Looking for pointers on Deflection Amplifier Output Stage Design
 

Can somebody point me towards books, websites, or published journal articles that fully explain deflection output stage amplifier design?

I have been reading Grob's "Basic Television" textbook. Bernard Grob was a head engineer and instructor at RCA during their golden era, when RCA was pretty much inventing every technological innovation in television.

The third and fourth editions of "Basic Television" are great as they cover black and white, color, tube, and solid-state TVs. Grob explains in sufficient detail everything about a TV needed to build a custom circuit from scratch... everything except the precise output voltage and current parameters for the output stage of a deflection amplifier.

Grob gets VERY close, describing the details of how to design sync separation, phase detection, oscillator, sawtooth waveform, and he covers the high level design of the power amplifier output stage... but when it comes to the exact formulas to use for the deflection output stage amplifier and impedance matching transformer, specifically the output voltage and current needed to deflect the raster by X-degrees horizontally and Y-degrees vertically... he just doesn't provide the necessary details.

I am looking for enough detail to DIY a custom tube TV circuit from scratch. Similar to the amount of detail available in the DIY tube audio community, where it is clear how much voltage and current the audio amplifier's output stage needs to provide to properly drive a loud speaker with a given specification. So if you have any pointers, even a retired TV engineer that I should try to contact, please help! I want to try to spark a DIY TV community by building a custom TV circuit and publishing the designs open source, similar to the DIY tube audio community.

Unlike tube audio, not many people have walked this path in the internet era. I would suggest primary sources. A lot of great info in the periodicals of the day. Start by digesting everything tv related you can find ih RCA Review, Electronics, Radio and TV News, etc. You can find much info at www.worldradiohistory.com, but it will be basic research - no spoon feeding here. The info is there but finding it is another question! If this project bears fruit I'm quite interested to learn about it!

At least these days it is online and free! This kind of info used to live only in corners of technical libraries.

Also keep in mind that no "tv cookbook" exists, because tvs were too complex to DIY, and commercial products contained trade secrets. Most educational materials were aimed at service men not designers
Speaking of which Google's patent search may also be useful.

Hi to all,
Hi LukeSimon,

Here's a similar endeavour by two French guys, Frederic A. & Pierre G. on their web site.

A Miniature 2" CRT Monitor using a Scope tube (Green phosphor).
- All tube design
- Every important section (Scanning, Video Amplication, Sync, Power supply) is studied in fine detail with an analysis of the Pro's & Con's of each design. The study is divided into 19 chapters for each sub-assembly.
- Loads of Scope photos, construction snapshots, voltage/current measurements.

http://www.cfp-radio.com/

scroll down to "Television"
then "Etude autour d'une TV-moniteur miniature"

Link via Google translate Fr to English :

https://www-cfp--radio-com.translate..._x_tr_sch=http

Sample off-screen photo of the finished monitor (see below)

Best Regards
jhalphen

This is amazing and very close to what I am looking for! Thanks!

My goal is to design something that will work with late 1990s era 90-degree deflection fine dot pitch color CRTs around 13-inch in size, such as the M34AFA13X02. There are plenty NOS CRTs from that era out there, especially from places such as https://www.relltubes.com/ My understanding is that the yoke designs starting in around the 1990s for "bubble curved" screen, 90-degree deflection CRTs (like the M34AFA13X02 ) are wound in a way that automatically corrects for pincushion distortion. So that will make the circuit design more simplistic. However, these yokes use parallel windings and a much lower series inductance than the yokes described in the 1950s thru 1970s television theory textbooks that I have access too. Grob mentions parallel yoke windings in his 1970s edition of his textbook, but it sounded like a new technology at the time and so he didn't go into any detail other than mentioning it.

That French project used a much smaller deflection degree than 90 degrees. It is also not clear from a brief viewing what type of deflection yoke is being used, and it may be deflection plates?

This is what I am currently doing. The guy that runs that website is providing an incredibly valuable service to the world. I hope to help slow the fading of this technology into distant history by reading and digesting as much as I can from that site... and hopefully creating a cookbook DIY TV project to give back.

Next on my list after Grob's "Basic Television" is read Zworykin's "Television" 2nd edition. Zworykin literally invented television, and founded RCA's television division. I've paged through the PDF copy of his book that is available on www.worldradiohistory.com. I see he has an explanation of many formulas for electromagnetic deflection. Zworykin is a once-in-a-hundred-years caliber person, and his 1000 page "bible" is intimidating. Hopefully it has enough details for what I am after.

If you are trying to design a tube sweep circuit to drive a solid state era yoke you may want to look into matching transformers....In the 1970s there were test jigs built with a deltagun CRT and tube deflection style yoke that came with matching transformers that allowed them to work with solid state chassis...If you can use a matching transformer to drive a tube yoke with a solid state chassis then you should be able to use a matching transformer to do the opposite.

I was thinking that an impedance matching transformer would be needed for both horizontal and vertical deflection output stages. Similar to how tube audio amps typically have an output transformer to step the output voltage down and the output current up. Grob describes a tube amp output stage for vertical deflection and the need for a transformer to couple the beam power pentode plate to the vertical deflection coil. But unlike his chapter on video amplifier design, which is a step by step cookbook, Grob doesn’t specify step by step formulas for determining the impedance matching transformer parameters and beam power amplifier parameters for a given yoke inductance.

Not knocking Grob, by the way. His book is incredible and provides cookbook level step by step details for literally every other aspect of a TV chassis. I think deflection tech is the one thing that changed the most throughout the era of CRT displays. So unlike his other chapters, he kept the explanation at a high level so that it wouldn’t become quickly outdated. Searching scholar.google.com, it appears as though active research was going on in computer driven yoke design up until the early 2000s.

If you tracked down a test jig with the impedance matching you could probably reverse engineer it. The 70's RCA jig I have supports everything from the 1955 RCA CTC-5 to stuff from the 90s according to the setup aeapter books the shop that used to own it sucribed to...

I presume you are planning to use an existing yoke? That would be a good idea, not only for image geometry and convergence, but because you could measure the yoke currents and drive voltages that you need to produce in an existing set. Then you "only" have the problem of winding transformers that will match a tube circuit to a yoke that was designed for transistor drives without the nearly impossible job of determining yoke design.

Some things to know:

Vertical sweep operates the yoke as mostly resistive, with some inductance. In other words, you need a class A (or close to it) drive, and will have a voltage waveform that consists of a sawtooth plus some amplitude of pulse during retrace.

Horizontal sweep operates the yoke as mainly inductive, with a switch-mode horizontal output that applies an essentially constant voltage to the yoke inductance, producing a sawtooth current (in concert with the damper tube), and a large voltage pulse during retrace. The inductance producing the sawtooth is the parallel of the flyback transformer and the yoke. Impedance matching is obtained by the winding ratio of the taps for the horizontal output device and the yoke. To use a tube HO to drive a low impedance (transistor) yoke, you will need a much greater ratio than was used in the transistor set the yoke came from. This means the tube capabilities will be matched (higher voltage retrace pulse and and lower current during active trace compared to a transistor). In tube circuits the damper tube may need a separate winding ratio of its own, whereas transistor circuits often put the damper on the same tap as the output transistor. Transistor circuits might have HO collector retrace pulse on the order of 1 kv or less, while tube circuits might have multiples of 1 kv. Retrace voltage and retrace time are also affected by the value of the retrace tuning capacitor.

Service info typically says "do not measure" and doesn't tell you the amplitude of the pulse, which takes a special high voltage scope probe to measure, so they didn't want to tempt technicians to blow out their scopes.

By the time sets were in wide commercial production, no designer in their right mind would take the trouble to go back to basic theory to calculate winding counts, inductances and resistances unless forced to by radical changes in design, like going to transistor circuits. Instead, they woud typically take a similar design and modify it slightly to match the new needs (the new yoke impedance is x % lower, so change the winding ratio by that much) and then fine tune the prototypes on the bench.

The flyback transformer is also responsible for generating the high voltage. For least high voltage variation with beam current, you actually want the HV winding to resonate at three times (or even five times) the primary retrace frequency, to present as flat-topped a pulse to the HV rectifier as possible. This is a refinement that I wouldn't expect any DIYer to make, as it depends on the distributed capacitance of the HV winding. Since this was affected greatly by whatever insulating compound was used for the HV winding, it always was trial and error based on previous designs, modifying the number of layers and therefore number of windings per layer. The head sweep designer at Zenith developed a method mixing a silicone liquid bath with the same dielectric coefficient as the potting compound, so a flyback prototype could be run in a beaker of the stuff instead of actually potting it. This way, if the tuning was off, you could modify the prototype instead of starting from scratch.

