Because you have skipped quickly over the decade before WWII when there were great advances in designing digital circuits with vacuum tubes, mostly in the UK, which have lead directly to the electronic circuits used in the British Colossus and in the US Atanasoff-Berry computer, and then in ENIAC, I want to fill in the gap between your 2 sentences:
> The Eccles-Jordan trigger eventually led to digital counters. In 1939, ...
After the Eccles-Jordan trigger mentioned by you, the following important steps were:
1929-07: “Hot-Cathode Thyratrons. Part II”, Albert W. Hull, where the inventor of thyratrons (at General Electric) describes a bistable circuit made with a pair of thyratrons, the second such circuit after the Eccles-Jordan trigger.
1931-01: “Note on the use of a thyratron with a Geiger counter”, N. A. de Bruyne and H. C. Webster, where it is suggested that the pulses produced by the Geiger counters for detecting radiation in nuclear physics experiments can be counted with electronic counters using thyratrons, since the existing mechanical counters were too slow for counting such pulses.
1931-07-02: “The Use of Thyratrons for High Speed Automatic Counting of Physical Phenomena”, Charles Eryl Wynn-Williams, where the first electronic digital counter is described. This counter was made with thyratrons and it was a decimal counter. For each decimal digit, a ring counter with 10 thyratrons was used. So the ring counter is the second kind of digital electronic circuit, after the simple bistable.
1932-05-02: “A Thyratron Scale-of-Two Automatic Counter”, Charles Eryl Wynn-Williams, where the first binary counter is described. It was also made with thyratrons, but it was an improvement over the decimal counter with ring counters per digit, as it required much less thyratrons and it was also more reliable.
1937-10: “A Scale-of-Two High-Speed Counter Using Hard Vacuum Triodes”, W. B. Lewis, where the first binary counter with vacuum tubes was described. This was an improvement over the counters with thyratrons, being much faster. The article "Trigger Circuits" from 1939, linked in TFA, cites this paper for its binary counter.
1938-01: “A Thermionic Trigger”, Otto H. Schmitt, where the Schmitt trigger made with vacuum tubes was described. The Eccles-Jordan trigger (a.k.a. R-S bistable) and the Schmitt trigger are 2 fundamental kinds of bistable circuits (i.e. the 3 values needed at the input of a bistable circuit, set, reset and keep the current value, can be encoded either with 2 binary inputs, as in the Eccles-Jordan trigger, or with 1 ternary input, as in the Schmitt trigger). Schmitt was American, but he invented the Schmitt trigger while working in the UK, where before WWII most of the work related to electronic digital circuits was done, mainly for applications in nuclear physics research.
Over on Youtube the Usagi Electric channel has several videos about reviving old vacuum tube computers and one of the interesting aspects is that tube failures are much less of an issue than the germanium diodes. The tubes are large and dramatic, but they're heavily outnumbered by the diodes. It was such a problem that the company built specialized test equipment that could identify bad diodes on the cards.
> I'm also curious, if the cathode was grounded to the machine frame, how dangerous was it to do work on those machines?
When ground means earth, then: The chassis isn't dangerous at all. Its potential relative to other nearby things that are also connected to earth is ~0v.
When that's used as a supply rail (like for the cathode), it still works fine. It's still tied to earth.
You don't want to put any of your dinguses betwixt that and the +150v plate rail, but... I mean. :)
Vacuum tubes don't have the longest "natural" lifespan, but you also have to remember that early manufacturing techniques mean a lot more "bad batch" variability.
When part performance is more unpredictable, ease of maintenance is a much bigger deal for everybody all across the price/quality range.
I think it would have been more likely to be some kind of mechanical analog computer full of gears and cams (which some planes and missiles did use for guidance). Probably needing more support from the ground.
Apollo did carry tubes also, in the radio transmitters.
Author here. I've powered up another tube module from the IBM 604, so let me know if you have questions...
Thanks for another interesting article.
Because you have skipped quickly over the decade before WWII when there were great advances in designing digital circuits with vacuum tubes, mostly in the UK, which have lead directly to the electronic circuits used in the British Colossus and in the US Atanasoff-Berry computer, and then in ENIAC, I want to fill in the gap between your 2 sentences:
> The Eccles-Jordan trigger eventually led to digital counters. In 1939, ...
