Electron gun

[makorihi]

Hey all,

I've been working on a project that requires a crt that I took from a mini TV to work. I've located the pins for the deflection and focusing coils as well as 4 pins from the back of the electron gun + the HV anode. I've also isolated the flyback transformer that came with it and figured out which pins are electrically connected (and I've tested the operation of the flyback to some degree).

Now the next logical step for me is to test the flyback with the electron gun, but I'm afraid of doing some irreversible damage to it. I'm pretty sure that electron guns normally use tungsten as a filament which I have to heat up to about 2700K to form a nice electron cloud. But I don't know how much voltage to apply across the filament to get it up to that temperature, and I don't want to accidentally melt it. The filament resistance that I measured is at about 38.7Ω

Does anybody have any insight or advice?

 

[Dick Cappels]

The aging rate of tungsten heaters is probably similar to that of the filament in tungsten light bulbs, which is to say that they probably age as the sixth power of the heater voltage. There is another danger, which is if the temperature of the cathode is too high, the electron emitting mixture on the cathode will boil off, leaving a very thin and useless electron cloud.

The only safe bet is to find thet heater voltage for which the manufacturer intended to drive the CRT. You might find a clue if you can find a schematic or service manual for the TV from which the tube was salvaged. Some sets took the heater voltage from the flyback transformer, so if you can find the specs for the flyback transformer (yes, almost impossible), that might help as well. What is the part number of the CRT? Color or monochrome?

The flyback transformer brings us to the topic of deflection. Make sure you have sufficient deflection current before you even start to apply anode voltage (if you make a flyback system, you can merely leave the anode cap off the CRT). Make sure G1 is negative enough below the cathode to assure cutoff (it may need to be many tens of volts), and slowly raise the G1 voltage.

 

[Nigel Goodwin]

CRT heaters are normally the standard 6.3V - but in any case are normally fed from the flyback transformer anyway, so connect it to the flyback transformer just as it was.

 

[makorihi]

CRT heaters are normally the standard 6.3V - but in any case are normally fed from the flyback transformer anyway, so connect it to the flyback transformer just as it was.

The problem with using the flyback directly is that I don't know the correct anode voltage (I hadn't tested it before taking it apart) I know that the anode has to be around some amount of kV, but I don't know the value for this specific one.

What is the part number of the CRT? Color or monochrome?

It's a black and white 14sx8y4 CRT. I've never been able to find a datasheet on it, but maybe I haven't looked hard enough . . . Also, by G1 do you mean grid voltage?

 

[Dragon Tamer]

Please post your circuit, if you're making the same circuit that I'm thinking of, we may have a problem. (not between us, with the circuit)

 

[Dick Cappels]

I'm not sure how much help this is, but the CRT was manufactured by:
Fu-Hong Computer co.
Phone:* 886-8-8321190
No.165 zhongzheng Rd., Donggang Town
Taiwan

 

[Vizier87]

Someone's gonna flip here... The nomenclature is generic ain't it..

 

[Dick Cappels]

No. Anything but.

 

[Nigel Goodwin]

You need to use the entire Line Output circuit, and the power supply that fed it, plus the frame stage as well - if you've just ripped the transformer out you're pretty well stuffed.

 

[Dragon Tamer]

Oh, never mind. I thought that your title was called Electron Cannon. (impossible for the most part) My bad.

 

[Externet]

Let me risk it:
The first 2 figures on the CRT part number sticker tell the filament voltage. That is 14V
Second, confirm if such mini TV supply was meant to work with 12VDC input, to make my affirmation safer.
Third; check continuity from the power switch to the filament pin on the CRT socket. If goes direct, good.
Last; apply 6V to filament. No CRT filament that I know of works with less. A faint glow will tell you if needs more. That would be 12V

 

[makorihi]

You need to use the entire Line Output circuit, and the power supply that fed it, plus the frame stage as well - if you've just ripped the transformer out you're pretty well stuffed.

Yes I've taken the transformer out, but I still have the original board and a guess at what the connections are.