You can get by without precise tuning of the HV winding, you just may get more HV variation than a carefully engineered production set.

Can you share schematics? Also, can you take impedance and inductance measurements of the transformers you have? We should take inspiration from the DIY tube audio, community and pull together DIY TV schematics as a community. Repair and restoration are only part of what is needed to keep tube television technology alive. Open source DIY schematics that can be simulated in LTspice would be an amazing educational tool for many.

One trend in CRT electromagnetic deflection technology was towards lower impedance, lower inductance yoke windings, as well as from series windings to parallel windings (which helps lower impedance and inductance). I don’t understand why that trend occurred though, but it appears to be aligned with the transition to solid state.

Basically, I want to back port many of the improvements that occurred in the solid state era to a pure tube chassis such as the improved yokes (better geometry), higher CRT G2, focus, and acceleration voltages (better focus and spot size), finer mask dot pitch, and smaller G1 aperture (smaller spot size).

Another improvement that I will try to back port: some of the compactron tubes that were designed for video amplification are small and lightweight enough to plug in to a printed circuit neckboard that has a design similar to a late 1990s solid state neckboard but with compactron amps as opposed to BJT transistors. The advantage would be reduced parasitic capacitance, inductance, and resistance in between the video amplifier output stage and the CRT cathodes. This would allow for higher video bandwidth at the same plate current, or, lower plate current at the same bandwidth. Lower plate current would prolong the tube amp’s lifespan.

I am torn on whether or not to cheat and use a 1990s DST flyback transformer. They have solid state rectifier diodes for the G2, focus, and high voltage anode lines. The majority of the tube audio community is OK with solid state power supplies, and keep the signal paths pure tube. Tube TVs have 2 power supplies: flyback power supply and B+ power supply. At first I will probably cheat and make those solid state, and later design non-solid state replacements. The non-solid state flyback power supply designs are very impressive, especially the ones that use a shunt diode tube for regulating the CRT’s high voltage. Most 1990s solid state TVs lacked high voltage regulation and yet pure tube TVs in the 1960s had it! A great example of newer tech sometimes taking a step backwards.

I am definitely going to use the yoke that was designed for the late 1990s M34 CRT. As I mentioned earlier, this style of CRT has the yoke designed using computer simulations that allow for pincushion correction to come “for free” due to the shaping of the windings. I have taken impedance and inductance measurements of the windings and they are much lower than the values I see in non-solid state TVs.

Grob does a good job of describing the variations in the waveforms required for vertical or horizontal deflection, as well as cookbook level details for how to design circuits for producing the waveforms. Where he stops short is the precise formulas for output stage voltage, current, impedance, and inductance matching. I am afraid I will blow up my yoke unless I have a formula for calculating reasonable starting values that I can then tune in a prototype circuit. Remember, I am using a late 1990s yoke in a circuit design that is coming from a 1960s textbook.

This is super helpful for me. I am going to add this to my project notes. Thanks!

I have a 3kv probe and a 40kv probe. The horizontal output pentode specs say the plate can handle up to 5kv. Having spent most of my life with solid state TVs and horizontal output transistors, I laughed when I saw the crazy voltages horizontal output tubes deal with. 5kv would immediately destroy a horizontal output transistor 😂. In general, tubes are better at handling high voltages, and transistors are better at handling high current.

It is totally understandable that practical engineering would work this way, but also for the current state of things, very worrying for keeping the technology alive today. As these engineers retire and move on, memories fade, and the art of the technology dies forever… unless we do what the DIY tube audio community has done. They have taken their favorite technology to the point where new tubes are being manufactured to this very day such as the Western Electric 300B vacuum tube (https://www.westernelectric.com/300b).

I know I will have to tune values of parts in a prototype circuit. I just wanted a more airtight understanding of the output stage so that my starting point is closer to not blowing up my yoke 😂

Actually, you said it yourself - it's because of transistors being high current, low voltage devices compared to tubes.

