It's so... fascinating to see him tearing apart these old devices and bringing them back to life. The amount of patience and experience required is out of this world !
Ah - regenerative dividers. I recognize that block diagram from the old 5105A + 5110A synthesizer combo. I bent my mind around that way back when this was a brandy new product in the early/mid 1960s. (1964?) The concept almost made it aboard GPS satellites but people were worried about the regeneration starting up correctly. So I broke the feedback and gave GPS a dither error that was proportional to its output frequency. At the accuracies involved here it is noticeable, in the lab if not on the signal from orbit. And it may be different on newer satellites. I worked on Phase IIB. {^_-}
It amazes me the things that people managed to build back in the 1950/60's with a mix of electro and mechanical components. Not only did it work as needed, we can still (well Marc anyway) repair and restore them to their original state.
13:39 I'm not sure whether you'll like this, Marc, but I (we) enjoy when your repairs are never a piece of cake 😆The harder the better, the more nail bitey. 😬
Old electronics that put up a fight makes for awesome content so i'm not complaining, though i imagine it's not the most fun for you Marc. Can't wait for the next video.
I loved seeing the synchronous motor, and hearing the sound. I’m reminded of the synchronous motor in my Hammond Organ, which was invented after Hammond heard the 60hz sound of a clock. Imagine my joy on learning that this particular device might have provided the 60hz reference to a power plant! Maybe even the one that powered the neighborhood where Green Onions was recorded, and thus the machine that kept that famous organ in tune.
Very awesome, super excited to follow along on this one! I have two predecessors to this fine instrument, the analog dial clocks. One 113BR, stock and working as expected. But also a prototype narrowly saved from demise, marked 113X. This also follows no schematics, so your journey here may prove to be extremely helpful! Story goes that it was donated to an IL College by HP (the name of which escapes me). From there I'm not entirely sure what use it served. I then acquired it about 5 years ago at a hamfest, just before it was to be parted out and lost forever. I just couldnt let that happen... Thank you again for letting us follow along with you on these projects! The time and effort you put into them are not in vain!
Thanks to your video, I recognized a brief appearance of an HP115BR in today's PBS Nova program called "Decoding the Universe:Quantum". I love your videos.
Indeed! This episode: www.pbs.org/wgbh/nova/video/decoding-the-universe-quantum/ . The HP 115BR appears briefly at 17:35, followed by the HP 5061A Caesium clock equipped with the later integrated Patek Phillipe divider and clock. Awesome episode by the way, they are right up there with Veritasium. Or is it the other way around?
Thank you sir for doing what you do and making it available to us! These kinds of electronics are the reason I got into the field, and your content genuinely makes me emotional!
As always Marc you bring to light these hidden gems of HP imaginative engineering. Always a good watch your videos are. I learn so much! And I hope my meeting you in person at VCF-West didn’t disturb you. You seem to have been not yourself that day. But I really enjoyed Mike’s presentation of the DSKY display repair. Thanks for appearing.
I will say, at least you don't have to first fix the oscilloscope or pen plotter that will help you fix the thing you really wanted to get running. It's encouraging!
If I understood even 1% of this I'm sure I'd get more out of it. But ... nonetheless I still truly enjoy this channel as much as any. Maybe you guys could do a colab with Mr. Carlson. His recent barn find is going to be a real barn burner (😉, .c 1930s frequency generator ... I think?). Anyway, great stuff Marc and keep it coming! Thanks
Interesting how much corrosion is on that synchronous motor considering its sealed and the rest looks fine. I really enjoy your explanations and your ability to reverse engineer nightmares.
Yeah. Although i rather suspect that there was one engineer in the instruments department who insisted on putting frequency mixers into absolutely every device 🙂
Great video, thanks! I have the dark gray version of the straight 115BR, former Air Force property. Haven't gone through it yet, still need to finish repairing my HP 105, which had a major power supply failure before I got it. I also have the HP K22-5321B Nixie clock to run off the HP 105 once it is repaired!
This has got to be one of the luckiest and unluckiest finds at the same time Can't wait for the upcoming episodes (although I'm not as on edge since you teased a happy ending :D)
One crazy box, what, you have to open it to start it.... Also surprised to see the HP105A, as I have a 105B (backup battery), as the video's usually feature gear I'll never come across.
A tightly sealed device, but you have to open it to turn on the ignition and crank it to start ;) I suppose you weren't supposed to ever switch it off?
That’s correct. Meant to be run on 24V with battery backup. If the reference 100kHz disappears, the divider stops, the clock stops, and won’t restart on its own so it does not indicate the wrong time. You then have to restart and resync it manually.
@@CuriousMarc Yes, another reason to have regen dividers: If there is an interruption, it stops, which makes sure that you KNOW that there was an interruption. Similar with the mechanical time-of-day display: If there is a fault/interruption, the user can see WHEN it happened. If the time display was electronic (i.e. Harwin displays, Nixies, whatever), then the time-of-fault would be lost.
Thanks for the video Marc, I love the Apollo vids, but a little palate cleanser now and again is greatly appreciated. PS. I sure hope those carbon comps are just used for setting transistor bias points and not for frequency setting.
How exciting. You think you have an amazingly cool new toy, but dig deeper and you find cool new toy super special edition. Curious Christmas came early this year.
Transporting an atomic clock around the world to synchronize local clocks is just amazing. We have much better ways now, but this sounds much more like science fiction than today's methods. The Future is boring :D
I would not be surprised if the syncho-resolver was specially selected for the 10kHz signal. The resolver coils are being operated at much higher frequencies and thus the parasitic inductances play around with the three port transfer characteristics of the resolver. But they are manufacturing at a test and measurement company.
I wonder if the early 1970's HP computer/control stuff like the HP9825 could use the precise 1kHz/60Hz signals as a time reference (for timestamping measurement data etc). This would especially make sense if you have to compare data (or syncronize actions) from systems at different locations.
As I understand it, there are three maneuvers needed every time the clock is powered on: kickstarting the regenerative dividers, cranking the synchronous motor and, of course, setting the time. But all of these require opening the (hermetically sealed) casing?
what a magnificent video :-)))) thanks for your hard work! Please add a description of the helper clocks for the 2 mixers that are activated internally by hand, i didn't see or recognize them in the original block diagram or wire plan.
Aha! That’s because they cleverly re-use the 90 kHz multiplier as the start oscillator. So it’s sort of hidden in plain sight… I showed it as a separate thing for explanation clarity.
Some ideas in that device seem to have been reused in the 203A "function" generator (which is more of an HP 200 on steroids than what we would call a function generator nowadays functionally, and with a very different internal working principle, also with resonant dividers and a resolver involved....).
@@t.a.7970 I won't quite deny that, though it is more a moral position than one based on the technology. HP printers are still pretty good machines, regardless of what one might think of the company. But my point is that "HP" != "HP". They aren't the same company. HP hasn't "fallen" from the instrument maker shown here. That is still doing just fine, it just had to change its name because a usurper stole the old name. It's kinda like saying that the fine old family name of Kruger has "fallen" just because they had a degenerate offspring named Freddie. No, the Krugers in general are still fine morally upstanding people. They just had a very famous offspring that should have been dumped down the well shortly after birth.
I have been vexed several times repairing old HP gear because HP loved to have their own part numbers for semiconductors. None of that JDEC 2N… foolishness for HP! Anyway the HP engineers liked to make their designs rather dependent on the specific characteristics of the semiconductors. If you don’t sub in the right transistor, the circuit card just sits there as if mocking “Is that the best you’ve got?”
But do you know that I have now scanned the fiches for the HP semiconductors and uploaded them to the internet archive? The ones for transistors are here: archive.org/details/c000_r006 . It gives the parameters for the HP only transistors, which usually are just selected versions of 2Nxxxx appearing a few lines above or below. Or replace the with something better.
wow... it physically implements the trig identity, sin(x+P) = sin(x)cos(P)+cos(x)sin(P). For a fixed frequency I guess, since a capacitor is used for the phase shift to make cos(x).
Wasn’t this stack of gear used in the Hafele-Keating experiment which provided some of the earliest empirical evidence of special and general relativity?
