The designers of older equipment that contained a CRT monitor rarely made the effort to design their own driver and deflection circuitry. Instead they were more likely to buy an off the shelf assembly from a monitor manufacturer, and simply supply it with their video. [TomV] has an old HP 16500A logic analyzer, and in it he found a Sony monitor chassis. With a quest for a microfiche service manual and a bit of reverse engineering, he was able to hook it up to a VGA port and use it as an extension monitor for his laptop.
The monitor chassis is a Sony CHM-9001-00, which sports their 10″ Trinitron tube. These were among the very best CRT tubes of the day, making it the type of module 1990s hacker would have been very pleased to get their hands on. Here in 2022 a look at the monitor’s 40-pin connector reveals a standard RGB interface which the service manual confirms is within the voltage range to be driven from a VGA output. A Thinkpad X220 is pressed into service, with a 576 by 360 pixel at 60 Hz video mode defined, and there we have it, a modern desktop on an obsolete piece of test equipment.
The intended destination for this monitor is a small arcade cabinet, so it needed to be independent of the HP chassis. The required 120 VDC supply comes from an inverter designed for solar battery charging, which balked at the inrush current from the monitor when fed with 12 V. Increasing the supply voltage on the low voltage side solved that, leading to a very serviceable monitor. We have no use for one, but we’d be lying if we said we didn’t want one.
Perhaps you may have wondered, what made Trinitrons so good?
The fun thing about CRT monitors is how they’re manufactured.
There’s the shadow or aperture mask which guides the image of each electron gun into the correct place on the screen. But how do they get the different phosphors to line up with the mask? Simple: they pour the phosphor resins in the tube and then shine UV light through the mask from each of the electron gun positions, which sets the resin only where it’s supposed to go. When the actual electron gun is placed in the tube, it goes in the position where the UV light source was.
Genius. Yeah, imagine the moire nightmare of trying to align the shadow screen after the phosphors were already placed! Not gonna work.
There’s so many interesting little tricks in tube manufacturing which are basically becoming lost to time, like Roman concrete. Not that we need tubes for much anymore, but they sure were nice for learning a lot of physics and electronics concepts when I was younger… Might just be me getting old and ossified, but it seems like they had a better illustrative quality than the solid-state stuff. They certainly tolerated a lot more abuse before going up in a puff of smoke!
The benefit of a CRT was the inherent anti-aliasing and nicer contrast. Everything else was kind of worse unless you paid a lot.
The difference is that the dot pitch of the display is not the same thing as its “resolution”. A pixel is not a sharply defined box, but more like a bell distribution of intensity, so the sharp square pixels of an LCD monitor are rather an artifact of the device rather than any real information present in the image. However, it creates the illusion of “sharpness” that CRT monitors did not have, which helps to mask the fact that the common display technology we have is actually fuzzier than what came before it.
I’ll take LCD with ClearType any time over even the best Trinitron.
There’s a video on YouTube about the worst VGA CRT monitor. Radio Shack had it as the cheap option with their PCs. Turns out it’s a tube made originally for CGA and EGA monitors in a VGA monitor chassis. Not actually capable of displaying a 640×480 pixel image but the circuitry splats it onto the front of it.
Looking at the dot pitch and comparing to ads for other low cost VGA monitors of the time, Tandy wasn’t the only company guilty of using old-tech tubes.
Great now also the A ones get up in price 😉
Reads summary, reads article, sees Tom’s Atari Joystick + I2C Expander board and only then realizes that it’s Tom’s project.
I used to use those logic analyzers. They were fantastic when you were trying to debug network hardware. They were also large, very heavy, and power hogs. I do not miss them. TFT displays have made the workbench so much less cluttered.
And FPGAs. Once the debuggee, now the debugger.
I got a 16500C and added some modern mods to it but adding VGA is a little too much.
The HP 16500 is by NO means an obsolete logic analyzer. It’s big, heavy and old, but still useful as hell. Mine’s currently loaded up with 170 channels of 500 MHz logic analysis, a 500MHz oscilloscope card, and a 60 channel pattern generator.
Noice! What do you work on with that beast? How much did it cost?
maybe a hundred bucks? I found it on the vintage test equipment FB group – the real bargain was that the guy was only 20 blocks away from me, so I didn’t have to pay another hundred bucks in shipping! (plus it came with all the cables and grabbers).
I’ve never hit anywhere near that channel count in practical use, but watching an 8 bit data bus, a 16 bit address bus, and a bunch of misc signals on an old 6502 or Z80 machine is a piece of cake.
IIRC 16500A works only with the oldest modules that don’t have anywhere near those specs. The 16500B and C can take higher speed modules, has an HDD, networking (optional on B) for remote X and file transfer, and more capable disassemblers. 16500A is also limited to HP LIF format DS/DD disks which can’t be written by a modern PC with a USB floppy drive, so unless you have a vintage PC to write the boot media, you’re going to need a friend with another 16500A or capable PC to create disks.
The 16702 is the sweet spot of the series as it’s old enough to be cheap-ish yet very capable for a highly parallel LA and somewhat smaller and lighter than the 165xx series due to having an LCD instead of a Trinitron. It works with many of the 165xx modules as well as 167xx boards. Also the PA-RISC mainframe controller running HP/UX doubles as a retrocomputing relic. The only concern when buying one is that HP used an adhesive backer on some of the modules that corrodes the PCBs over time, though you can usually fix these with a little cleaning and jumpering.
You’re correct – mine is a ‘B’ Variant, and I forgot that the ‘A’ couldn’t take the newer cards, like the 16555A or the 16534A. The lack of HDD and LIF floppies were enough to scare me away from the ‘A’ version.
+1! It’s only a power hog (and deafening) when it’s turned on!
I got an oscilloscope, and some nifty extra cards, high speed, so it’s well stuffed.
Mine was only $70 or so, but I bought two, they had to be shipped from the states. The other one was bought for the company I worked for at the time, so I got both shipped in the same box, job done!
I worked out how to make extra HP “pod” connectors – they are basically a 91k series resistor in each logic line, so I can add and use multiple pods at once pretty cheaply.
I like the idea of splitting out the VGA and inserting an RPi inside it, which could then take over or augment the running of the beast.
What extras do you have? I hooked mine up to Ethernet via an AUI adapter and can use it remotely. Back in the day I also wrote a parser that could scan over captured files and extract binary data. I wonder if I still have the that code somewhere?