Re: Ampliphone HV-other options?!

On Sun, 13 Jun 1999 01:51:38 -0500, Rodger Boots <rlboots@cedar-rapids.net>
wrote:

>Zonn wrote:
>
>> On Sat, 12 Jun 1999 02:34:05 -0500, Rodger Boots <rlboots@cedar-rapids.net>
>> wrote:
>>
>> >You can be connected to the CRT, just unplug the
>> >neck board.
>> >
>> >Resistor? Just where do you think you're going to find
>> >a resistor that can take that kind of voltage? A HV
>> >probe would do it, but without the capacitance of a
>> >CRT present the reading on the probe will be wrong.
>>
>> Doing a little calculation, it looks like a 10meg ohm resistor would give you a
>> nice 2ma load (not bad for testing). By doing a little more calculations you'll
>> find you'll need a 40watt 10meg resistor! (All assuming 20,000v)
>>
>
>2 mA!?! That would represent a raster monitor with a
>shorted CRT going supernova!

Cool! That would be fun to see! (From a distance of course..)

I was under the impression the HVT was using something like 5ma!! Making the
whole thing about a 100w supply, which seemed real high, but I just had no idea
how much current it takes to get an electron to travel through a vacuum. (Had I
thought it through I would have realized that some smaller TV's run on less than
100w.)

>For a RASTER monitor you don't need more than 1 mA
>MAXIMUM. For vector I doubt you even need half that.

I've assumed that components design for a 19" raster monitor would be overkill
for any of the vector games (including the 25" models).

>> >Be careful what you use for the transistor. The flyback
>> >pulse at the transistor collector is REAL close to 1,000
>> >volts! And the resonating capacitors must be able to
>> >withstand that, too. Without the capacitors that pulse
>> >will TRY to go to infinity, that's why monitor makers
>> >are real picky what they use there. It's also why they
>> >use multiple smaller value capacitors, if one goes open
>> >the pulse doesn't get too wild.
>>
>> I assuming by using the same one WG does I'd be safe. Isn't there a very fast
>> diode (whether internal or external) across the transistor to short this flyback
>> pulse? While the transistor is on, the voltage drop across it should be quite
>> low (1v or less), and when off it is (hopefully) protected by the bypass diode.
>
>Not quite right. When the transistor is on isn't a problem.
>When you turn it off the flyback pulse happens. When that
>happens you get several hundreds of volts across the
>transistor. You want to make REAL sure that transistor
>is turned ALL the way off during the pulse or it will instantly
>go into secondary breakdown and short out. Right after
>the flyback pulse (half a cycle into the flyback/capacitor
>resonance) the polarity will reverse. When that happens
>either the damper diode turns on OR the transistor is
>blown to kingdom come.

I suspected that the transistor needed to be turn off very quickly. Which made
me think the horz drive transformer does more the simply boosting the current
drive going into the horz output transistor. Since the output of this
transformer will also change polarity when the drive transistor is turned off, I
suspected attribute was being used to quickly clamp the horz output transistor.
I also suspect that this assembly is also somewhat a tuned circuit. There is
always a cap and a resistor across the primary of this transformer in all the
designs.

When dealing with circuits that are in danger of exploding, I have no problem
with wearing protective goggles. Upon initial power up I place a piece of
cardboard between me and the circuit. (Something I do anytime I power up a
Cinematronics monitor, or sound card, for the first time. Old tantalum
capacitors can go off like bullets!)

I might want to invest in a variac when playing with this circuit so I can start
everything with real low voltages.

>Or to quote the G08 manual's description of Q900 (the
>HVPS flyback transistor) "PWR Transistor 5A, 1400 V".

And the gain is something like 1.5 to 2! Which I assumed was the primary reason
for the drive transformer.

>> I need to brush up on resonant circuits, the raster monitors seem to run the HV
>> in a series resonant mode, while the vector supplies seem to place the tuning
>> capacitor in parallel with the Flyback. I would think we'd want a parallel mode
>> and "ring" it with the pulses from the transistor, kinda like a type C RF
>> amplifier. (I think you'd want parallel, for that, I just don't remember...)
>
>What makes it more confusing is that BOTH circuits are
>the same! The power supply is actually a virtual ground
>thanks to the power supply filter capacitors. So either
>way you are returning to "ground".

That makes a lot of sense, so either way the capacitor is in parallel with the
flyback, whether directly, or through the power supply. That's how I would
expect an L/C resonant circuit to work.

>Just connect the transistor emitter, damper diode anode,
>and resonating capacitors to a common ground point and
>you should be OK.

I plan on using a transistor with an internal diode (the same one WG uses). I
figure the closer to the emitter / collector junction the diode is, the better.
I'd be concerned that inductance in the connection between an external diode and
the transistor might be a problem when dealing with very fast switch rates.

Thanks for your suggestions Rodger, as usual they're enlightening!

-Zonn
Received on Sun Jun 13 13:22:18 1999