Almost. The Hafele-Keating used a later divider clock that allowed them to get into the nanoseconds, with the famous Patek Philippe clock face. It’s very rare. We are still looking for that clock…
Given the way this would be used it isn’t something you would ever see needing to triggered after starting the whole system. This sort of thing you’d start once and keep powered for months or years at a time. Same reason the knurled knob to get the synchronous motor running is only accessible when the box is open.
@ If it’s potentially only months or a few years, I don’t think it’s a good idea. But I might be too steeped in the idea that everything must be remotely controllable, per the current paradigm.
My guess is because the switch, like the motor, is something that could completely mess up the timing if it was accidentally pressed. This is the kind of device that you set up once and then never touch again, so it makes sense that it is locked away
It’s on purpose. It’s to prevent to accidentally restart the clock once it has lost lock. If the reference 100 kHz disappears even for a split second, the divider stops and the clock stops, instead of restarting and indicating the wrong time. You then have to open it up, reset it and resync it, which is an hour long procedure anyhow.
The high quality photo of the flying clock is actually not the original, but was put together by hpmemory.org. The original had a prototype Caesium and a prototype modified clock. However, triple kudos to hpmemory for using the correct model of clock in their setup!
How does the regenerative divider stay stable? For instance say the starter had a slight error and was at 90.1KHz, then the mixer produces the low signal of 9.9KHz, which gets tripled twice to 89.1 KHz, with a 9 times larger error than at the start. If it was working in reverse that would make more sense to me; using two 3x dividers to get the 10KHz and a mixer to generate the 90KHz signal which then gets divided again.
That is an excellent point. It’s not exactly implemented as I drew here. There is no separate 90 kHz start oscillator. The 90 kHz tripler circuit itself is used. The tripler is in essence a very resonant amplifier, almost an oscillator, so the start pushbutton just enables a slight additional feedback that brings it to oscillation. When the push button is released, the oscillator will not jump instantly from 90.1 to 89.1 kHz, it will start sliding towards 90 kHz and reach equilibrium. That said, it must be an interesting design challenge to make the triplers resonant enough to make this stable, but not too resonant that they self oscillate.
@@CuriousMarc Thanks! I was thinking about it more and had wondered if it might have had something to do with some sort of resonance/feedback in the tiplers that the diagram was glossing over, it sounds like it is something to that effect. I had realized that the frequency would have to slide through the correct value but figured that there had to be some sort of delay across the triplers which could cause it to overshoot the correct value so that it would actually end up oscillating around the 90kHz just a little in an amount of time on the scale of that delay. Clearly the people who designed it did something clever that I'm not quite wrapping my head around. Or maybe that delay time is just so small that it is would just change the wave shape a bit in a way which would probably get smoothed out by other components.
Same atomic clock but different divider. They did that later with a better all electronic divider specially made for the atomic clock, allowing them to get down to nanoseconds instead of microseconds.
In this crazy time, nothing calms like some CuriousMarc.
Isn’t that the truth.
The Bob Ross of electronics!
It's so... fascinating to see him tearing apart these old devices and bringing them back to life. The amount of patience and experience required is out of this world !
@ Mr Marc has the gift to explain very complex things in away that they are sound simple.
It's a crazy time, yes, but with nanosecond accuracy !
By this point, I feel there’s no greater human archive than that of “The Doodly-Doo”.
Every time I see one of these videos with any HP equipment I just have to marvel at how brilliant the HP engineers were!
Ah - regenerative dividers. I recognize that block diagram from the old 5105A + 5110A synthesizer combo. I bent my mind around that way back when this was a brandy new product in the early/mid 1960s. (1964?) The concept almost made it aboard GPS satellites but people were worried about the regeneration starting up correctly. So I broke the feedback and gave GPS a dither error that was proportional to its output frequency. At the accuracies involved here it is noticeable, in the lab if not on the signal from orbit. And it may be different on newer satellites. I worked on Phase IIB.
{^_-}
It amazes me the things that people managed to build back in the 1950/60's with a mix of electro and mechanical components. Not only did it work as needed, we can still (well Marc anyway) repair and restore them to their original state.
I love the smell of old electronics in the morning. Smells like, well, old electronics.
