The Thyratron and Using a 2D21 Thyratron to Move Some Current
ฝัง
- เผยแพร่เมื่อ 20 ก.ย. 2020
- In this episode, we take a look at a pretty unique little tube, the Thyratron. It’s surprisingly difficult to drive power hungry loads from vacuum tubes, like a halogen bulb, but Thyratron’s can move a lot more current due to their unique construction.
Thanks for watching!
Here’s an excellent read on Thyratrons:
ir.library.oregonstate.edu/co... - วิทยาศาสตร์และเทคโนโลยี
![CAMPปลิ้น | EP.76[2/2] เดินมาสิบกว่าป่า วันนี้รู้แล้วว่าโจโฉกลัวอะไร!](http://i.ytimg.com/vi/R7Yor-t2Au0/mqdefault.jpg)
What thyratrons are _really_ good at, is moving huge amounts of current in very short times. They are used as pulse formers for currents of hundreds or thousands of amperes. They are also very good at making very short pulses. If you have a capacitor charged at a very high voltage, a thyratron can get all of the charge out of it very quickly. They are still, to this day, better at doing that than semiconductor devices like BJTs, FETs, thyristors and triacs. And they do that in a very reliable and reproducible manner. A special version of thyratron called krytron, is used in the ignition circuit of nuclear bombs.
Excellent video! Thyratron tubes were actually used extensively in industrial applications as rectifiers, and in speed controls for DC motors. I have a Monarch 10ee metal lathe from the late 1950s that uses a Thyratron variable speed DC drive. It uses four thyratrons in total: two large C16J argon thyratrons and two smaller (but still big) 3C23 mercury vapor thyratrons. The mercury thyratrons glow a brilliant blue when operating, and the larger argon tubes glow, but much dimmer.
The operation of the drive is very interesting. There are two sets of thyratrons to form two full-wave rectifiers. The larger C16J tubes control the voltage on the armature of the DC motor, and the smaller 3C23 tubes control the field. The speed control consists of two wire-wound rheostats which control the two rectifiers. Each set of thyratrons has a capacitor connected to the grid. When the thyratron fires (once each cycle of the input AC), the grid current causes the capacitor to charge a bit, eventually charging it enough to create a sufficient bias on the grid to prevent the tube from firing. This stops the operation of the drive.
The speed control rheostat is connected to a control tube that discharges the capacitor at a controlled rate. When the capacitor discharges enough to allow the tube to fire again, the tube fires until the grid current charges the capacitor again. Based upon how fast the tube discharges the capacitor, the circuit will begin to oscillate, eventually getting to the point where the discharge rate of the capacitor exceeds the grid current, causing the tube to fire on every cycle.
When the armature is at full voltage, the second half of the speed control does the opposite: it ramps down the voltage applied to the field of the DC motor, allowing it to spin faster.
The whole system is basically a gigantic tube-controlled PWM speed control. It was particularly advanced for the time as it also included circuitry to detect load on the lathe and automatically adjust the voltage on the motor to keep it running at a constant speed.
I have thought about eventually making a few videos outlining the operation of the machine, and showing how to diagnose it, as most people with the machine are stumped by the drive. Unfortunately, the common remedy is to rip it out and shove in a generic modern drive that does not come close to the performance of the original.
Thank you so much!
That's awesome that you have a lathe that actually uses thyratrons. What a brilliant way to build it too! One day I'd love to build essentially a switching power supply like that, and the capacitor/rheostat method you described is pretty brilliant. I absolutely would love to see some pictures or video of the lathe in action (I love old metalworking tools in any case, so it's like a win-win, haha).
Those old thyratron supplies were surprisingly advanced and really, really smooth. Ken Schiriff has an excellent write up on a switching thyratron power supply for an old Teletype on his website here: www.righto.com/2018/09/glowing-mercury-thyratrons-inside-1940s.html
He even compares the ripple to that of a modern Apple SMPS and it's actually surprisingly close in performance, despite there being 80 years between them!
