HP9825<9600Bd>HP9825+MultiIO= Rocket Console

[Tony Stewart]

My career started in 1975 as EE at the University of Manitoba in Winnipeg. I was started in Aerospace as an Electronic Instrument Designer at Bristol Aerospace R&D group. (now Magellon). Walking every day thru bldg 3 to the lunch room past fighter jets and choppers getting upgraded and parts for Candu reactors being welded and L1011 turbines being made for Boeing, I was like a kid in a candy store. Although an average student by grades, I preferred to get my "hands dirty" by design. So I scanned every back-issue magazine in the Eng. library for old designs and the HP Journal was like a dream.

My first calculator was the Slide Rule, my 2nd was the magnificent HP-9825. In 3 months, I will have to create a complete system to remote control Black Brandt sounding rockets used for scientific research in the Arctic thru 2 umbilical cords in the remote launch building. These rockets reach speeds of Mach 7 and are gone from sight in about 5 seconds coast up to where future satellites go then often come down somewhere in Hudson's Bay.

What I was about to learn was how to design a replacement for the old manual console with 100 power buttons and a rotary switch with 40 positions for each umbilical. Churchill Mb was an interesting place to launch rockets. Where else can you find a WWII landing strip for B-52's with roaming Polar Bears going from the pub to the grocery store for early spring food with a skating rink, auditorium, bowling alley hospital, school, gym, pool, library all under one roof facing Hudson's Bay?

The HP9825 Computer Silicon on Saphire, fast, sleek and smart. My 1st computer)

In my spare time, others would ask me to measure the phase modulation on their RF GOES transmitter design using the HP Time Interval Counter using the HPIB port. That would only take a couple of hours of reading the manuals and writing a small program that spits out the PPM error and modulation index on power up after 1ms, 10 ms by measuring the change in the RF carrier time interval.
This sure beat my slide rule and scope. Meanwhile, my mentor would play games on his NOVA minicomputer with the 3D chess program he wrote.

The HP9825 task
Rocket launching requires a small team to follow their own payload checkout procedures for a few days then every night, wait until the Head Scientist from NRC to find the best conditions for Northern Lights> Suddenly he says , it's starting to pickup on his kilo-Rayleigh gun then the countdown starts, T-180 Once, it jumped to T-30 seconds, right away and the payload chief was hurrying thru all the rotary positions to check every battery state of charge and turn on the experimental instruments for launch with feedback lamps, then when all was green, he said Payload is a GO. The adreneline was pumping, to check everything is ready at the last second.... Zero, Liftoff, says the launch director and the squid ignites the solid Nike booster which leaves the launch-rail about 250 MPH.

The equipment
This new Auto-Console would replace all the mechanical switches and voltmeters with two HP9825's talking through a Limited Distance Dataset at 9600 Baud to the HP I/O Multiprogrammer and a big box full of 96 power relays for backpanel interconnection of DC power with extra contacts for relay feedback.

In the end, I wanted to have 2 pages of data in form fills on the smart HP terminal like a spreadsheet with inverse video for altered-state and flashing if out of tolerance. One remote HP9825 was the remote console. To use single key toggle action to a remote job, I used keycodes with the numeric keypad for adjusting up to 10 power supplies for charging payload silver-ion battery packs and dozens of instruments to enable such as Plasma Ion-detectors, RF + optical spectrometers.

In order to config all the variables and assign them to each specific task, a 1st program was generated with prompts from the Payload Engineer to type in with tolerances for control limits on analog processes. This created the table used by the main program. I forget most of the details but in modern data, it might be considered a Supervisory Control And Data Acquisition or SCADA control system. I had to use DMA to process the incoming UART data with feedback on every change to validate the control, for continuous frames of payload data every second like a Real-Time Operating System in HPL with 96 channels of data. The LED display was used like a Digital Multimeter to read every battery voltage on board and adjust each power supply for umbilical line loss.

The system state was stored in non-volatile tape so that if AC power failed, it would resume to the previous state automatically.

The HP equipment worked flawlessly with an overlay I made for the keyboard to control remote and instant feedback on the LED display and CRT Smart terminal. The mercury switched differential analog converter ADC in the IO Multiprogrammer worked well. Although one board failed during development but worked on an extension board. I found the culprit IC with cold spray and got 'er fixed. The HP power supplies were remote-controlled with Ohms/volt and the I/O resistor DACs.

I learned how to integrate a system with the HP Basic calculator doing something it does not... control a rocket on the ground. It was a fun project.

I also learned how to prevent relay failures from oxidation. The P&B extra signal contacts were only rated for <2A but at that time were not gold plated. Today every relay maker uses gold contacts as the silver alloy contacts do not work well for TTL signals drawing microamps of current. The signal relay contacts were intermittent, and with no margin for rocket failure, I had to find a solution. Today all signal relays < 2A are gold-plated. The silver oxidizes and becomes an insulator but is a better conductor, but TTL will not "wet" the contacts.

The gattling gun test

When the relay box was built with 5V resistor pullup on the spare contacts. I wrote a self-test program to toggle the next relay after the first one in a looping sequence with the HP9825. rat-tat-tat-tat- pause .. It would go for spurts like a machine gun then run out of bullets. I could even bump the box and it would continue on for a spell. So I that Carl Jung AHA experience that every engineer has when the "light bulb" goes on. Wetting requires at least 10% rated current to burn off the oxide layer and keep the resistance low. So the solution was add 22 uF Tantalum low ESR caps across the contacts. Ran it for 10 minutes and annoyed everyone in the lab. No signs of wear. Fixed.

A lot of projects were like this in my career. Bleeding edge. Like the SCADA robotic eddy current heat-exchanger sanning subsystem or DS1 to 100 home trial in 1980 with an ISDN broadband WAN , with alarms, meter reading, digital voice, opinion polling , pay TV, Graphic weather channel, opinion polling.

But the HP9825 had a quirk soul with whizzing tape drive scrolling 32 column LED display that takes a special design to multiplex this many LED digits and "Basically" do anything I wanted with any of the wall full of high tech instruments at my disposal.

[KeithB]

The museum at the White Sands Missile range has a lot of this kind of cool stuff.

[Martin Hepperle]

Tony,

thank you for sharing your interesting story. It is always interesting to learn about the various applications of the HP "calculators" in the field.

I have been involved for some years in wind tunnel testing and there HP 1000 systems were used for many years, switching to PCs just before I started. During my first week in the new office, the old HP 1000 equipment was thrown away, but at the time I was not yet interested in HP history.
Only the very accurate HP 3456 DVMs were used for many years with PCs and HPIB (I have 3 of them at home for usage with my HP gear).

You might also want to have a look at this discussion group:
https://groups.io/g/VintHPcom/topics

Martin

[Guenter Schink]

Off topic

(08-03-2022 09:11 PM)Tony Stewart Wrote:  My career started in 1975 as EE at the University of Manitoba in Winnipeg.

Ahh, [memories] In 1974 the German Air Force started to carry German army troops to Winnipeg. From there they went to Shiloh Range for their 3 weeks combat training. I was a young B-707 copilot then, flying into Winnipeg almost every three weeks during summer.
Perhaps you've seen our nice aircrafts, white with a blue writing "LUFTWAFFE" on it. Last time for me was in 1984, as captain then, or as we called it: aircraft commander. [/memories]

Günter