The future of my country is being decided today. And I would much rather watch Marc and his friends.
precisely my friend
Condolences.
I hear Canada, Cuba and Mexico are all fast tracking the construction of walls.
We did it 🥲
@DrinksInHighPlaces Tyranny for all!
@@richardkaz2336yeah it’s not like there’s an absurd level of emigration from those countries to the US or anything… You are actually mentally ill
13:39 I'm not sure whether you'll like this, Marc, but I (we) enjoy when your repairs are never a piece of cake 😆The harder the better, the more nail bitey. 😬
"It shouldn't be too hard" the famous last words :) Amazing machinery with special history. Looking forward for future episodes.
Old electronics that put up a fight makes for awesome content so i'm not complaining, though i imagine it's not the most fun for you Marc.
Can't wait for the next video.
Not to worry, it’s more fun for me too when it resists a little bit. That’s when you learn more.
Watching Marc getting curious while I have breakfast, bliss.
The only sad thing about this episode is having to wait to see the next one!
This is one of my favorite channels!
Top notch again, Marc! These restoration videos should be standard material in every electronics engineering class.
I loved seeing the synchronous motor, and hearing the sound. I’m reminded of the synchronous motor in my Hammond Organ, which was invented after Hammond heard the 60hz sound of a clock.
Imagine my joy on learning that this particular device might have provided the 60hz reference to a power plant! Maybe even the one that powered the neighborhood where Green Onions was recorded, and thus the machine that kept that famous organ in tune.
I watch this channel regardless of my understanding of electronics. Always mesmerizing.
Very awesome, super excited to follow along on this one! I have two predecessors to this fine instrument, the analog dial clocks. One 113BR, stock and working as expected. But also a prototype narrowly saved from demise, marked 113X. This also follows no schematics, so your journey here may prove to be extremely helpful! Story goes that it was donated to an IL College by HP (the name of which escapes me). From there I'm not entirely sure what use it served. I then acquired it about 5 years ago at a hamfest, just before it was to be parted out and lost forever. I just couldnt let that happen... Thank you again for letting us follow along with you on these projects! The time and effort you put into them are not in vain!
Ongoing outstandingness. This channel is the best! 👍👍
Thanks to your video, I recognized a brief appearance of an HP115BR in today's PBS Nova program called "Decoding the Universe:Quantum". I love your videos.
No kidding! I must watch that now.
Indeed! This episode: www.pbs.org/wgbh/nova/video/decoding-the-universe-quantum/ . The HP 115BR appears briefly at 17:35, followed by the HP 5061A Caesium clock equipped with the later integrated Patek Phillipe divider and clock. Awesome episode by the way, they are right up there with Veritasium. Or is it the other way around?
4:49 I love it when you guys describe the smell of things! Cause we’ll never be able to smell it ourselves.
Yess!!!!! So today is not such a bad day after all! Thank you Marc
Thank you sir for doing what you do and making it available to us! These kinds of electronics are the reason I got into the field, and your content genuinely makes me emotional!
Very Very Interesting! Ill be on the lookout for the next episode!
As always Marc you bring to light these hidden gems of HP imaginative engineering. Always a good watch your videos are. I learn so much! And I hope my meeting you in person at VCF-West didn’t disturb you. You seem to have been not yourself that day. But I really enjoyed Mike’s presentation of the DSKY display repair. Thanks for appearing.
I'm really looking forward to this series. Thanks!
I will say, at least you don't have to first fix the oscilloscope or pen plotter that will help you fix the thing you really wanted to get running. It's encouraging!
finally some hp stuff again!!! thats why i watch! hp is the best!!!
@marc ! You have to warn us before spinning up time in the wrong direction! @6:40
The aircraft at 3:24 appears to be a Douglas DC-7, which is powered by four (piston engine) Wright 3350 turbo-compounds, not turbo props.
Thanks for the correction. That still flew with pistons? Thanks goodness they didn’t use regenerative dividers in the engines ;-)
Excellent, as always. I wish my physics classes had used elevator music explanations.