Thank you for the awesome writeup! I'll do some Googling and see if I can find some schematics for that old metal lathe and maybe I can recreate the power supply on a more compact scale!
@@UsagiElectric I have a full set of schematics for the machine, I'll scan them and send them your way if you wish. It took me about a month of research and reverse engineering to figure out how the circuit works with very little data online other than an old patent. (then a week after I had figured it out I was flipping through an old book from 1960 about industrial control systems I had on my shelf, and it had a whole chapter on the exact circuit. That was a facepalm moment)
Something I didn't mention about the system before is how the thyratrons actually power the motor. The motor is wired across the secondary of a large 1:1 transformer with a center tapped primary. The center tap is connected to 240 volts, with the anode of one thyratron on each side. By alternating the thyratrons on the positive and negative portions of the mains AC input, it creates pulses of DC on the secondary of the transformer.
Here's the schematic that is printed on the cover of the electronic compartment:
www.practicalmachinist.com/vb/attachments/f10/95965d1389024864-10-ee-tube-drive-relay-small_wiad.jpg
It's rather cramped, and is laid out with page space in mind rather than readability. I have a proper service schematic in the manual that I'll take a pic of tomorrow. (additional funny fact: This version of the Monarch 10EE drive was referred to as the "WIAD", for Works In A Drawer, due to the fact that the whole electronic compartment quite literally slides out on drawer slides for serviceability. A nice feature, and a memorable name)
One of the odd quirks about that particular drive system is that it doesn't put out true DC, more like 'pulsed' DC. It has no smoothing caps, as caps big enough to smooth out the power going to the 5 HP motor would be absurdly large.
I can also take some short video of the machine. Here's some links to images I uploaded to imgur of the machine:
i.imgur.com/9KXXeO5.jpg
i.imgur.com/JpfuHhv.jpg
Those are old pics from when I first got the machine, I've since electrically restored it and cleaned it all up. I'll take some more pics tomorrow of the whole thing. It's hard to tell how big those tubes are just based upon that picture, but the large ones are over 10 inches tall.
@@davidk8893 You weren’t kidding, that’s one of the mot confusing schematics I’ve seen in a long time, haha. It’s wild that it doesn’t smooth the rectified AC, but like you said, the capacitors that would be needed for that would be intense.
Confusing as it may be, that schematic is a treasure trove of wild ideas! The center tapped transformer that goes to the plates also splits and goes through another smaller transformer that operates the grids. What a wild setup! It’s crazy how nowadays, in the interest of making things compact and lightweight, there’s generally only one, small transformer, but in this schematic it looks like there is six transformers. That’s insane, but a lathe already weighs a ton, no need to skimp on the power supply.
In your photo, it’s so deceiving because I know those mercury filled thyratrons on the bottom are quite large, but they look dwarfed next to the National Electric tubes above them!
Thank you so much for the pictures and schematic, that is absolutely awesome!
They were also used as timebases in early computers, radiation-counters and televisions and oscilloscopes. They're some of the most diverse tubes you'll find. Especially the later, high-speed tubes like the EC50 🙂
Thank you for the video! I did not realize thyratrons could be operated at such low voltages. I have never had these to play with.
Thanks for checking the video out!
I actually ordered a bunch of these 2D21 Thyratrons off eBay simply because I was curious about them! I was quite pleased with how well they work at just 24V. Now I have a bunch of them and need to come up with some projects to build with them, haha.
I first heard about thyratrons this afternoon (some $375K turntable [yes, apparently people buy such things] uses one in its power supply). Now, thanks to your video I feel I know a lot about them. Thank you, again!
Perfect explanation! Thanks!
Thank you so much!
Thyratrons were used in spot welding controllers for the car manufacturing industry. Difficult to fault find because the circuits were high impedance.
From what I read Thyratrons were used in many applications, including relay controllers, early computers, tool motor controllers, switches, stage light dimming, fluorescent dimmer ballasts, stepper controls, jukeboxes, lasers, tesla coils, radar systems, industrial equipment, welders, elevators, and other applications back in the day.