Let me risk it:
The first 2 figures on the CRT part number sticker tell the filament voltage. That is 14V
Second, confirm if such mini TV supply was meant to work with 12VDC input, to make my affirmation safer.
Third; check continuity from the power switch to the filament pin on the CRT socket. If goes direct, good.
Last; apply 6V to filament. No CRT filament that I know of works with less. A faint glow will tell you if needs more. That would be 12V

Yes, the mini TV worked on 12VDC, but the filament is tied directly to one of the transformer pins. I suppose I can use this to figure out what voltage to run the transformer at, assuming the frequency is at about 15MHz.

Thanks for the replies

 

[Dick Cappels]

Perhaps you mean 15 kHz (rather than MHz).

On small thing: Check our your gate drive circuit's behavior during start-up and shut-down. It is quite common in TV monitor prototypes for an extra long drive pulse to be generated just at the circuit is switched on or switched off. An extra long drive pulse stores more current in the inductor, and when the drive goes away -poof! so does your output stage.

One other small thing: In traditional horizontal flyback circuits, the flyback primary is the energy feed choke for the horizontal deflection. The horizontal yoke is typically AC coupled to the collector (or in your case the drain) of the output transistor, the value of the capacitor being selected add a little parabolic voltage to the drive signal, resulting in what is referred to as "S" correction, which distorts the scan current waveform to compensate for the fact that the screen is not realy spherical. An additional nonlinear inductor is also usually placed in series to compensate for the loss in voltage across the inductance on the right-hand side of the screen, which is the result of the high IR drop across the resistance of the yoke and other things in series with it at the end of the sweep. In addition to that, an electrically controlled saturable reactor is usually used to modulate the entire horizontal deflection current waveform with a vertical rate parabola, so as to reduce the amplitude of the scan current at the top and bottom of the secreen; this is called pincushion correction because without this correction, the raster will resemble a pincushion. A very high quality, low loss capacitor will be across the horizontal deflection circuit (the flyback capacitor), typically between the collector and emitter of the deflection transistor to control the rate of rise of the collector voltage, indeed to control the maximum collector voltage since the half-sine flyback pulse amplitude is proportional to its width. On color CRTs, most of the magnetics are mounted up or integrated into the pre-aligned (so called "Yammed") deflection yoke itself. A diode, similar to the intrinsic reverse biased diode in the FET is used to catch the "free wheeling" current, through the yoke when the transistor is off; the yoke current then switches from going through the diode to the transistor when the transistor is turned on somewhere around 40% of the way through the active line., This diode (called the "flyback" or "damper" diode often has either a DC offset in series with it, or is driven by a small, but high current winding on the flyback transformer so that there is no change in voltage across the yoke circuit when the current swiches from going through the diode to going through the transistor.

And those are only the highlights. Maybe that gives you some cause to consider Nigel Goodwin's admonition to use the old circuitry. All those parts would be hard to source, and the whole circuit was "born" with each of its parts finely tuned to the others.

 

[makorihi]

I'm not sure why I had 15MHz in my mind, but it does sound pretty unreasonable . .

So I understand that there are many many many things that go into building this and that I don't know most of them, but that doesn't change the fact that I still want to get this to work. Actually, because this seems really hard to do, I feel like building it even more. It's never a bad thing to gain knowledge. But thank you for the warning and brief overview.

One thing that I've been wondering is regarding the horizontal output transistor. In all the diagrams I've seen, it is driven by a sawtooth waveform. Is this because more energy will be able to go into the transformer, as the inductor has a very high resistance to the current to start off?

 

[Nigel Goodwin]

You can't just run a LOPTX like that, it's a carefully tuned resonant circuit, you need to duplicate the original circuit.

 

[Dick Cappels]

The collector current builds up over time because that is what happens when you put voltage across an inductor -the current is a sawtooth.

The base drive waveform to a bipolar horizontal output transistor is usually pretty complex, and most designers go thorough lots of dead transistors before getting a good, stable design that also, but the way, makes a nice looking picture.

For some very lower power monochrome monitors deflecction circuits, where the deflection angle is small, the frequency is low, and the anode voltage is low, it can very simple. As simple as a pulse coupled through a capacitor to the base of a Darlington deflection transistor. Those were used in millions of early CRT terminals in the 1980's.

