The hardest part to build on an oscilloscope (while still maintaining
bandwidth) is the final deflection amplifiers.
Something had occurred to me a little while ago. There is a readily
available supply of chips that will do the job in a specific product. That
product is CRT SVGA monitors.

The video amplifiers that drive the grids of the electron guns.

Normal supply voltage is 80 to 100V.
Normal output swing is 10 to 75V or 10 to 95V
Normal signal input voltage 0 to 5V
They are capable of driving 8pf loads which is close to the value of a
deflection plate.
Output response times are 10us or faster. Some of the newest are down to
2.5us.
Corresponding bandwidth is DC to 50 or 200MHZ
They normally come in a self contained package that contains three
amplifiers.

Since most scope most scope tubes take less than 80 volts per plate for full
deflection (all the ones i have seen), then they make a great deflection
amplifier.

They would be great for upgrading an old tube type scope up to a solid state
100Mhz unit. The rest of the amplifier chain would be easy to built. They
would also be great for building a scope from scratch. Remember to keep lead
lengths short, and use shielded interconnect cables, because at 100Mhz it
will oscillate if it can hear it's self.

That is what I am currently doing to an old EICO scope. I already have the
vertical amplifier in place. It uses two sections of a VPS10S. For The
horizontal amplifier I am going with discrete components, because it doesn't
need to be as high of bandwidth. I have got the horizontal amplifier built,
and I just need to get it mounted.

The next stage is the amplifier stages between the input, and deflection
circuits. They are all going to be 5V rail to ground. There will be a couple
+10 and -5V point in the circuit for level transition, and signal
inversions. And I have to build the horizontal oscillator and triggering
circuits.

The only thing that has been bugging me is, what can I do with the old tube
sockets. I have been trying to think of some tube circuits I could use. The
only things I could think of is a high impedance input buffer. Or use a 6AL5
as a wide band RF rectifier. Or I could put in a sweep oscillator and a IF
amplifier and make a rudimentary spectrum analyzer.

On Feb 28, 12:16*pm, "N9WOS" wrote:
The hardest part to build on an oscilloscope (while still maintaining
bandwidth) is the final deflection amplifiers.
Something had occurred to me a little while ago. There is a readily
available supply of chips that will do the job in a specific product. That
product is CRT SVGA monitors.

....

Wow, I admire you for even thinking of tackling something like this!
You're right that the video drivers from modern fast monitors should
work fine for deflection amplifiers, perhaps with a bit of care about
frequency compensation, and so long as they have enough output to
drive your particular CRT.

On the other hand, it does seem like a LOT of work to go to
considering that you can get a really nice used analog scope for under
$100. I have an HP1715A 200MHz 2-channel scope with nice main+delayed
sweep that I need to put on ebay (or equivalent), and I seriously
doubt I'll get $100 for it.

If you do go to all the trouble to revamp your Eico (maybe the same
scope model I started with almost 50 years ago??!!), I do hope you'll
put a good triggered sweep in it, and not just a "horizontal
oscillator." It's not too difficult to make a very respectable linear
sweep with op amps and fast comparators these days. For triggering,
for example, have a look at the Analog Devices ADCMP02 and its
siblings: fast, built-in hysteresis... I'm using one in the input of
a counter, and it goes to 100MHz easily for low level inputs, and to
well over 200MHz if I jack up the input a bit.

Cheers,
Tom

On the other hand, it does seem like a LOT of work to go to
considering that you can get a really nice used analog scope for under
$100. I have an HP1715A 200MHz 2-channel scope with nice main+delayed
sweep that I need to put on ebay (or equivalent), and I seriously
doubt I'll get $100 for it.

I have the old scope laying around so why not? And almost everything I am
using in it is salvaged parts. Including capacitors, transistors,
transformers, and the video chips in question. I will probably have under 10
dollars in it after I am done. (That includes the $5 I paid for the scope a
few years ago.)

If you do go to all the trouble to revamp your Eico (maybe the same
scope model I started with almost 50 years ago??!!), I do hope you'll
put a good triggered sweep in it, and not just a "horizontal
oscillator." It's not too difficult to make a very respectable linear
sweep with op amps and fast comparators these days. For triggering,
for example, have a look at the Analog Devices ADCMP02 and its
siblings: fast, built-in hysteresis... I'm using one in the input of
a counter, and it goes to 100MHz easily for low level inputs, and to
well over 200MHz if I jack up the input a bit.

For the sweep, I was planning on using a transistor as a constant current
source to charge the capacitor for a nice linear sweep. And use JK or D flip
flops for trigger and state control. The flip flops would be triggered and
reset by comparators.

The clock input to the flip flop will be feed from a comparator and used as
the trigger input. When the capacitor reaches a certain value (end of sweep)
a comparator would reset the flip flop. When the flip flop rests, it will
discharge the cap. After a dwell time (cap discharge period), the data input
to the flip flop will be brought high. The next clock input (trigger) will
set the flip flop and the cap will start charging which commences the sweep.

I should be able to do the sweep and triggering circuit with a dual D flip
flop, a quad comparator, and a transistor with some other components. with a
comparator and D flip flop to spare.

N9WOS wrote:
On the other hand, it does seem like a LOT of work to go to
considering that you can get a really nice used analog scope for under
$100. I have an HP1715A 200MHz 2-channel scope with nice main+delayed
sweep that I need to put on ebay (or equivalent), and I seriously
doubt I'll get $100 for it.

