising and falling edge geometry of the pulses, in addition to being able to output double pulses. For an all-analog instrument made in 1974, it was in decent shape, and it still powered up and produced at least the square wave output. But [Marc]’s exploration revealed a few problems, which are detailed and partially addressed in the first video below.
In part two [Marc] goes after the problem behind the pulse delay function. He traced it to a bad IC, which was bad news since it was a custom H-P part using emitter-coupled logic (ECL) to achieve the needed performance that can no longer be sourced. So naturally, [Marc] decided to replace the chip with a custom circuit. The design and simulation of the circuit are detailed in part two, while the non-trivial details of designing a PCB to handle the high-speed signals take up most of part three. We found the details on getting the trace impedance just right fascinating.
In the end, [Marc]’s pulse generator was salvaged. It’ll go into service helping him probe the mysteries of vintage electronics from the Apollo era, so we’re looking forward to seeing more about this great old instrument.
Just because something is newer than something else doesn’t automatically make it better. Of course the opposite is also true, but when it comes to displays on bench multimeters, a fancy LCD display is no guarantee of legibility. Take the Hewlett Packard HP 3478A multimeter; the stock transflective display with its 14-segment characters is so hard to read that people usually have to add a backlight to use it.
That wasn’t good enough for [cyclotronboy], though, who chose to completely replace the stock 3478A display with Nixie tubes. He noticed that with a little modification, six IN-17 tubes just fit in the window vacated by the LCD. He sniffed out the serial data stream going to the display with a collection of XOR gates and flip-flops, which let him write the code for a PIC18F4550. The finished display adds a trio of rectangular LEDs for the + and – indicators, and an HDLO-1414 four-character alphanumeric display to indicate units and the like. And the decimal points? Tiny neon bulbs. It already looks miles better than the stock display, and with the addition of a red filter, it should look even better.
If you’re stuck with a lame LCD multimeter but Nixies don’t quite do it for you, worry not – an LED conversion is possible too.
In the world of oscilloscopes, as in the rest of the test equipment world, there’s always some trickery afoot. Companies will often offer different models to the market at different price points, in an effort to gain the widest possible customer base while also making the most profit. Cheaper, less capable models are often largely identical to more expensive hardware, save for some software or a couple jumpers that disable functionality. [Alexandre] found just this when working to repair his HP 16533A scope card.
Work began when [Alexandre] received his HP 16533A in the mail after a long wait, only to find the trigger functionality was inoperable. This is crucial on a digital scope, so this simply wouldn’t do. After some research online, a post was found discussing which signals to probe to troubleshoot the issue. It noted that corrosion is a common problem on these units, and that occasionally, a certain resistor goes open circuit and causes problems. Initial measurement showed there was still resistance there, but reading closer, [Alexandre] noted this fateful line:
You might not be able to measure it accurately in circuit.
Removing the 100K resistor from the board, the part was indeed open circuit. After replacement with a new component, the trigger circuit was again fully operational. With the scope still open, it was then a simple job to execute a further resistor swap which gives the 16533A the functionality and range of the higher-spec 16534A model.
LEDs are now a mature technology, with all manner of colors and flavors available. However, back in the 1970s, it was early days for this fledgling display tech, and things looked very different. [IMSAI Guy] happened to work at the optoelectronics division of Hewlett-Packard during their development of LED displays, and has a handful of prototypes from those heady days.
The video is a great look at not only vintage display hardware, but also rarely seen prototypes that seldom left the HP offices. Matrix, 7-segment and even 16-segment devices are all in attendance here. There’s great macro photography of the packages, including the now-forgotten bubble displays as well as hermetically sealed glass packages. The parts all have a uniquely 1970s look, drenched in gold plating and otherwise just looking very expensive.