No lab in almost any discipline was complete in the 70s and 80s without an X-Y plotter. The height of data acquisition chic, these simple devices were connected to almost anything that produced an analog output worth saving. Digital data acquisition pushed these devices to the curb, but they’re easily found, cheap, and it’s worth a look under the hood to see what made these things tick.
The HP-7044A that [Kerry Wong] scored off eBay is in remarkably good shape four decades after leaving the factory. While the accessory pack that came with it shows its age with dried up pens and disintegrating foam, the plotter betrays itself only by the yellowish cast to its original beige case. Inside, the plotter looks pristine. Completely analog with the only chips being some op-amps in TO-5 cans, everything is in great shape, even the high-voltage power supply used to electrostatically hold the paper to the plotter’s bed. Anyone hoping for at least a re-capping will be disappointed; H-P built things to last back in the day.
[Kerry] puts the plotter through its paces by programming an Arduino to generate a Lorenz attractor, a set of differential equations with chaotic solutions that’s perfect for an X-Y plotter. The video below shows the mesmerizing butterfly taking shape. Given the plotter’s similarity to an oscilloscope, we wonder if some SDR-based Lissajous patterns might be a fun test as well, or how it would handle musical mushrooms.
” Anyone hoping for at least a re-capping will be disappointed; H-P built things to last back in the day.”
Maybe one should start a “back in the day, things lasted” museum? There should be plenty of exhibits.
And the mandatory XKCD reference
https://xkcd.com/1827/
I think, generally, any museum is–to some extent–a “built to last” museum. Right?
Needs more analog. Analog computer for lorenz attractor: https://www.youtube.com/watch?v=DBteowmSN8g
Chua’s circuit in some form would be a nice thing to drive an old plotter with.
It seems to me that HP’s soul has moved on to Keysight. HP’s not what it once was but Keysight is.
High end best of the class test gear to low end PC crap. What a transformation……
I saw a lot of what was called “chart recorders” back in those days. They had a roll of chart paper, a motor driven x axis and a long analog “needle” for the y axis. The needle was like a big VU meter with a small pen at the end.
They were used to record overnight or through the week.
They still make and use them. One company I worked for puts one on all of their large walk-in curing ovens. It’s records the exact temp for each batch being cured, and the paper chart goes to the customer (mostly automotive makers) with every shipment or goes in a file, depending on the what the customer wants.
They want to be able to verify even the ramp up and ramp down temps over the entire 8 – 12 hour cure, and they’re apt to refuse a batch that’s not right on the money.
I would like to have one with some kind of disappearing ink.
Steampunked-up and brass everywhere. Let it run constantly for weather data.
Uh, now I want one like you describe. Crap, another project on the list….
We use something similar for recording in Biology, but the recording surface is a disc of paper. The paper is mounted on a disc that rotates slowly; once a day to once a week, depending on what you are measuring. You can either replace the paper once a week, or just show up to write down the observations and then replace the paper once it approached being to cluttered to read.
We the same disk based thermometer in the “computer room” (a couple of PDP-11s) to make sure the cooling system worked over night.
IIRC, the plotters were great…when used with the very expensive HP pens (roller, fiber or ink). Surprised he got it working well with what looks like a BIC ball point.
Of course, ball point pens may have improved since then….
CALCOMP drum plotter…pin feed paper and a whole ink-stained table full of partially used pens :-)
I think that’s the model we used in the CAD lab.
Ours was in the back hallway of the Univ. Computer Center. The operators hated the damn thing, because plots took a good part of an hour and the damn pens would run out of ink halfway through.
It was hooked through a peripheral controller to the CDC Cyber 74 timesharing system.
Good times :-)
We had something like this at a place I used to work. I think it was a Calcomp, but it’s been 27 years now, so it’s hard to remember. The pens always did run out at the most inconvenient times during the very long plotting process. We used it to do colored plots of hybrid circuits (screened onto alumina and fired in an oven before the parts were soldered on, used underhood in cars) After we got a color electrostatic plotter, the pen plotter was mostly used for off-hours art projects. This was all connected to a Data General 16 bit Eclipse with 1M of RAM, 2 washing machine sized 300M disk drives, running 4 graphic terminals and one green screen terminal. The software was Calma GDSII.
Pretty much this:
https://ub.fnwi.uva.nl/computermuseum/calcomp565.html
I’ve got a rather nice plotted in my local Fab Lab.. Made by Roland DG, it’s actually a vinyl cutter, but you just throw it HPGL(Strait out of Inkscape!) over usb(Good ‘ol cat > ) and if you replace the cutter with a pen, it does a nice job of that too.
Did you guys buy it, or was it donated? Sounds very cool…especially the vinyl cutting part…
7044A is a X-Y Recorder, not plotter. There is a difference.
I have used them for 30 years. They are great for 1/10 sec
to 10 sec rates. Faster leaves no ink. Slower leaves too
much ink.
Acknowledged. Other than the motors, the mechanics are generally similar, though. Servos for the X-Y Recorder and stepper motors for the plotter.
With some kind of paper feed mechanism and an appropriate renderer, I’m certain some enterprising soul could get a few frames per minute of DOOM out of this.
SDR based Lissajous patterns? These plotters move pens in real-time. I don’t think you need SDR to generate signals at 1 Hz.