I have the old scope laying around so why not? And almost everything I am
using in it is salvaged parts. Including capacitors, transistors,
transformers, and the video chips in question. I will probably have under 10
dollars in it after I am done. (That includes the $5 I paid for the scope a
few years ago.)

If you do go to all the trouble to revamp your Eico (maybe the same
scope model I started with almost 50 years ago??!!), I do hope you'll
put a good triggered sweep in it, and not just a "horizontal
oscillator." It's not too difficult to make a very respectable linear
sweep with op amps and fast comparators these days. For triggering,
for example, have a look at the Analog Devices ADCMP02 and its
siblings: fast, built-in hysteresis... I'm using one in the input of
a counter, and it goes to 100MHz easily for low level inputs, and to
well over 200MHz if I jack up the input a bit.

For the sweep, I was planning on using a transistor as a constant current
source to charge the capacitor for a nice linear sweep. And use JK or D flip
flops for trigger and state control. The flip flops would be triggered and
reset by comparators.

The clock input to the flip flop will be feed from a comparator and used as
the trigger input. When the capacitor reaches a certain value (end of sweep)
a comparator would reset the flip flop. When the flip flop rests, it will
discharge the cap. After a dwell time (cap discharge period), the data input
to the flip flop will be brought high. The next clock input (trigger) will
set the flip flop and the cap will start charging which commences the sweep.

I should be able to do the sweep and triggering circuit with a dual D flip
flop, a quad comparator, and a transistor with some other components. with a
comparator and D flip flop to spare.



Thanks for posting this interesting topic! I thought no one else cared
about CRTs. I've been messing with scopes and CRTs a long time. It's
always fun. I always see the 3" tubes built into people's projects,
running on 1-3KV. They are never very bright. I like the larger 5" ones
of recent manufacture (recent being a relative term). Half the fun is
doing it yourself. Even the HV supply is a fun project when you get into
serious scope tubes. In general, it will take at least 6KV to get a real
bright one going. - but the tube has to be able to take it! In case
anyone needs a good CRT, there are many here, for cheap, and quite a few
HP or TEK types:
http://www.magnetosphere.net/

PJ

Hey OM:

I remember my days at RETS electronics schools we had a instructor his
name was Armstrong the cousin of the inventor of FM. In his class you
got to build a 100Mhz tube only scope, you bought all the parts from
him, nice scam, including the CRT, he made the CRT's himself from
scratch. I didn't get him as an instructor so I missed out on a great
deal. We got to play with computers the relay tube type. When I
graduated, Armstrong finally broke down and put an FET into his scope.
I have to add that his scope was flat + - 3db all the way out to
100MHZ. dual trace with a triggered sweep.
Don't forget your astigmatism

73 OM
de n8zu


On Feb 28, 3:16*pm, "N9WOS" wrote:
The hardest part to build on an oscilloscope (while still maintaining
bandwidth) is the final deflection amplifiers.
Something had occurred to me a little while ago. There is a readily
available supply of chips that will do the job in a specific product. That
product is CRT SVGA monitors.

The video amplifiers that drive the grids of the electron guns.

Normal supply voltage is 80 to 100V.
Normal output swing is 10 to 75V or 10 to 95V
Normal signal input voltage 0 to 5V
They are capable of driving 8pf loads which is close to the value of a
deflection plate.
Output response times are 10us or faster. Some of the newest are down to
2.5us.
Corresponding bandwidth is DC to 50 or 200MHZ
They normally come in a self contained package that contains three
amplifiers.

Since most scope most scope tubes take less than 80 volts per plate for full
deflection (all the ones i have seen), then they make a great deflection
amplifier.

They would be great for upgrading an old tube type scope up to a solid state
100Mhz unit. The rest of the amplifier chain would be easy to built. They
would also be great for building a scope from scratch. Remember to keep lead
lengths short, and use shielded interconnect cables, because at 100Mhz it
will oscillate if it can hear it's self.

That is what I am currently doing to an old EICO scope. I already have the
vertical amplifier in place. It uses two sections of a VPS10S. For The
horizontal amplifier I am going with discrete components, because it doesn't
need to be as high of bandwidth. I have got the horizontal amplifier built,
and I just need to get it mounted.

The next stage is the amplifier stages between the input, and deflection
circuits. They are all going to be 5V rail to ground. There will be a couple
+10 and -5V point in the circuit for level transition, and signal
inversions. And I have to build the horizontal oscillator and triggering
circuits.

The only thing that has been bugging me is, what can I do with the old tube
sockets. I have been trying *to think of some tube circuits I could use.. The
only things I could think of is a high impedance input buffer. Or use a 6AL5
as a wide band RF rectifier. Or I could put in a sweep oscillator and a IF
amplifier and make a rudimentary spectrum analyzer.

"raypsi" wrote in message
...
"I remember my days at RETS electronics schools we had a instructor his
name was Armstrong the cousin of the inventor of FM. In his class you
got to build a 100Mhz tube only scope, you bought all the parts from
him, nice scam, including the CRT, he made the CRT's himself from
scratch."

Doesn't sound like a scam at all unless he was seriously inflating the prices
of the parts compared to what you'd normally get them for. Indeed, assuming
the prices were in-line, it sounds like a great dead, since especailly for
something like the CRT, he's the one warrantying it and you know exactly where
the guy works. :-)