First a correction. Almost all solid state ( SS ) sets do have HV regulation.
Its all done with the main power supply.

Its an interesting project. If you use a modern in line jug I would
try to work with the yoke it has. You may look into Zenith EFL tubes also.
They could light up a room & very reliable.

Another thought was what Panasonic & Sony did in the early 70's. They
had separate hoz sweep & HV stages. Just a cool idea if you want to
be different. It may also make some things easier.

Last note is if you eventually want to build a full blown OTA NTSC set
you can cheat with a Zenith IF strip (SS) & tuner of your choice.
Or if you want tubes look at the Zenith IF's on the sets from abt 1958.
They had 3 6BZ6 or 2 6EH7's & a 6EJ7. Those Zeniths had better
sensitivity & selectivity than ANY TV. They saw things other sets
were blind to.

73 Zeno:smoke:
LFOD !

I am hesitant to use a vintage CRT because they are rare and therefore should be reserved for restorations. Restorations are paramount for keeping this tech alive. My “Frankenstein” TV will use vintage vacuum tube diodes, triodes, and pentodes… which are less rare, thanks to the tube audio community hoarding them since the 1970s. I also have 6 NOS M34AFA13X02 CRTs, which is another motivator for taking a Frankenstein approach. I keep using the DIY tube audio community as a sort of guiding analogy. They often use vintage triodes with modern loud speakers, and they even use solid state power supplies. They keep everything non-solid state in the signal path. I am taking the same approach, with a 1990s solid state diode split flyback transformer.

This is such an awesome project! I have often thought it would be cool to do something similar but my level of knowledge is way below what's required to even start. Here's hoping this really takes off.

Speaking of which, and following on with the general theme of deflection and yokes, and not intending to hijack this excellent discussion, I wonder if someone can tell me if all of these huge pulses and inductors, etc., are strictly necessary. I mean, if one were to start over from scratch, in a vacuum, and set out to build a circuit whose requirements are to move an electron beam from A to B over a precise period of time, and given that we are already in possession of a waveform that exactly graphs that movement, would it be possible or feasible to simply drive a yoke winding from a simple power amplifier circuit? Both V and H?

I don't know how much juice is really needed to deflect that beam five inches on a 34CM CRT, but I do remember as a kid back in the 80s building a lissajous project from Radio-Electronics out of a 19" tube B&W set and a second yoke, and powering it from a 6L6 amplifier, and that it could very nearly move the beam off the screen. I've often wondered why, if that's the case, they didn't just put an STK chip or similar in these sets along with a SMPS for the HV.

I totally agree!

Regarding output amplifier type:

The horizontal sweep is operating a load that is mostly inductive at horizontal frequency. This is because of the number of windings needed to create a magnetic field to deflect the beam. This dictates that the output circuit apply a constant voltage to the yoke during sweep, so the output ends up just switching on and off. This is convenient because it reduces the power handling requirements (wasted energy) tremendously. The horizontal output device has to conduct for a bit more than the right half of the sweep (to make up for resistive losses) but most of the energy is stored in the magnetic field and transferred to the resonant flyback capacitor and then to the damper diode to make the left half of the sweep.

In both tube and transistor horizontal outputs, you want the switch-off to be as fast as possible, to prevent having both current and rising retrace pulse voltage present simultaneously. In tubes, it's a matter of total power dissipation, and in transistors it's much more critical because the safe operating area (limit of simultaneous voltage and current) is very small. A bad design will kill an output transistor on the first retrace, before you can even figure out what happened.

The vertical output amplifier is much closer to linear, modified somewhat, because at vertical frequency the yoke is mainly resistive but partly inductive. So, it already operates mainly as you proposed.

Electrostatically deflected tubes use linear drive waveforms with incidental power losses in the amplifier but none in the deflection plates.

I finally understand what the damper is doing! Thanks for the explanation.

Just an update to keep people posted on my progress. I am still doing research. Thanks to https://worldradiohistory.com/ I have put together a pretty extensive reading list for deflection aka "sweep" circuit theory and design. This reading list also has great chapters on video amplifier design, but honestly, the video amplifier design is easy compared to deflection. The only complexity in video amplifier design is the use of peaking coils and such for high frequency compensation.