If I understood even 1% of this I'm sure I'd get more out of it. But ... nonetheless I still truly enjoy this channel as much as any. Maybe you guys could do a colab with Mr. Carlson. His recent barn find is going to be a real barn burner (😉, .c 1930s frequency generator ... I think?). Anyway, great stuff Marc and keep it coming! Thanks
“how do you call this” - the most french french i’ve ever heard marc say!!
Mais oui mais oui!
I would love some HP gear like that just to have standards quality devices at home
I love these videos and vintage HP gear… thanks for posting it!!
"That shouldn't be too hard" is the delusion we use to mask any doubts in our abilities in order to achieve extraordinary things. It's a good thing :)
What a wild little unit, there's a lot of history to it's mods
great work
I like the way you first introduce RF wizardy to get things done analog way, then go to simple repair, just to end up with part 2 video 😀
Fantastic!
Can't wait for the next episode!
OHHHHH ALAN KILIAN! HEY ALAN!!!! no friggen way my friend Alan Kilian is on the Curious Marc Channel LOL! wow!
It’s a little bit of fame!
And this device landed in the one place on the planet it needed to be!
Thanks Marc.
Interesting how much corrosion is on that synchronous motor considering its sealed and the rest looks fine. I really enjoy your explanations and your ability to reverse engineer nightmares.
I didn't get a notification for this! I'm subscribed and set to all.
Boo. The Google has failed us again.
How does the frequency tripler work? I find this stuff fascinating.
Just incredible what those HP engineers were capable of those days ...
Yeah. Although i rather suspect that there was one engineer in the instruments department who insisted on putting frequency mixers into absolutely every device 🙂
A problem worthy of attack, proves it's worth by fighting back. 🙂 Love your videos!
Great video, thanks! I have the dark gray version of the straight 115BR, former Air Force property. Haven't gone through it yet, still need to finish repairing my HP 105, which had a major power supply failure before I got it.
I also have the HP K22-5321B Nixie clock to run off the HP 105 once it is repaired!
Nice collection! Good luck on the all the HP repairs. Keep them going!
Beautiful device 😌
I am so looking forward to the not-too-hard stuff.
This has got to be one of the luckiest and unluckiest finds at the same time
Can't wait for the upcoming episodes (although I'm not as on edge since you teased a happy ending :D)
One crazy box, what, you have to open it to start it....
Also surprised to see the HP105A, as I have a 105B (backup battery), as the video's usually feature gear I'll never come across.
@4:04 Maybe you should change this to: "...but as with everything else on this channel, it doesn't work - yet!"
If you've managed to repair this one of a kind equipment, you've earned the right to wear a fancy pant.
It shouldn't be hard to repair...
Bold!
A tightly sealed device, but you have to open it to turn on the ignition and crank it to start ;)
I suppose you weren't supposed to ever switch it off?
That’s correct. Meant to be run on 24V with battery backup. If the reference 100kHz disappears, the divider stops, the clock stops, and won’t restart on its own so it does not indicate the wrong time. You then have to restart and resync it manually.
@@CuriousMarc Yes, another reason to have regen dividers: If there is an interruption, it stops, which makes sure that you KNOW that there was an interruption. Similar with the mechanical time-of-day display: If there is a fault/interruption, the user can see WHEN it happened. If the time display was electronic (i.e. Harwin displays, Nixies, whatever), then the time-of-fault would be lost.
Thanks for the video Marc, I love the Apollo vids, but a little palate cleanser now and again is greatly appreciated. PS. I sure hope those carbon comps are just used for setting transistor bias points and not for frequency setting.
You make it look and feel so easy. Its not :D I would really like to be an apprentice of yours one day.
I think that's fabulous that it's all point-to-point and tag-strip, yet there's an Arduino in there.
The Arduino has been exorcized, then burned as an offering to the Gods of horology. Hopefully they'll forgive us now.
How exciting. You think you have an amazingly cool new toy, but dig deeper and you find cool new toy super special edition. Curious Christmas came early this year.
You probably have the only working combination of that equipment anywhere in the world!
Transporting an atomic clock around the world to synchronize local clocks is just amazing. We have much better ways now, but this sounds much more like science fiction than today's methods.
The Future is boring :D
10:46 please turn it continuously and vary the rotational speed, would love to see that oscilloscope image!
