Sunday, 7 September 2014

Simple TV using DG7-32 'scope tube

A while I go I set myself a challenge... to design an make a simple TV using only valves. Not a semiconductor in sight.

WARNING. Do NOT try this at home. This project calls for some high HT voltages. The main EHT rail is some 700 VDC. It will floor you or kill you if you come into contact with it. It will not warn you, blow a fuse, nor pull your safety trip. It may not give you a second chance. RESPECT IT'S AUTHORITY

I will not, under any circumstances, accept any liability if you decide to re-create this project for yourself.

I procured an electrostatically deflected DG7-32 oscilloscope tube from the usual sources, a base , a handful of very nice Russian 6N1P valves, a couple of EF91s and a ECL83 for the sound. An EZ81 provides (some) rectification. Enter some transformers from the scrap bin, some tag strips and a couple of scrap aluminium enclosures.

My first design was too complicated, used too many valves, and, most importantly didn't work.


You'll notice a few things:

1) The raster is all "in one corner", this is because I didn't employ a differential drive to the X &Y plates.
2) There's precious little video modulation. I gave most of it away to Miller capacitance.

All-in-all, rubbish.

It sat on a shelf for a year, sneering at me.

I took it to pieces, and completely re-vamped it. I'd subsequently robbed the mains transformers for another project, so I ended up with a nice transformer from an old audio amplifier to provide the +300V and 6.3V for the heaters. I also needed to develop the 700 V EHT. In the original design I had a separate transformer, with a nice high voltage winding and an isolated heater winding. No such luck this time. I had a small insolating transformer from an old bathroom shaver point. 240VAC in, 240VAC out. This would have to do.

Let's have a quick look at the schematic:-

Yeah, you're going to want to zoom in on that a bit...

The Power Supply

Mains comes in and is fused by F1. TR2 and TR3 are actually one transformer, with centre-tapped heaters and HT winding. R1 and R2 provide a little peak current limiting to keep our EZ81 rectifier happy. The cathode of the rectifier is connected to our main smoothing capacitor C1. This must be no larger than 50uF, so 22uF is fine here, and we're not pulling much current. R3 and C2 provide additional filtering, and R4 and C3 provide a lower voltage supply for the screens of our video amplifiers. TR1 is the shaver transformer, and provides and quick and dirty 350V supply via D1 and C4. This 350V supply is not referenced to earth, but our 300V HT supply, giving around 650 VDC (More like 700VDC off-load)

I'll repeat my warning at the start of the article at this point. This power supply is more than capable of supplying enough current to kill you. Yes. Really. This means you. Don't do it. 

Whilst I was developing this gadget, I kept a voltmeter permanently attached between GND and 700V. Also, I kept a 10W 10K resistor connected to GND, and to an insulated lead, which was connected to the +700 VDC rail whilst I was working on the circuit. I've also installed 2x220K resistors in series directly across C1, C2 and C4 to bleed away any charge within a few minutes. These resistors aren't shown on the diagram. 

The original version also had separate fuses for each HT supply, and, stupidly, for the EHT supply. I used the ordinary type of 20mm fuse. During development, I inadvertently shorted out the EHT supply. The 20mm fuse blew violently, vapourising the wire against the inside of the glass, which carried on conducting! The resulting burned mess of fuseholder resulted in a lesson learned. Fuses have voltage ratings. 20mm fuses are rated to 250V. 

The Signal Stages.

Video comes in on X1, and is loaded by a 75 ohm resistor. Video is passed via C1 to the grid of our first video amplifier VT2, an EF91 (6AM6) pentode. This is further amplified by VT3 and passed to the first grid of the CRT. 

Video is also taken from VT2 to VT4a, the sync amplifier. Here some of the video information is lost (mainly through miller capacitance), but the line and frame sync pulses are retained. The sync pulses are passed from the anode load to VT4b, the frame sync separator, whereby the higher frequency line sync pulses are filtered out by C13 and R19, and the remaining frame sync pulses are used to sync the fantastron oscilator formed by VT5a & VT5b. This gives us a nice sawtooth waveform, locked to our video frames at 50Hz (in the UK). Although I haven't tried it, I'm sure it will work just fine with NTSC at 60Hz.

The frame sawtooth waveform is then passed to VT9A & VT10B (what happened there? It should read VT9B!)  which forms a differential amplifier and drives the Y-plates of the CRT. The plates are coupled to the anodes of the differential amplifier by 10nF, 1.6KV capacitors, as the deflection plates are referenced to the EHT by R46 and R47.

The sync pluses are also fed from VT4a via a high pass filter formed by C16-19 and R23-26, to pass the line sync pulses, to another fantastron amplifier, this time synced to the line pulses at 15.625 KHZ (again, I've not tried it, but it should be OK with NTSC too).  Once again this syncs another fantastron oscillator, and provides us with our sawtooth waveform to drive the X-plates via differential amplifier VT7a and VT8b (Again? This should read VT7B!)

The CRT is provided with a focus voltage via R50, and the cathode is supplied by R51 to slightly elevate it above 0V, to allow our negative going video to modulate the grid. 

Audio is provided by a very simple ECL83 amplifier. It's not going to win any awards for design or performance, but it functions. The output transformer is simply a 9V mains transformer. I know all about the fact that the transformer is totally unsuitable due to DC on the windings etc etc, but I had it to hand, and it works! 

Construction is fugly to say the least. 

I intended it to be a sort of homage to the home constructor of the 1940's / 1950's. 

There it is... exposed HT wiring an' all...

Performace isn't fantastic. The circuit is simple, the tube was never designed to show a raster scanned image... but it's definitely Dr. Sheldon Cooper!

Some retrace (flyback) lines are visible, as there's no supression. 

Feeling cheated? Yes, me too. I promised a design with no semiconductors, and yet the EHT rectifier is clearly a grain of sand. As I said originally, the EHT was originally supplied by a separate transformer, and this allowed me to use a valve rectifier (I actually used an EZ80), and I elevated it's heater winding to 350VDC, so as not to compromise the heater/cathode insulation. I did try putting the EZ80 into this circuit (if you look carefully you can still see it's Noval base below the speaker magnet) but the heater cathode insulation was just not good enough to hold off the ~700 volts and flashed over :(