Thyratrons had a ton of applications! IBM in particular used them to drive the relays for the punch card machine, since they pulled just a bit too much current! Also, CuriousMarc has an excellent video talking about a power supply for an old Teletype that uses two Mercury Vapor Thyratrons as a kind of switch mode power supply to control the voltage precisely. It's an awesome watch: th-cam.com/video/WX74GoHuwHk/w-d-xo.html
Great video! I was experimenting a while back with using a vacuum tube to drive a high current load. I found that even a small 6AU6 can drive a 24 volt relay coil. You could potentially use this property to reset the latched thyratron tube.
Thank you!
Modern miniature relays are so incredibly efficient, it's amazing how little power they actually require to operate. IBM actually used the 2D21 Thyratron to drive the relays for the card reader and punch machines for the IBM 604, 650, etc. Of course, those were much beefier relays!
Using a 6AU6 to trigger a relay to reset the Thyratron definitely could work, although, even a miniature relay can switch a lot more current than a Thyratron can, so it would make more sense to just use the relay to drive the heavy current.
Where the Thyratron will really excel though is switching speed! The Omron G5V1 type relay has an engagement speed of about 5ms, which equates to about 200 Hz switching speed I believe. The 2D21 Thyratron is limited by its de-ionization time, which can be up to 75 us (the ionization time 0.5 us), which equates to about 13 kHz!
@@UsagiElectric Or connect the plate of the 2D21 to the plate of a triode, so that when the triode grid goes positive, the plate voltage is pulled down towards ground, and thus pulls the plate of the 2D21 closer to ground.
A Thyratron can be used to easily make a voltage-controlled sawtooth oscillator.
Thyratrons are really great for all sorts of things! I've experimented with a sawtooth oscillator a couple time and it's amazing how simple it is to make! I think my favorite use for thyratrons is as a very early switch mode power supply!
Brilliant. You should have used that 16vac transformer when demonstrating the Diode valves in part 1 of your series.
Thank you!
I absolutely should have done that, unfortunately, I didn't salvage that transformer until quite a bit after that episode, haha.
Would the regular 6AU6 tube manage to power the bulb if it were powered with it's regular rated voltage of at least 100V or so? Also I did not know 47 bulbs were halogen. Is that all of them or is the 47 denomination just the base identifier? love your videos!
nice differential shirt🚗
Rocking my OS Giken shirt for this one! Their twin-cam head for the L Series is an absolute masterpiece: th-cam.com/video/WoaT-WcOTEg/w-d-xo.html
nice; i too am fascinated at these "parents" of the "transistor"! (Btw for fun did u catch the hollywood movie with the ENTIRE plot centered on the "krytron"? guys nite to watch?)
This is the nice looking version of a Thyristor ! Makes me want to latch ON/OFF all my circuits with it. I was expecting that we would see the ionized xenon. Anode is masking the glow ?
Thanks! Interestingly, the term Thyristor is actually a portmanteau of "Thyratron" and "Transistor"!
I was also hoping to see the ionized xenon, but I think you're right, the Anode/Plate completely encapsulates the grids and cathode, blocking the view of the cool colors. Someday, I'll have to get my hands on some mercury vapor thyratrons and really get the glow going!
@@UsagiElectric Or light the tubes in a dark room
this is LIT, you're awesome man
if you connected the control grid to the positive, would that stop the ionization and stop the current flow?
@@mazengomaa27 Thank you!
Unfortunately, once the gas is ionized, the actual grid wires get encased in positive ions. The effect is that no matter what charge we put onto the grid, it doesn't have any affect on the ionized gas. In rectification service, this isn't a problem since the incoming AC waveform goes negative for half the cycles, that's like essentially cutting the flow of current, which causes the gas to de-ionize and the tube to essentially reset.
@@UsagiElectric
Wow, that's awesome, thanks for the very good explanation
Keep it up, you're damn good 😅
@@mazengomaa27 Thank you so much! I've got an ever growing Excel sheet with episode ideas that I'm working through, so lots of content still to come!
@@UsagiElectric in the wait, all support man, you're very good 👌👌
Thank you for this interesting video. I'd like to build an oscillator circuit with the 2D21=PL21, that generates tilting oscillations as are used for sound generation in the so called 'Trautonium'. Do you have an idea how to generate tilting oscillations with the thyratron?
Thank you sir.
Can thyratron be used in 12 kv circuits n so on?
It's like a tube version of a Thyristor?
Exactly!
Interestingly, the term "thyristor" is actually a portmanteau of the words "thyratron" and "transistor"!
What would happen if the heater was disconnected from the thyratron while turned on? Would the ionisation continue unaffected, or would the voltage drop change due to temperature effects on ionisation? Something to consider for science.
Sure you learn a lot about it. However you could use miniature incadescents like those used in Christmas lights. These are 3v and their current is low.
Those little thyratrons have been obsolete since SCRs were invented. But I've used larger hydrogen thyratrons as switches in radar transmitters. A 7890 will switch hundreds of amps at 40 KV. I've also used ignitrons to fire somewhat more power but they can be jittery. Plus they contain maybe a pound of mercury each. Small thyratrons (from EG&G) were also used as triggers for nuclear bombs.
I'm always amazed at just how much voltage and current certain vacuum tubes can handle. I had to look up that 7890 - what a phenomenal tube! It's hard to get a sense of sale from pictures, but the heater alone requires 30 amps! From the datasheet, it looks like it's about 4 1/2 inches in diameter and foot tall! But, given the voltages and amperage (up to 40kV and 2,400A!!) that it can handle, that's a surprisingly compact package.
High voltage tends to frighten me, but there's just something amazing about tubes at really, really high voltages. I can't imagine working around big thyratrons and ignitrons, I bet you have some awesome stories!
@@UsagiElectric Yeah, I worked in a building that housed a 390KV @ 1 amp power supply. We tested BMEWS radar klystrons, among other things. One of these days I'll come up with a show & tell about my HV work. There are still applications for such gadgets.
Could you please explain the use of shield grid?
I'll try my best, though my understanding of the shield grid is also a little spotty.
The control grid in a thyratron, particularly this 2D21, is not shaped like conventional grids. It's more like a single chunk with a hole in the center of it. The control characteristic of the tube can be controlled by the size of the hole in the grid (since the plasma must flow through the hole) as well as the size of the grid itself. The larger the grid becomes, the lower the input impedance and the higher the starting grid current. To reduce these effects, the shield grid is added.
The shield grid is held at a constant potential and "shields" the control grid from the anode and cathode. The shield grid now absorbs most of the starting current, allowing the grid to be made smaller, but the control characteristics can now be controlled by changing the potential difference between the shield and cathode. The result is a low impedance control grid input and control characteristics that can be tailored to a specific purpose.
The shield grid also serves one more purpose, which is to help with cleanup of positive ions after the tube has been de-enegized. This allows deionization time to be decreased and lets the tube operate at higher frequencies.
I hope that helped!
@@UsagiElectric Thank you ❤️
Are those bulbs actually halogen or are they just ordinary incandescents?
15:37: "We need a manual button, and I haven't come up with a good way to get around that issue." How about another thyratron? 😛
PS: It's thyratrons all the way down.
5:19: "The control grid gets kind of encased in positive ions."
From the gas.
Here's an interesting little bit of trivia for you, the word Thyratron comes from the Greek word Thyros, which means Door, so, a Thyratron is literally a Door Tube, you could think of a Thyratron as being like a voltage-controlled switch.
That's cool, I had no idea it had its roots in Greek!
Another interesting bit of trivia is that the word Thyristor is a portmanteau of Thyratron and Transistor!
15:33 what about relays?