There is no easy in a multi part video :)
But there is awesome tech in all!
Got yourself a rare one there. Very interesting.
I would not be surprised if the syncho-resolver was specially selected for the 10kHz signal. The resolver coils are being operated at much higher frequencies and thus the parasitic inductances play around with the three port transfer characteristics of the resolver. But they are manufacturing at a test and measurement company.
I wonder if the early 1970's HP computer/control stuff like the HP9825 could use the precise 1kHz/60Hz signals as a time reference (for timestamping measurement data etc). This would especially make sense if you have to compare data (or syncronize actions) from systems at different locations.
The atomic clock transition is "optical" as in purely electro-magnetical, no vibration involved. But that is a tiny nit to pick.
As I understand it, there are three maneuvers needed every time the clock is powered on: kickstarting the regenerative dividers, cranking the synchronous motor and, of course, setting the time. But all of these require opening the (hermetically sealed) casing?
✌
what a magnificent video :-)))) thanks for your hard work! Please add a description of the helper clocks for the 2 mixers that are activated internally by hand, i didn't see or recognize them in the original block diagram or wire plan.
Aha! That’s because they cleverly re-use the 90 kHz multiplier as the start oscillator. So it’s sort of hidden in plain sight… I showed it as a separate thing for explanation clarity.
'shouldn't be too hard' famous last words
I hope you get to fly it around the world and test GR! I guess you'd need two of them.
Some ideas in that device seem to have been reused in the 203A "function" generator (which is more of an HP 200 on steroids than what we would call a function generator nowadays functionally, and with a very different internal working principle, also with resonant dividers and a resolver involved....).
for anyone curious, look into "hyperfine transition of cesium atomic frequency standard"
The most amazing thing about this vid is how far HP has fallen.
"HP" as we know it is a spinoff of the printer division from the 1970s. The real HP turned into Agilent and then Keysight.
@@lwilton Regardless of what became what..the HP name is trash now.
@@t.a.7970 I won't quite deny that, though it is more a moral position than one based on the technology. HP printers are still pretty good machines, regardless of what one might think of the company.
But my point is that "HP" != "HP". They aren't the same company. HP hasn't "fallen" from the instrument maker shown here. That is still doing just fine, it just had to change its name because a usurper stole the old name.
It's kinda like saying that the fine old family name of Kruger has "fallen" just because they had a degenerate offspring named Freddie. No, the Krugers in general are still fine morally upstanding people. They just had a very famous offspring that should have been dumped down the well shortly after birth.
I have been vexed several times repairing old HP gear because HP loved to have their own part numbers for semiconductors. None of that JDEC 2N… foolishness for HP! Anyway the HP engineers liked to make their designs rather dependent on the specific characteristics of the semiconductors. If you don’t sub in the right transistor, the circuit card just sits there as if mocking “Is that the best you’ve got?”
But do you know that I have now scanned the fiches for the HP semiconductors and uploaded them to the internet archive? The ones for transistors are here: archive.org/details/c000_r006 . It gives the parameters for the HP only transistors, which usually are just selected versions of 2Nxxxx appearing a few lines above or below. Or replace the with something better.
So, all you need now is a battery pack and a prop plane, and if you configure it right... We can set the clocks back in time.... Right?
wow... it physically implements the trig identity, sin(x+P) = sin(x)cos(P)+cos(x)sin(P). For a fixed frequency I guess, since a capacitor is used for the phase shift to make cos(x).
I think you got it absolutely right, that's how it does it.
I have an HP 107BR that would be a nice match for that unit.
That would be the perfect match! They were meant to go together.
Wasn’t this stack of gear used in the Hafele-Keating experiment which provided some of the earliest empirical evidence of special and general relativity?
Almost. The Hafele-Keating used a later divider clock that allowed them to get into the nanoseconds, with the famous Patek Philippe clock face. It’s very rare. We are still looking for that clock…
Hang on.. Why is the “starter” oscillator switch internal rather than being an external protected switch? That seems inconvenient.