In higher power circuits, the transistor is "turned on" gracefully by the flyback in the base drive transistor (when the drive transistor turns off). This reduces the chance of a vertical stripe at the transistor's turnoff, usually, the base current decays just as the collector current is reaching its maximum at the end of the scan line. Far from ideal, but not too serious of a problem. At the end of the scan line, the driver transistor switches on and active pulls the horizontal output transistor's base negative to quickly sweep minority carriers out of the junction, with the goal of getting the collector current to a low enough value before the collector voltage rises to too high a value to still be under the transistor's safe operating area.

Other drive circuits use a "forced beta" approach in which an emitter tap on the base transistor is used to keep the base to emitter current constant, so base current tracks the collector (emitter) current, and an very clever one from Motorola in the 1990's in which the base current is ramped up as collector current increases, by switching the base drive voltage through an inductor, then when the base drive voltage is switched off, the energy stored in the series inductor is used to dump the base charge. A very simple and clever circuit.

 

[makorihi]

You can't just run a LOPTX like that, it's a carefully tuned resonant circuit, you need to duplicate the original circuit.

I don't need it to run exactly how it used to. There is no need for all of that if all I want is to drive an electron gun. I mean, circuits don't come out of thin air. Many people had to experiment, record data, theorize, etc. Anyways, arguing isn't really productive so I'll end that there.

I've found out that with the circuit I posted earlier, it works fine(for curious definitions of 'fine') with the transformer being driven by 5volts, but the transistor burns/dies when I step up the voltage to 12volts. It should be able to handle that because it's rated to handle quite a bit, and especially since I'm switching at many kHz, right?(I even tried putting a heatsink on it) But then why is it dying?

Also, is there any way to measure the duty cycle and/or frequency of a square wave without an oscilloscope or any expensive equipment?

 

[Nigel Goodwin]

I've found out that with the circuit I posted earlier, it works fine(for curious definitions of 'fine') with the transformer being driven by 5volts, but the transistor burns/dies when I step up the voltage to 12volts. It should be able to handle that because it's rated to handle quite a bit, and especially since I'm switching at many kHz, right?(I even tried putting a heatsink on it) But then why is it dying?

It's dying because (like I said) it bears no resemblance to the minimum circuitry required to run. Look at the original circuit, notice all the required components you've just ignored.

What use is just running the electron gun anyway?, all you'll do is burn a hole in the phosphur coating.

 

[Dick Cappels]

A couple of thoughts on the dead FET:

Two major differences between the original flyback circuit and the one you posted are apparent. The frequency is (probably) quite different, and maybe more importantly, there is no flyback tuning capacitor across the primary.

In a basic flyback circuit, current in the inductor builds up as long as the transistor is on. The higher the power supply voltage, the faster the current builds up. When the transistor switches off (assuming instant turn-off and everything is ideal), the voltage on the drain will rise in an attempt to keep the current flowing as the magnetic field in the flyback transformer core collapses. This voltage can get to be very large. In TV sets, most of the energy that was stored in the magnetic field in the core is transferred to the flyback tuning capacitor after the transistor switches off, and then (in the "falling" half sine wave of the flyback pulse, is transferred back to the flyback transformer.

The maximum voltage on the drain is determined by the amount of energy stored in the core (related to the volt-seconds across the core during the time the transistor conducts and the inductance of the primary of the flyback transformer) and the size of the capacitor. The larger the capacitor, the lower the power supply voltage, or the shorter time the transistor was on, the lower the peak drain voltage (and also, the lower the output voltage). With no flyback capacitor, you have very little capacitance, so the voltage could easily exceed the breakdown voltage of the FET, thus dumping all of the energy stored in the flyback transformer's core into the FET. And it might only take one or two cycles to kill it (depending upon the amount of energy in each cycle, and the design of the FET).

By the way, the flyback transformer was probably designed to work best with one particular flyback pulse width, and even with monitor flyback transformers designed to operate independently of the deflection circuit, a high Q, low value inductor is put in parallel with the flyback transformer's primary. This allows high frequency operation (the inductance of the flyback primary is usually pretty high) and it has the added benefit of providing a lower tolerance resonating inductance, and therefor more accurate flyback timing that would otherwise be obtained.

One other thing that the flyback capacitor does is slow down the risetime of the voltage on the drain, giving the drain time to turn off so the peak power is not so high.

In case I did not make the point about frequency clearly: If the gate drive pulse is too wide, the drain current will be too high.