After the Eccles-Jordan trigger mentioned by you, the following important steps were:
1929-07: “Hot-Cathode Thyratrons. Part II”, Albert W. Hull, where the inventor of thyratrons (at General Electric) describes a bistable circuit made with a pair of thyratrons, the second such circuit after the Eccles-Jordan trigger.
1931-01: “Note on the use of a thyratron with a Geiger counter”, N. A. de Bruyne and H. C. Webster, where it is suggested that the pulses produced by the Geiger counters for detecting radiation in nuclear physics experiments can be counted with electronic counters using thyratrons, since the existing mechanical counters were too slow for counting such pulses.
1931-07-02: “The Use of Thyratrons for High Speed Automatic Counting of Physical Phenomena”, Charles Eryl Wynn-Williams, where the first electronic digital counter is described. This counter was made with thyratrons and it was a decimal counter. For each decimal digit, a ring counter with 10 thyratrons was used. So the ring counter is the second kind of digital electronic circuit, after the simple bistable.
1932-05-02: “A Thyratron Scale-of-Two Automatic Counter”, Charles Eryl Wynn-Williams, where the first binary counter is described. It was also made with thyratrons, but it was an improvement over the decimal counter with ring counters per digit, as it required much less thyratrons and it was also more reliable.
1937-10: “A Scale-of-Two High-Speed Counter Using Hard Vacuum Triodes”, W. B. Lewis, where the first binary counter with vacuum tubes was described. This was an improvement over the counters with thyratrons, being much faster. The article "Trigger Circuits" from 1939, linked in TFA, cites this paper for its binary counter.
1938-01: “A Thermionic Trigger”, Otto H. Schmitt, where the Schmitt trigger made with vacuum tubes was described. The Eccles-Jordan trigger (a.k.a. R-S bistable) and the Schmitt trigger are 2 fundamental kinds of bistable circuits (i.e. the 3 values needed at the input of a bistable circuit, set, reset and keep the current value, can be encoded either with 2 binary inputs, as in the Eccles-Jordan trigger, or with 1 ternary input, as in the Schmitt trigger). Schmitt was American, but he invented the Schmitt trigger while working in the UK, where before WWII most of the work related to electronic digital circuits was done, mainly for applications in nuclear physics research.
What was the failure rate of those vacuum tubes, that they had to be so quick & easy to replace?
I'm also curious, if the cathode was grounded to the machine frame, how dangerous was it to do work on those machines?
Over on Youtube the Usagi Electric channel has several videos about reviving old vacuum tube computers and one of the interesting aspects is that tube failures are much less of an issue than the germanium diodes. The tubes are large and dramatic, but they're heavily outnumbered by the diodes. It was such a problem that the company built specialized test equipment that could identify bad diodes on the cards.
> I'm also curious, if the cathode was grounded to the machine frame, how dangerous was it to do work on those machines?
When ground means earth, then: The chassis isn't dangerous at all. Its potential relative to other nearby things that are also connected to earth is ~0v.
When that's used as a supply rail (like for the cathode), it still works fine. It's still tied to earth.
You don't want to put any of your dinguses betwixt that and the +150v plate rail, but... I mean. :)
Vacuum tubes don't have the longest "natural" lifespan, but you also have to remember that early manufacturing techniques mean a lot more "bad batch" variability.
When part performance is more unpredictable, ease of maintenance is a much bigger deal for everybody all across the price/quality range.
integrated circuits were only invented a couple years before the Apollo moon missions started in 1961
imagine an alternate timeline where that didn't happen, could moon landing have even happened with vacuum-tubes? highly unlikely
Not sure if you've seen this video about how the Apollo computers were made, but it's really fascinating.
https://www.youtube.com/watch?v=ndvmFlg1WmE
(They also anachronistically refer to women as girls which feels weird nowadays.)
I think it would have been more likely to be some kind of mechanical analog computer full of gears and cams (which some planes and missiles did use for guidance). Probably needing more support from the ground.
Apollo did carry tubes also, in the radio transmitters.
In between IC's and vacuum tubes, isn't there discrete transistor logic?