Looking for pointers on Deflection Amplifier Output Stage Design

[AlanInSitges]

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

[LukeSimon]

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.

[LukeSimon]

Originally Posted by AlanInSitges

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.

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.

[AlanInSitges]

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.

[old_tv_nut]

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

[LukeSimon]

Originally Posted by AlanInSitges

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.

Originally Posted by old_tv_nut

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.

[old_tv_nut]

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 ?

[LukeSimon]

Originally Posted by old_tv_nut

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.



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.

[old_tv_nut]

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.

[etype2]

“Zworykin: this guy invented TV …. “

Pem Farnsworth Philo ‘s wife will dispute that comment.

Good read this thread.

[LukeSimon]

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.

Originally Posted by etype2

“Zworykin: this guy invented TV …. “

Bev Farnsworth Philo ‘s wife will dispute that comment.

Good read this thread.

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.

[Electronic M]

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.

[LukeSimon]

Originally Posted by Electronic M

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!

[etype2]

Originally Posted by LukeSimon

.
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.

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.

[LukeSimon]

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:



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: