Re: Ampliphone HV-other options?!

On Thu, 10 Jun 1999 20:31:18 -0700, "Aaron Howald" <Ahowald@bilbo.w-link.net>
wrote:

>
>----- Original Message -----

From: Zonn <zonn@zonn.com>
>To: <vectorlist@mcfeeley.cc.utexas.edu>
>Sent: Thursday, June 10, 1999 3:42 PM
>Subject: Re: Ampliphone HV-other options?!
>
>
>> On Thu, 10 Jun 1999 15:27:52 -0500, you wrote:
>>
>> >Look at http://randyfromm.com/techdept/ for those schematics.
>>
>> Well I just checked out the WG schematics (JPG!, uck!, and parts of
>> the schematics are chopped off! Geeze!)
>>
>> Anyways we struck out there. The flybacks' primaries are designed for
>> 123v. It looks like most all the raster monitors run of this voltage.
>
>Yeah, almost all flybacks run off this (Except some AC/DC sets-some of those
>have a 15 volt B+!!).... but no has mentioned this: there is a capacitor
>wired from the collector of the H-out to GND on every H.V. system. This is
>called a safty cap. It sets the speed of the discharge though the FB when
>the transistor cuts off. Higher/larger value= LOWER High voltage= wider,
>lower spike on the HOT. Lower/smaller values give more "Spike" to it
>(higher/sharper), and generates more voltage...
> Maybe adjusting this will provide enough HV on a lower voltage B+?
>If a certain cap value works at 120 volts in (with 30KV output) then if the
>value is halved, 60 volts in would produce the same HV! (2X current in of
>course!)
>
>decimal places like .0015 uf.
>
>Getting 20 KV out of a t.v. transformer ran on 40-50 volts should be
>possible!

Yeah, I noticed that capacitor, I noticed it is always labeled "critical" with
high tolerances and had intended on playing with it. It looks very critical in
the tuning of the flyback frequency. If you look at the Williams "19" Raster
Monitor Type C instruction manual supplement" you'll see that a capacitor is
switched in parallel to this capacitor, to lower the resonant frequency (the
schematic says that adding the capacitor changes the sweep rate from 40us to
50us or from a 25khz to 20khz horz freq monitor).

I imagine a fixed input frequency with a changing capacitor would really mess
with the waveform! (Or a changing input frequency and a fix cap.)

With no changes to the input voltage, reshaping the wave will give you a higher
*peak* voltage, but it's not going to give you more current (the wattage E*I=P
will be the same -- nothings free), your current (at the higher voltage) will be
less, though you RMS voltage should remain the same (discounting core
saturation). When you drop the input voltage to compensate for the peaked
output HV, your current going in is also going to drop. (The ohms law thing.),
and will be reflected by less current coming out.

Sticking to the 123v also allows me to use the same horz drive transformer they
do, which also needs a 123v primary voltage.

I noticed that the capacitor exists on the Amplifone design (and I'm sure it's C
in the L/C tuning of the flyback, there are no other caps!) and the HV is
adjusted by changing the drive frequency, so this circuit acts like you describe
it! (Cool!)

I also noticed that this capacitor does not exist in the WG X/Y display, but
instead a cap is placed in parallel with the flyback, but they're not as
interested in a sharp resonance either. There is no yoke to be controlled with
the flyback voltage.

I'll have to play with those caps, and look for resonances while changing the
input freq's whenever I get something wired up on the bench.

>> So maybe the deal is to design a HV supply using an isolation
>> transformer, a 123v regulator (with error correction for adjusting the
>> output HV), and both the Horizontal transformer and Flyback of a WG.
>
>If the drive winding is isolated (either by winding it externally or by
>internal design) no isolation Xformer is nessasary. In fact, most Flys have
>a "Floating" drive winding, since the other end goes to the B+ supply-all
>other windings go to ground.
>The WG and AMP flybacks are the exceptions...

Good point!

The WG is almost isolated, it's only connected to GND through a 56k resistor.
And this is only because the primary side and the rest of the monitor share a
power supply. If a separate power supply were used for the HV section, the GND
could be removed.

The Amplifone ran the 555 off of half the power supply (and saved themselves a
zener diode), so they needed the GND as reference.

In normal use the line voltages would be isolated from the rest of the monitor,
but you'd damned well want an isolation transformer before working on the HV
section!

I'm a little (ok a lot) afraid of supplying a product that could kill you
(unlikely but it could happen) if simply touched a heatsink to see if the
transistor was overheating. Ok so that's a stupid thing to do in general, but
punishable by death seems kinda harsh. ;^) as Clay said earlier, no matter how
foolproof you make something, there is always someone stupider.

But your right, since the power supplies will be kept separate, you could safely
operate the monitor with the HV section plugged directly into the wall. But I
think for liability reasons I'll deny that that's the case! (Everyone please
delete this paragraph before archiving this message. Thanks, I knew I could
trust you all.)

> Also, the High voltage itself sould be regulated; NOT the B+! High voltage
>can vary 2-3 kv or more depending on screen brightness, causing blooming,
>lines not meeting, etc. Big-screen sets DO directly sense/regulate the high
>voltage, otherwise the picture would change size a lot (1-3 inches) with
>brightness changes.

None of the schematics I've looked at directly regulate the HV voltage. There
are not many transistors (if any) that can handle the 2-3kv break down voltage
that would be needed (and this is assuming you float the regulator) and on power
up time, when the picture tube is completely discharged, your going to have 20kv
across the pass transistor!

All of the schematics I've looked at so far regulate the output voltage by
regulating the primary sides voltage. The HV itself is not measured, but
instead a secondary winding is held at a constant voltage, by the regulator.
Apparently loading the HV winding, will drop the voltage in a secondary
(isolated) winding also. By keeping the output of the secondary winding
constant, you can keep the output of the HV winding constant.

On the WG K7000, this voltage is fed back to a regulator IC, from an isolated
winding, that controls the 123v supply.

On the WG K7100 the line voltage is rectified and then used to run a switcher
that supplies voltage to the rest of the monitor through a high frequency
isolation transformer - like a standard PC power supply. This allows use of the
monitor without an isolation transformer. The feedback to the switching
regulator is through a secondary winding of the flyback that is there just for
regulation. By not tapping into the HV directly (like on the GND side of the
Focus/G2 resistor divider) the switcher can be kept completely isolated from the
rest of the monitor.

On the WG X/Y monitor the feedback is a tap in the *primary* side of the FBT!
They are measuring the voltage developed during the collapsing of the magnetic
field (while the transistor is off), and using the developed voltage as an
indicator of the HV, kind of an auto-transformer arrangement. This is also
their B+ voltage.

The Williams "Type C" monitor has no feedback from the HV section (as far as I
can tell), the primary side voltage is simply regulated to 115v, and the HV is
allowed to rise and fall, if your picture goes from black to white. Hey, it's a
software problem, you don't want the screen to bloom, don't change brightness!,
and remember it's only a game, who's going to notice?

>> I didn't check the G2 voltages, but they could probably be made
>> usable, especially if they're too high. (Too low might be a problem!)
>
>Both foucus and G2 will be low on a T.V. fly-since the out put is only 20KV
>for vector monitors, not 30 KV as originally designed (wrong divider
>ratios). Hopfully the controls will have enough range...

I think that's a good thing, it looked like the voltages being generated by the
raster HVs were pretty high compared to the ones I've seen on the X/Y monitors.
If everything drops accordingly (and it should), then they just might be close
enough. After all the vector and raster games do use the same picture tubes,
you'd think the grid voltages would need to be similar in ratio.

-Zonn
Received on Fri Jun 11 03:20:31 1999