Here is the reading list, in case anybody wants to read along. I skimmed thru each of these books very quickly to ensure that they have chapters on deflection circuit design and theory. As usual, many of the books come from the top engineers at RCA, back in the era when RCA was the leader in TV. I have finished reading both Grob's "Basic Electronics" and "Basic Television" textbooks, and I am now working through Kiver. Grob's "Basic Television" chapter on Vertical Deflection does show how an impedance matching transformer is used to match the high impedance vertical deflection amplifier pentode to the low impedance yoke. He shows the current and voltage wave forms at both primary and secondary side of the impedance matching transformer for an example circuit, but does not explain the theory in enough depth to adapt to a different yoke impedance. Still, very enlightening and a good read.

Basic Electronics 4th Edition by Grob: this book was so popular that it's 13th edition was just released last year. The benefit of the 4th edition is that it still has a chapter on vacuum tubes. The 5th edition and later have, sadly, removed that chapter. Grob is a great teacher and this book is great if you need to refresh your fundamentals, especially when it comes to AC circuit concepts such as reactance, resonance, etc.

Basic Television 4th Edition by Grob: if you are into Vintage TVs, this book is like a gripping novel. Extremely entertaining and you learn a thing or two by the end.

Television Simplified 7th Edition by Kiver: seems comparable to Grob's textbook. More on this once when I finished reading it.

Practical Television Engineering 2nd Edition by Helt: a quick glance suggests the chapter on deflection seems good, and the book seems approachable.

Television Cyclopedia by Coyne: similar to Helt's text.

Television 2nd Edition by Zworykin: this guy invented TV and transformed RCA from an audio-only radio company to an audio and video company. Quickly paging through the book, it is definitely not the first TV theory and design book to read because it is written for an advanced level engineer, but the formulas that explain the physics and circuits for deflection are likely what is needed to create a DIY tube TV, especially when mixing and matching modern yokes with vintage amplifier tubes.

RCA's Journal of Television Volume V: RCA published a 6 volume journal containing an editorially selected collection of the best scientific research articles by RCA's top engineers and researchers. This volume contains a detailed article by a Mr. Shade on deflection circuit theory. Since these are scientific research articles, they are written for an audience of researchers and engineers with a PhD or equivalent experience in electronics, magnetism, and circuits.

RCA's Journal of Television Volume VI: this volume contains another article on deflection circuit theory and design by Mr. Shade, and a quick glance shows that it is exactly what I am looking for. However, it is written at such an advanced level that I am going to have to work my way to up being able to read this.

Once I am done with this reading list, I will start to design, in LTSpice, an end-to-end vacuum tube only circuit for taking a YPbPr aka "component" video input, separating the horizontal sync signal from the Y input, doing phase detection, multivibrator horizontal oscillation, horizontal output amplifier, flyback transformer, and horizontal yoke. Then I will prototype this circuit in real life. Success will be if I can get the CRT to display a horizontal white line that spans the width of the screen. That means horizontal deflection works, and the flyback transformer is generating the proper screen, focus, and acceleration anode voltages. Adding vertical deflection will be easy after that to get a full raster. Then I can add color differencing to transcode YPbPr to RGB, and finally video amplifier to drive the CRT's cathodes.

Creating the tuner and IF circuits will be done last. Most people have content sources coming from DVD players, game consoles, and other devices that are already in component video format or can easily be converted into that format. So the tuner and IF circuits are just for completion and preservation of the vintage TV tech. I think most people will just directly use the component video input since that is going to be both easier and provide better video quality than doing RF modulation and then RF reception only to convert to YPbPr.

The output stage of both vertical and horizontal deflection circuits is a power amplifier. However, the horizontal deflection output stage has to deal with the high voltage created by the shock excitation caused by the rapid change in current flowing through the yoke during horizontal retrace or flyback time. In fact, this is one reason why the horizontal yoke inductance is lower than the vertical yoke inductance. Inductive impedance is higher, the faster the current is changed. The vertical deflection current's most rapid change is during vertical retrace, which is still significantly slower than horizontal retrace.

Thanks for that Luke! It really is interesting to consider the number of different, precisely-timed events that have to happen every second so these early sets worked.