Given the way this would be used it isn’t something you would ever see needing to triggered after starting the whole system. This sort of thing you’d start once and keep powered for months or years at a time. Same reason the knurled knob to get the synchronous motor running is only accessible when the box is open.
@ If it’s potentially only months or a few years, I don’t think it’s a good idea. But I might be too steeped in the idea that everything must be remotely controllable, per the current paradigm.
My guess is because the switch, like the motor, is something that could completely mess up the timing if it was accidentally pressed. This is the kind of device that you set up once and then never touch again, so it makes sense that it is locked away
It’s on purpose. It’s to prevent to accidentally restart the clock once it has lost lock. If the reference 100 kHz disappears even for a split second, the divider stops and the clock stops, instead of restarting and indicating the wrong time. You then have to open it up, reset it and resync it, which is an hour long procedure anyhow.
@@CuriousMarc Yes, and the mechanical display (rather than electronic) tells the user exactly WHEN an interruption occurred, which might be important.
Worth noting that the flying clock was an H20. It's *barely* visible in the photos of the flight case.
The high quality photo of the flying clock is actually not the original, but was put together by hpmemory.org. The original had a prototype Caesium and a prototype modified clock. However, triple kudos to hpmemory for using the correct model of clock in their setup!
Wow.
👏👏👏👏
❤wow❤
Again, another episode that is too cool to be true (but it is).
You mention 190KHz a couple of times right after the 9:00 mark, but I think you meant 90KHz?
Hey that's my time machine i really need it back
How does the regenerative divider stay stable? For instance say the starter had a slight error and was at 90.1KHz, then the mixer produces the low signal of 9.9KHz, which gets tripled twice to 89.1 KHz, with a 9 times larger error than at the start. If it was working in reverse that would make more sense to me; using two 3x dividers to get the 10KHz and a mixer to generate the 90KHz signal which then gets divided again.
That is an excellent point. It’s not exactly implemented as I drew here. There is no separate 90 kHz start oscillator. The 90 kHz tripler circuit itself is used. The tripler is in essence a very resonant amplifier, almost an oscillator, so the start pushbutton just enables a slight additional feedback that brings it to oscillation. When the push button is released, the oscillator will not jump instantly from 90.1 to 89.1 kHz, it will start sliding towards 90 kHz and reach equilibrium. That said, it must be an interesting design challenge to make the triplers resonant enough to make this stable, but not too resonant that they self oscillate.
@@CuriousMarc Thanks! I was thinking about it more and had wondered if it might have had something to do with some sort of resonance/feedback in the tiplers that the diagram was glossing over, it sounds like it is something to that effect.
I had realized that the frequency would have to slide through the correct value but figured that there had to be some sort of delay across the triplers which could cause it to overshoot the correct value so that it would actually end up oscillating around the 90kHz just a little in an amount of time on the scale of that delay.
Clearly the people who designed it did something clever that I'm not quite wrapping my head around. Or maybe that delay time is just so small that it is would just change the wave shape a bit in a way which would probably get smoothed out by other components.
Weren't these used for LORAN before Cesium Clocks were a thing?
Nice ;) can you make a pocket watch version ;)
Is this the clock where they put one in a plane and one on the ground to prove time difference ?
Same atomic clock but different divider. They did that later with a better all electronic divider specially made for the atomic clock, allowing them to get down to nanoseconds instead of microseconds.
"Atomic steam punk clock"
You have a small bug on your modified block schematic: 60KC rather than 60HZ. Or 60CY in the speak of the era. :-)
Oh, indeed! Keen eye!
They missed the opportunity to add a crank
a couple milliseconds off I really need my watch to keep a more precise time than that that is just too big of an error for me 😁
Interestingly, the stamp says 60 Hz, not 60 Cps. Your schematic has the mistake of writing "60 Kc" when it is "Cps" (Kc = kilocycles)
Would you share the arduino code for others to play about with?
It’s just a few lines of code to set whatever synthesizer I used to 1,000 HZ
I can’t remember which one I grabbed that went down to 1,000 HZ. Sorry.
Cut the blue wire with the green stripe... not the green wire with the blue stripe. ... as he looks at it using a yellow glow stick.
3:25 "turboprop" is not accurate
I know, it's pedal powered.