You can use the 6V DC as his own Heater Voltage :O)_)
Haha, got a bit of a Catch 22 going on there. In order to generate the 6V, the thyratron needs the heater to be nice and warm, but the thyratron won't conduct until that heater is hot.
Can you see the glow from the tube though?
Unfortunately, the 2D21 is a very self contained tube so you can't actually see any glow from the plasma. It's such a shame too, I was hoping for a nice warm glow like the 5651 regulator tube or something.
Someday I'll get my hands on an old Mercury thyratron and bask in the awesome blue glow!
7:19: "It's essentially like a switch."
You're definition of fuse is quite narrow. If you check any good dictionary what fuse means you realise that resetting them is not always easy. For example I've workend with a lathe that had a wooden gear as the fuse on the transmition line. If the lathe jammed the wooden gear would break and the electric motor can ran free and not burn.
also deuterium
Put you button on the controlgrid, the you should be able to control the bulb.
Thanks for checking the video out!
I'm not sure I follow your comment though. The button and the potentiometer are directly controlling the control grid in my example. Also, Thyratrons fundamentally work using an arc of plasma created within the tube. By nature, this arc of plasma will send as many electrons as it can from the cathode to the plate. That's why we end up with very switch like behavior out of the tube, as opposed to the more analog behavior we see in standard vacuum tubes.
@@UsagiElectric If you would take the button setup from the middle sample to the last. Would the the thyratron turn off when the AC go into the nagative state?
@@nielserikmeier9721 I think I'm following you. For example, on the third circuit where I rectify AC, if instead of having the grid tied to the plate via a resistor, what if I tied it to positive with the button? In that case, the Thyratron would indeed turn off whenever the AC went negative, just like you're thinking. It wouldn't turn back on again until the button was pressed. However, since the AC from the transformer has the same frequency as mains AC, 60Hz. This means that as soon as I release the button, the Thyratron will stop conducting as soon as the AC signal falls below 0V again, which would happen essentially immediately.
At first, I was trying to wrap my head around how that setup could be used, but it's actually perfect for something like a binary display. Four halogen bulbs, four Thyratrons, and you've got yourself a four-bit display that can be controlled from other tubes. The only caveat is that you need a transformer outputting AC into the plate for it to work, but that's something that can be worked around.
Thanks for the suggestion, that's awesome!
Usagi Electric if you tie this setup to you RS flip flop ... 👍
@@nielserikmeier9721 I think you're on to something here!
Never saw a #47 bulb filled with Halogen gas. Who knew?
Actually, you're definitely right in that the #47 bulb is actually just a standard incandescent bulb and not a halogen. I was just mixing up by terminology here and totally didn't pick up on it in the edit, haha.
well, there's another way to make a single bit of memory using a tube, thyratron sram...
all sram is, is just bunch of buffered S/R latches
15:35 well you could if the tube drives an igbt or mosfet, technically its not cheating, functionally fets, igbt, sicfet, moset.... whatever, its functionally equivalent to a triode and thyratron in one package and remove filament pins, so just anode, grid , and cathode, just make sure gate voltage for dat mosfet is 24v, not 5v or 12v, tube might punch a hole through dat wafer
how about a latch with a thyrathron
edit: oh well we thought the same thing i guess xD
Looks like we were on the same page, it just took me a little longer to get there, haha!
it acts like a thyristor
Exactly!
Interestingly, the term "thyristor" is actually a portmanteau of the words "thyratron" and "transistor"!
@@UsagiElectric I worked with high voltage ceramic thyratrons that would stitch 30 kV and KA. Made of ceramic. They use a ballast heater to control the internal pressure.
one bit memory
Yup, it makes a nice little SR Flip Flop. Though, I did a lot of testing with trying to get one of these to work as a memory cell for the vacuum tube computer, and I had all sorts of trouble. I ended up not going with them, but there are a lot of really interesting ways that these could be used with some proper designing!
@@UsagiElectric Pull the plate down with a triode!