If I can, I'd like to suggest one more book that might be interesting. I came across it a while back after a Google search for something turned up a passage from it. It was written later than Grob and the rest, and this guy has a different way of explaining things that allow them to sink in better than the other books, at least for me, likely because of more modern language. Coincidentally I was reading the chapter about video amplifier circuits, peaking coils, etc. just a few hours before I saw your post.

Anyway it's called Color Television Principles and Practice, by R. R. Gulati. Don't worry the book covers B&W in extreme detail before it even touches on color. I think it was originally a textbook used in India and the color stuff deals more with PAL than NTSC, but again the text is system-agnostic. Maybe it's a good resource to check out if you're stuck on something. You can buy it on Amazon https://www.amazon.com/Colour-Televi.../dp/8122400086 or since it's out of print you can download it for free as an ebook from Libgen.

Here's a second edition that looks very thorough:
https://docs.google.com/file/d/0B6AB...VU2CsQmqdAJ9_A

Thanks to both of you. This forum is a great place to collaborate and get help on projects. I love it!

I just flipped through the PDF copy of the second edition of Gulati. It was published in 2005 and yet still has in-depth coverage of pre-solid state TV design! Amazing! Also it seems to go into depth of the transient response formulas for the deflection circuits. Sadly in the USA, television circuits books published after the 1970s had all vacuum tube circuit theory removed and went pure solid state (of course except for the greatest vacuum tube ever, the CRT).

I am almost done reading Kiver. I am placing his text up in the top tier with Grob's text. There is a reason why both authors had their TV circuit design textbooks published over and over again in several editions. They are just plain great!

Getting the same theory explained by different authors is useful when teaching yourself. Sometimes for stuff to resonate with your brain, you need to have it explained multiple times by different people. Also, Grob and Kiver use snippets from real life popular models of RCA and Zenith TVs. Since they use different example brands and models in their books, reading both books means you also get to study more example real life circuits. That is another important thing. Learn the theory and study real life applications of it.

I now have a pretty good understanding of how to adapt both vacuum tube based vertical and horizontal deflection circuits to be able to drive a late 1990s diode split flyback transformer and late 1990s low impedance yoke (0.4 mH and 0.9 Ohm horizontal winding) and its matching ultra fine dot pitch "M-class" CRT. What I need is a small high efficiency ferrite core 20 to 1 transformer that steps down the plate voltage by 23 to 1 and steps up the current by 1 to 23 is what is needed to translate the high voltage, low current output of the beam power tube to the low impedance yoke. The ideal would be for these two windings to be on the diode split flyback transformer, as that would be more efficient, but for my first design, I wanted to stick to a "modern" diode split flyback transformer because they are super reliable and safe. Yes these diode split flyback transformers have 3 built in semiconductor rectifiers, so my TV design will have semiconductor rectifiers in the high voltage power supply... but I am also going to use a semiconductor rectifier in the B+ supply too, at least for my first DIY tube TV. In a later revision, I will try to modify the design to not use any semiconductors, but for the first version, it will be pure tube in all signal paths and a few rectifiers in the power supplies to convert from AC to DC.

Regardless, I want to refrain from circuit designing and circuit simulation until I finish working through the reading list. No reason to reinvent something that is written in one of these textbooks.

A voltage or current ratio of 23:1 sounds way too big. Horizontal output tube pulse voltage is not 23 times horizontal transistor pulse voltage. Maybe more like 5:1 ?

My numbers are off. I have a Panasonic BT-H1390Y professional video monitor that uses the M34AFA13X02 CRT and yoke that I have several NOS units of and plan to use in my DIY TV.

https://i.ibb.co/vdTZTM1/IMG-0509.jpg

I just checked the Panasonic BT-H1390Y with my 3kV oscope probe to check the voltage waveform at the collector of the horizontal output transistor and it is about 860 volts peak to peak. The math I am using is the max pulse plate voltage on the beam power pentodes that were commonly used for vacuum tube based color TVs, which is around 7500 volts divided by 860 = 8.7. So 8:1 is closer to what is needed. But a more accurate calculation would be to measure the vpp of the waveform inbetween the flyback pulses, right? Matching on the flyback pulse seems a bit of a backwards way of getting to the transformer step down ratio.

