HP in the Time of Apollo

In July of 1975 the USA and the USSR conducted a joint spaceflight called Apollo Soyuz Test Project (ASTP) or, in Russian, Экспериментальный полёт «Аполлон» – «Союз» (ЭПАС). The purpose was primarily political theater; secondarily it was a substantial engineering and cultural challenge — cultural both in the sense of bridging the basic cultural norms of the two countries, and also in the sense of interfacing their very different engineering cultures. It was not a trivial undertaking. While the overt political nature of the flight diminished the public interest that had surrounded previous American manned flights, it was seminal in forging the engineering experience and expertise that would eventually be required to accomplish the Shuttle-Mir dockings — which were a technical prelude to the International Space Station.

The Apollo spacecraft crew cabin normally used a 100% oxygen atmosphere at a pressure of 5 psi; the Soyuz cabin normally used a mixed oxygen-nitrogen atmosphere at 14.7 psi. Apollo and Soyuz both used a probe-and-drogue docking system, but the implementation on the two spacecraft could hardly be more different. These were the main (but by no means only) engineering challenges in accomplishing docking and crew transfer between the two craft. The atmosphere difference could only be handled with an “airlock,” and since each craft docked to opposite ends of this airlock, the docking incompatibility was moot. Despite that, the Soyuz used a completely new, jointly-developed, androgynous docking system, forerunner of the APAS system used today on the ISS. The airlock — which was dubbed the Docking Module (or DM) — was built by Apollo contractor Rockwell International, and launched along with the Apollo CSM in a compartment below it, just as the LM had been on lunar missions.

The HP‑65 was the very first handheld programmable calculator, introduced by Hewlett Packard in 1974. Like all HP calculators of the era, it was operated using a modified Łukasiewicz notation called RPN, for Reverse Polish Notation, based around a 4-level stack. It included the basic arithmetic, trigonometric, logarithmic and exponential functions, square root, and degree-minute-second arithmetic and conversions. Programmability was implemented as a recording of keystrokes, with comparison and branching capabilities. Labels provided destinations for branches, as well as a means to invoke individual routines. Subroutine calls were also supported. A program could contain up to 100 “steps.”

It had a significant innovation beyond its programmability: a motorized magnetic card reader/writer was squeezed into the tiny package, which afforded the capability of developing a set of programs for a particular purpose, recording these programs onto tiny magnetic cards, and later loading the programs back into memory. Since the HP‑65 pre-dated non-volatile memory in calculators, this was a crucial feature.

The HP‑65 was powered by a rechargeable, replaceable 3-cell NiCd battery pack. The display was implemented using the so-called “bubble” LEDs, which had an integral magnifying lens for each 7-segment digit. Inside, the HP‑65 utilized advanced miniaturization techniques, incorporating a hybrid integrated circuit on a multi-layer circuit board a mere 2″ × 3″ in size. The May 1974 Hewlett Packard Journal carried several informative articles on the creation of the HP‑65.

If you’re curious to experience what it was like to use an HP‑65, there is a site with a microcode-level emulator that runs in a browser. If you have an iPhone or iPad, there’s a wonderfully realistic app that’s a lot of fun to use. And finally, you can browse a complete original Owner’s Manual online.

There were five major maneuvers, or “burns”, that were planned to accomplish Apollo’s rendezvous with Soyuz — two “phasing” burns (NC1 and NC2), a “corrective combination” burn (NCC), a co-elliptic burn (NSR), and finally the “terminal phase initiation” burn (TPI). Two Terminal Phase Midcourse (TPM) “tweaking” burns were performed to refine the final approach. The parameters for each of these maneuvers needed to be calculated in real time. The onboard Apollo Guidance Computer (AGC) calculated these parameters and, independently, mainframe computers at Mission Control (in the Real-Time Computer Complex, or RTCC) performed computations based on data from ground-based radar tracking data. NASA being what it is, mission planners preferred to have a third solution in the later phases of the rendezvous to compare with those computed by the AGC and the RTCC. When any two of the solutions agreed, that would be the solution used. If all three agreed, that simply boosted confidence. Programs were written for the HP‑65 to provide that third solution for the NSR, TPI and TPM burns. In the latter phases of the rendezvous, particularly when the Apollo crew was not in communication with ground control, the burn solutions were calculated on the HP‑65, and compared with the AGC and (when available) the RTCC values.

When it was decided to utilize the ATS-6 experimental telecommunications satellite to enhance communications coverage for the flight, a program was developed for the HP‑65 that yielded pitch and yaw angles required for aiming Apollo’s High-Gain Antenna (HGA) at the satellite. Without such a relay, spacecraft in low Earth orbit have only sporadic periods of communication with ground stations; use of the ATS-6 satellite as a relay increased communications coverage from 17% of each day, to 63%.

In poring over the voice transcripts of the flight for mention of the HP‑65, I found only the very briefest of asides, indicating simply that Slayton was using it at scheduled times. (The ASTP Technical Air-to-Ground Transcription ( Part 1 | Part 2 | Part 3) and the ASTP Onboard Voice Transcription (Part 1 | Part 2 | Part 3) are available online.)

In the ASTP Technical Crew Debriefing (JSC-09823, available via FOIA request), commenting on the rendezvous maneuvers, the crew reported “All the maneuvers were accomplished in good shape and the solutions came out good and matched with the ground solutions. There was very little scatter [i.e. difference in the solutions], including the HP‑65.” When questioned on the usefulness of the HP‑65, Deke Slayton’s evaluation was that it “worked like a charm.” What sounds like glowing praise, however, was simply Slayton’s way of saying he had no problems with it — it was a phrase that he used frequently.

Qualification Test Report for Programmable Calculator Kit — This document provides a fair bit of information about the “kit” that the HP‑65 was packaged into, but nothing at all about the programming. Among other things, we learn about the procurement; about the soft materials (i.e. the cases that were fabricated using non-flammable Beta cloth); and about the testing that was performed on the calculator to ensure that it would continue to function properly in the extreme conditions it would encounter, and would not compromise spacecraft functionality or crew safety.

As to procurement, in all 3 HP‑65s were procured by NASA’s general Apollo program contractor, General Electric, along with 20 spare battery packs. Two calculators were for qualification, training and flight, and one was for program development. GE was further tasked with fabricating soft cases for the two calculators to be carried on the flight, a holder for the program cards, and a soft case to hold the complete complement of two calculators, the program cards, six spare battery packs, and a copy of the HP-65 Quick Reference Guide.

Testing included evaluation of the HP‑65’s tolerance to shock and vibration, temperature, humidity, acoustic noise, acceleration, radiation, and low atmospheric pressure; and its offgassing properties (which were tested at JSC’s White Sands Test Facility) and electromagnetic interference. While it passed all the other evaluations, the HP‑65 failed the EMI emissions standards, so a test was scheduled and performed on 19 August 1974 in CSM 119 (the ASTP backup vehicle — the ASTP primary vehicle was CSM 111) at KSC to determine if any CSM systems were affected by this EMI. They were not, and a waiver was requested to allow the HP‑65 to be used on the flight. That waiver was approved on 26 September 1974.

Two pertinent engineering drawings extracted from this rather lengthy document are available here: SJF12100337 Page 1 (0.9Mb) | SJF12100337 Page 2 (1.2Mb)

Mission Techniques Rendezvous Book — This planning document describes all of the maneuvers employed to accomplish rendezvous with the Soyuz. It explicitly describes the HP‑65 programs as a replacement for the onboard paper charts that had provided backup solutions on previous flights. Procedures are given for both the Apollo Guidance Computer (AGC) and the HP‑65 as well as rationales for those procedures. Since the HP‑65 was nominally a backup to the AGC, the crew was to obtain input values for it independently from the AGC: range values were to be obtained from the VHF ranging capability of the EMS (Entry Monitoring System), and angles were to be measured using the COAS (Crewman Optical Alignment Sight). Solutions were computed on the HP‑65 all during the rendezvous — both in order to compare to the AGC solutions, and also to be prepared to continue the rendezvous in the (admittedly very unlikely) event of a loss of AGC functionality.

CSM Rendezvous Book [Part 1 (2.3Mb)] [Part 2 (2.7Mb)] — This onboard document provides some insight into when and how often the HP‑65 was used. The Rendezvous Book is essentially a timeline for performing the tasks required to rendezvous with the Soyuz. Alongside the timeline are rules, checklists, and places to write values supplied by the AGC, the HP‑65 and Mission Control.

Page 1-12 has the first reference to the HP‑65: UNSTOW HP-65 KIT (U1) (“U1” identifies the compartment where the HP‑65 was stored). This is followed by HP-65 CHECKOUT, indicating that it is to be verified for proper operation. Three program cards were supplied for this purpose.

Shortly thereafter the direction ACQUIRE ATS HGA is listed, along with a place to write in the pitch and yaw values which were calculated by the HP‑65.

Numerous times we see items like R HP-65 DATA FOR NSR-28. This is a direction to record (“R”) the values calculated by the HP‑65 — in this case, for the NSR (coelliptic) maneuver at 28 minutes prior to the burn. “Record” simply means to write down the values in the Rendezvous Book.

Closer to the time of the burn, there are “decision boxes” that summarize how the crew was to decide which values to use for the maneuver — the possible options were: AGC, STDN (values relayed from the ground), or HP‑65.

HP-65 Rendezvous Targeting Checklist (2.2Mb) — This onboard document is essentially a user’s manual for employing the programs supplied for the ASTP mission. It’s moderately interesting, but really doesn’t give any insight into the internals of the programs, nor even the use of the calculated results. It gives step-by-step instructions on how to use:

• Diagnostic programs that prove the calculator is working properly (3 cards);
• NSR – coelliptic maneuver (5 cards);
• TPI – Transfer Phase Initiation (4 cards);
• TPI TIG – TPI Time of Ignition (1 card);
• TPM – Terminal Phase Midcourse (2 cards);
• The ATS pointing program (4 cards)

Previous to the ASTP flight, an HP-35 calculator was carried on the last two Skylab flights. Jack Lousma, the CM Pilot on Skylab 3, recalled in Homesteading Space: The Skylab Story:

I had to make ... backup calculations on the closure rate. I was sitting there with this little HP calculator and punching all those numbers in, going through this formula and backing up what the ground saw and what we saw in the spacecraft. There had to be a third vote and that was me. I never enjoyed making that calculation. You had to get it right. If you missed one keystroke, you had to start all over again and it was a long one.

As a result of the positive experiences on the Skylab flights, an effort was undertaken by the JSC Crew Training and Procedures Division to reduce the manual calculations to a series of programs that could be loaded into the programmable HP-65 from magnetic cards as needed. This project was led by Duane Mosel, of that organization. Mosel was well qualified to lead the project, as he had earlier authored both the Apollo 7 and Apollo 10 Rendezvous Procedures documents, and the Skylab SL-4 CSM Rendezvous Book.

Mosel recruited Mason Mines (who worked for contractor McDonnell Douglas Technical Services Company) to code the rendezvous programs. As a natural outgrowth of that task, Mines also authored the Rendezvous Targeting Checklist document.

Mike Hollars, at the time a co-op student who was studying at the University of Texas at Austin, was tasked with coding the ATS pointing program.

The Smithsonian National Air and Space Museum has an HP‑65 in its collection that is clearly from the ASTP flight kit. (The Smithsonian National Museum of American History also has an HP‑65 in its collection, but that one has no connection to the ASTP mission.) The Air & Space Museum’s HP‑65 (which is assigned Inventory number A20120307000) includes the fireproof soft cases that were fabricated from Beta cloth, and the magnetic card holder (inventory number A20120307001), complete with the magnetic cards that contain the flown software. The Museum notes that “NASA transferred this device to the Museum with a variety of crew equipment when the Space Shuttle program ended in 2012.”

No mention is made of a second HP‑65, and only a single calculator and small Beta cloth pouch is shown in photos of this object. Further, the curator’s notes indicate that there are 3 spare battery packs included (inventory number A20120307002), which is only half of the complement of 6 that were included in the flight kit. So somewhere in the world, there is another HP‑65 that was flown into space on ASTP, including the calculator case and 3 spare battery packs.

