wrote:


Thank you all around!

I went through a couple of display motherboards, and looked into the
specs of all power devices.

What their applications have in common is being rather high current
and low frequency. You don't find much rated to work beyond fery few
MHz, and with very large capacitances.

Examples:

- power MOSFET capable of switching 600V, 10A peak, with 2200pF gate-
drain capacitance
These can be handy for switching stuff on and off, regulators, DC-DC
converters etc.

- bipolars that can absorb several watts, but no higher than a couple
of MHz

- horizontal finals yjat are included in fairly complex ICs with lots
of ancillary components in feedback and protection circuitry - at
least in CRT VDUs from the last 10+ years.

I doubt this - the vertical output is in an integrated IC, but the
horizontal output is normally just a transistor, often with a built-in
diode across the collector-emitter, but in the direction such that it won't
bother you in any sensible circuit. They are mostly pretty slow, since
they're high voltage.

I have no doubt some RF can be teased out of these, but it seems to me
it would be at the price of major design complications.
Or you could say that you would get a sense of accomplishment when you've
succeeded. Some of the transistors are mot so bad - e.g. I got a few
2SB772 (PNP) and 2SD882 (NPN) (10-Watt 30V, 3A, fT80MHz) transistors out
of a monitor, that'd be OK for a QRP transmitter I think.

The advantage of tube finals was that their high-voltage low-current
and low-capacitance nature lent itself to functioning also at higher
frequencies than those of typical video circuits.
I did find a 2SC3675, which has an output capacitance of only 2.8pF and a
VCEO of 900V. It is slowish (fT=6MHz), so I might use it as a cascode in a
high voltage audio amplifier for driving electrostatic headphones.

The video amplifiers in a monitor have to be pretty fast for high resolution
signals, and often these are discrete transistors to handle the power. I
got some 2SC2682 devices out of monitors - 180V, 3.2pF, fT=200MHz, 10W - a
few of these could make a useful amount of RF, I reckon.

Next I'll look into TV distribution amps. I remember that transistors
in those were designed to linearly pump a lot of different signals up
to UHF, and that they tended to run hot - and 24/7.

Good luck!

Chris

I already found something, thinking a bit sideways: these SS devices
impressed me more on the current- than on the voltage handling side.
I reckoned that scaling to higher powers was mostly done in current,
so I headed for a low power video unit I had: a long deceased bible-
sized 1970's TV + radio combo, with a 2" B/W screen. As suspected, I
found 2-3W low capacitance bipolars that could run into VHF... so
much so I had to bead them to keep them from producing spurs in my
ugly prototype.



On Jul 30, 11:37 pm, Chris Jones wrote:

2SB772 (PNP) and 2SD882 (NPN) (10-Watt 30V, 3A, fT80MHz)

The video amplifiers in a monitor have to be pretty fast for high resolution
signals, and often these are discrete transistors to handle the power. I
got some 2SC2682 devices out of monitors - 180V, 3.2pF, fT=200MHz, 10W - a
few of these could make a useful amount of RF, I reckon.


Hmmmm.... I'll look harder. The pile of monitors is still there
gracing a hidden nook of my place.

W.r.t. the high definition requiring high frequency: KVM switches
often quote their bandwidth, which must match what^s required from a
given definition (if a KVM only quotes that, and not pixels, good luck
finding the equivalents). Bandwidths now run in the 200-300MHz range,
which is a good omen.

Where would such transistors be?

In the vicinity of the HV coil?

Or near the CRT cathode?

wrote:


I already found something, thinking a bit sideways: these SS devices
impressed me more on the current- than on the voltage handling side.
I reckoned that scaling to higher powers was mostly done in current,
so I headed for a low power video unit I had: a long deceased bible-
sized 1970's TV + radio combo, with a 2" B/W screen. As suspected, I
found 2-3W low capacitance bipolars that could run into VHF... so
much so I had to bead them to keep them from producing spurs in my
ugly prototype.



On Jul 30, 11:37 pm, Chris Jones wrote:

2SB772 (PNP) and 2SD882 (NPN) (10-Watt 30V, 3A, fT80MHz)

The video amplifiers in a monitor have to be pretty fast for high
resolution
signals, and often these are discrete transistors to handle the power. I
got some 2SC2682 devices out of monitors - 180V, 3.2pF, fT=200MHz, 10W -
a few of these could make a useful amount of RF, I reckon.


Hmmmm.... I'll look harder. The pile of monitors is still there
gracing a hidden nook of my place.

W.r.t. the high definition requiring high frequency: KVM switches
often quote their bandwidth, which must match what^s required from a
given definition (if a KVM only quotes that, and not pixels, good luck
finding the equivalents). Bandwidths now run in the 200-300MHz range,
which is a good omen.

Where would such transistors be?

In the vicinity of the HV coil?

Or near the CRT cathode?

The video amplifiers are normally on the PCB that plugs directly onto the
back of the tube. The unusual thing about those transistors is that they
are moderately fast and also capable of handling quite high voltage.

Chris

On Fri, 14 Sep 2007 23:37:57 +0100, Chris Jones
wrote:

wrote:


I already found something, thinking a bit sideways: these SS devices
impressed me more on the current- than on the voltage handling side.
I reckoned that scaling to higher powers was mostly done in current,
so I headed for a low power video unit I had: a long deceased bible-
sized 1970's TV + radio combo, with a 2" B/W screen. As suspected, I
found 2-3W low capacitance bipolars that could run into VHF... so
much so I had to bead them to keep them from producing spurs in my
ugly prototype.



On Jul 30, 11:37 pm, Chris Jones wrote:

2SB772 (PNP) and 2SD882 (NPN) (10-Watt 30V, 3A, fT80MHz)

The video amplifiers in a monitor have to be pretty fast for high
resolution
signals, and often these are discrete transistors to handle the power. I
got some 2SC2682 devices out of monitors - 180V, 3.2pF, fT=200MHz, 10W -
a few of these could make a useful amount of RF, I reckon.


Hmmmm.... I'll look harder. The pile of monitors is still there
gracing a hidden nook of my place.

W.r.t. the high definition requiring high frequency: KVM switches
often quote their bandwidth, which must match what^s required from a
given definition (if a KVM only quotes that, and not pixels, good luck
finding the equivalents). Bandwidths now run in the 200-300MHz range,
which is a good omen.

Where would such transistors be?

In the vicinity of the HV coil?

Or near the CRT cathode?

The video amplifiers are normally on the PCB that plugs directly onto the
back of the tube. The unusual thing about those transistors is that they
are moderately fast and also capable of handling quite high voltage.

Chris

I pulled parts from 17" Viewsonic/Optiquest, and I found a treasure
trove of parts. The horizontal deflection amp is a VPS10S 3-channel
module (each indentical) tested at 100MHz. The data sheet is on
www.datasheet catalog.com. Vcc is 80 or 90 Volts, but Ic is only 75
mA. It's the only heavy-duty-looking thing on the board that mounts
on the CRT.

However, that monitor' shielding cages would be perfect for tube RF
cages, and there are a lot of large sections of it; a couple of square
feet, at least. The video amp is an LM1205N, 3 channel 28 pin device.
It apparently drives 3 2SC1921s, 200 V Vceo, 50mA, 600 mW, Ft=130
devices. Lots of C0G and Z7R decoupling caps on all the boards. This
is a 1997 vintage monitor, and it was a good'un.

On Jul 30, 11:37 pm, Chris Jones wrote:


..... I forgot..... Thank you Chris!