Where does your 7500 volts number come from? A maximum tube rating, or measurement of a tube type color set? You should use a measurement of an actual set, as this would include some reasonable derating.

Note that tube sets used unregulated power supplies, so line voltage variations had to be accommodated. This is a problem with running vintage sets these days, as AC line voltages have crept upwards over the years. Nominal for sets in the 50s was 117v, but line voltages these days are often well above 120. Another characteristic of tube circuits is that there is relatively large variation in bias and operating currents/voltages even if the supply is regulated. This is due to both tube variations and passive component tolerances. Tube circuits were designed with plenty of head room between nominal plate voltage and B+ to accommodate variations in B+ voltage, tube conductance (see spec sheets), and resistor tolerance (typically 10%), with nominal tube operating current and voltage appropriately derated so that operation was still possible under extreme variations. Transistor circuits, on the other hand, especially with a regulated supply, had well defined variability determined mainly by resistor values, due to the high current gain of the transistors.

I'm not sure what you mean by "the vpp of the waveform inbetween the flyback pulses." It's the voltages across the active device that count for its operation; and it's the p-p of the waveforms on the flyback and yoke that you need to duplicate to get normal width.

“Zworykin: this guy invented TV …. “

Pem Farnsworth Philo ‘s wife will dispute that comment.

Good read this thread.

Any chance somebody knows of a place that makes one-off custom high frequency (~15khz) transformers? I've wound my own transformers before by hand, and also done epoxy potting. Since I lack a winding machine, it was incredibly tedious to do it by hand and the results were not as cleanly wound as a machine wound transformer. Also, not having a vacuum chamber for the epoxy potting... my potting results ended up with tiny air bubbles, which is also not ideal.

A Google search shows dozens of custom transformer companies. I will contact them, but I worry that they will all only be interested in corporate customers that want to order at least a few thousand copies.


The dispute is historic. I think they both simultaneously invented electronic TV, but Zworykin accomplished much more... and Zworykin's book is amazing. If Farnsworth wrote a textbook on TV, I'd read it too. If it helps spark more controversy, I will repeat that RCA was the top TV company in the golden age of television. Zenith made amazing TVs, no doubt, but RCA provided more technological leadership.

You can make a vacuum chamber easily and cheaply enough. Get a pressure cooker from goodwill remove the safety valve and install a fitting for an $89 harbor freight air-conditioning repair vacuum pump (found in the automotive section).

Coil winding machines do show up used for reasonable if you keep your eyes open at various venues... I've considered one that has showed up at my local antique radio club swapmeet a few times.

My wife does resin work, which also would greatly benefit from a vacuum chamber (similarly for removing tiny air bubbles in the resin)... she would be ecstatic if I built a vacuum chamber. Poking in the DIY tube audio forums, it looks like there are shops that do one-off custom transformers. Looks like they have started getting big into toroidal transformers due to the greater power efficiency. Toroidy looks like a promising vendor.

The DIY tube audio enthusiasts make me feel much more normal and boring. They go super extreme with their builds, like using industrial strength tubes such as the 833A for power amplification. Wow!

Farnsworth was definitely first in inventing the camera and tube by 5 years. Zworykin couldn’t get his image orthocon to work so Snaroff sent Zworykin on a spy mission to Farnsworth’s lab. Zworykin tried to reverse engineer the image dissector at RCA from notes he took while visiting Farnsworth. It didn’t work. Then Snaoff payed visit to Farnsworth and tried to buy him out. That didn’t work, so Sarnoff unleashed his high power lawyers in a 7 year court battle and Sarnoff lost. The crucial point was when Farnsworth was a boy of 14 he drew this theory of camera and tube scanning electrons for electronic television on a blackboard for his science teacher. The teacher was so impressed, he preserved the drawings and presented them in court. This showed the inventions were conceived when Farnsworh was 14. Farnsworth created working electronic television by 1927 in a demonstration to his colleagues. The first public demonstration was 1934 at the Franklin museum in Philadelphia 5 years ahead of of Snaroff’s 1939 demonstration.