In addition to an advertisement, HP issued a press release about the HP calculator carried on the ASTP flight. The release included a photo of three technicians putting together the “kit.” It carried this caption:

D. Mosel of NASA (center) and G. Riddle (left) and M. Mines, both of McDonnell Douglas Corporation, display the HP‑65 calculator that will be carried aboard the Apollo Command Module during the upcoming Apollo/Soyuz Test Project scheduled for launch on July 15, 1975. The 10-day mission is to demonstrate and test a common docking system and the performance of joint experiments between the American and Soviet crew members (NASA photo)

Note that in the photo you can see two calculators (one is face-down near Mines’ left hand), and two small soft cases, as well as the largish soft case that contained the entire “kit.” It appears that the Mines is applying tape over the battery contacts, as specified in the Test Report.

Note that the photo is credited to NASA, but no NASA photo ID is provided.

The press release:

Palo Alto, June 23.

An 11-ounce, $795 pocket calculator that can be programmed like a computer will play an important role in the historic Apollo/Soyuz rendezvous in space on July 17.

The Hewlett-Packard HP‑65 fully programmable pocket calculator will be used to calculate two critical mid-course correction maneuvers just prior to the linkup of the U.S. Apollo and the Russian Soyuz spacecraft. These maneuvers will take place 12 and 24 minutes after terminal phase initiation (the beginning of the last part of the flight before rendezvous).

It will also be used as a backup for Apollo’s onboard computer for the final maneuvers prior to rendezvous and docking. The first use will be for the coelliptic maneuver (putting both spacecraft into the same orbit) when the vehicles are within approximately 100 miles of each other. The second will be for the terminal phase initiation calculations when Apollo is 22 miles from Soyuz. In both instances, the HP‑65 will be used to solve the problems, and its answers will be compared with those of the onboard computer.

In the event of an onboard computer failure, however, the HP‑65 will have the only available solution for the mid-course maneuvers, since the spacecraft will not be in communication with ground stations at that phase of the mission.

A third set of calculations to be performed by the battery-powered HP‑65 will allow the astronauts to precisely point Apollo’s high-gain antenna at an orbiting satellite to assure proper communications with Earth.

NASA scientists have written programs of up to 1,000 steps and recorded them on the HP‑65’s magnetic cards (100 steps per card) that the astronauts will feed into the HP‑65 to automatically perform the critical calculations. In previous space flights, backup maneuver calculations were made manually, using charts. The HP‑65 will substantially reduce the time needed to make the complex calculations and improve the quality, accuracy, and confidence in resulting solutions.

Two HP‑65s will be taken on the space flight, along with four sets of program cards and six spare battery packs.

The HP‑65 is not the first HP pocket calculator to venture into space; an earlier model, the HP-35, went along on the Skylab missions.

HP later created a poster using a staged photograph of a space-suited figure loading a program card into an HP-65. It is painfully obvious that this was staged — there are just so many things wrong with it. The most obvious is the fact of the full Earth being visible through the window — the spacecraft would have to be very, very far from the Earth in order for the full globe to be visible at that size. ASTP was in low Earth orbit, so even just Florida wouldn’t fit in the window. The spacesuited figure is complete with helmet and gloves — these would be removed as soon as the spacecraft reached orbit: by the time the HP-65 was actually used for the first time, they had removed their spacesuits completely and were in “shirt-sleeves.” And for anyone familiar with the layout of the Apollo CM, that window is depicted in the lower equipment bay — no windows exist in that part of the spacecraft. There are lots of other anomalies but, after all, it is just a publicity image, so it really doesn’t matter. My point is this: don’t believe the spurious captions that claim it shows the HP-65 “in flight.”

If you have information about the HP‑65s flown on the ASTP mission, or about the programs written for them, I’d be delighted if you’d share it. You can send it anonymously if you prefer, but if you decide to share your e-mail address with me, be assured I will only use it if I have a question about the information you’ve shared.

Your name Your E-mail address Your comments