When Sarnoff lost his 7 year court battle with Farnsworth, it was the first time he had to pay patent royalty fees and cross licensing fees to his competitors. This was his mode of operation, and did the same to Westinghouse, Philco and Akein. The government was chasing Snaroff for 39 years for violating anti trust laws and finally in 1958 RCA has to release all their color television information to their competitors to avoid a monopoly.

Snaroff operated heavy handed as a bully, but I agree, after Sarnoff destroyed Farnsworth and became king of the hill, RCA spent millions to promote and develop color television more than any other.

There is a statue of Farnsworth at statuary hall in the Capital. Farnsworth was inducted posthumously into the Academy of arts and sciences in 2003 and in 2013 Pem Farnsworth accepted posthumously, an academy award for the inventions of the image dissector and electronic television. Ironically, Sarnoff’s son was in attendance.

I finished reading all of the textbooks in my reading list, and none of them described the particle physics theory for deflection... except Zworykin, who walked step by step through the derivation in chapter 14 of the second edition of his textbook "Television". From first principles, he describes the derivation of the formulas that link the circuit theory of current through an inductor generating a magnetic field to the amount of magnetic energy needed to deflect a cathode ray by alpha degrees:

https://i.ibb.co/6tJvBXv/Screen-Shot...1-52-13-AM.png

K = ratio of deflection core diameter to inner deflection diameter (typical value is around 1.66)

D = CRT neck diameter

V = CRT acceleration anode voltage

alpha = peak deflection angle required after recycled damper current is exhausted

W = the magnetic energy in joules.

This can then be combined with the standard formula for the magnetic energy (L*I^2)/2 generated by current I in amps through an inductance L in henries, to calculate the peak current and average current needed to sweep a screen. I have created an automated calculator for these formulas in a Google spreadsheet here. Just copy the spreadsheet and update the input parameter values to see the output amplifier design requirements for a given CRT, yoke, and output transformer. Below is a screenshot of what the calculator looks like, with example values for the CRT and yoke that I am using in my DIY TV:

https://i.ibb.co/N3nfx90/Screen-Shot...2-11-20-PM.png

I have updated the CRT Deflection Calculator to also calculate the specs for the impedance matching output transformer, as well as the flyback capacitor, peak primary, and peak secondary side voltages, using the derivations from Gulati's textbook.

That's really great work you're doing!

Thanks for the encouragement. I finished the reading list I posted earlier, and found the the two papers from RCA's engineer Otto Schade to ultimately be exactly what I was looking for. His papers are in RCA's Journal of Television Volume V and RCA's Journal of Television Volume VI. It turns out that Schade's two papers were quite famous for their time, as they were the first publication that fully explained what became the standard circuit design for the horizontal output stage.

Schade has horizontal deflection formulas that are equivalent to those from Zworykin, which makes sense. Zworykin was Schade's colleague at RCA. Anyway, their formulas for the required deflection current have a parameter for the outer to inner deflection ratio. Read the papers for the detailed definition. The value that they use in their examples for this parameter is 1.66, but based off of my real world measurements for color CRTs, the parameter value is 6.82. This makes sense, as the published formula and examples they used were for black and white CRTs.

Another great paper is "Deflection and Convergence of the 21-inch Color Kinescope". Again, an RCA research publication, but about 6 years after Schade and Zworykin's publications, and this time describes the horizontal output stage for an early color TV. When I apply Schade/Zworykin's formula for the current required for horizontal deflection for an X-degree CRT, a Y-inductance yoke, and Z-volts ultor... as long as I use 6.82 for the parameter mentioned above, the required current is exactly what "Deflection and Convergence of the 21-inch Color Kinescope" specifies. So the Schade/Zworykin formula tests out for both early color CRTs and late color CRTs (by my personal measurements of my late 1990s CRTs).

One last recommended reading, is just 8 pages. There is one of the most amazing vintage TV restoration enthusiasts that I just learned about, Dr Hugo Holden. Looking at some old forum threads on another site, it looks like he realized, like me, only decades earlier than me, that television textbooks do not provide enough detail about the design of the horizontal output stage. Dr Holden wrote this 8 page article to give an overview of the theory and design of the horizontal output stage, and Dr Holden has clearly read Schade's work and more. It is a terrific read. Any chance Dr Holden lurks on these forums?