Maker Pro
Maker Pro

homebrew sampling oscilloscope

  • Thread starter Jorgen Lund-Nielsen
  • Start date
J

Jorgen Lund-Nielsen

Jan 1, 1970
0
Hello,

a few years ago, in a magazine, there was a "design idea" to build a
homebrew sampling oscilloscope with about 1 GHz bandwidth.
Diode bridged input circuit, timing and so on.
I lostthat article and Google find
not the original article, only notes about it.
anyone have a copy of it and will let me participate?

Thanks,

Jorgen
dj0ud
 
J

John Larkin

Jan 1, 1970
0
Have not looked to find a schematic.
Basically, a full wave bridge is DC reverse biased, and a floating
winding from a pulse transformer is used to turn the bridge on for a
short period of time (the sampling time).
One of the "AC" legs is the signal input, and the other is the sampled
output.
For decent operation, all diodes should be the same type, and for
critical operation should be matched.
They should have a fast turnoff and minimal stored charge; maybe PIN
diodes would be useful her, but good fast diodes certainly can be used.
Almost all sampling systems are 50 ohm input, but higher impedances
are possible (maybe 10K or so?) at reduced bandwidth.


A very good book is "Sampling Oscilloscope Circuits", part of the
Tektronix Concepts Books series. They show up on ebay now and then,
and the bookfinder services like Alibris can usually get them.

The oldest samplers (1950's) used point-contact silicon or germanium
radar diodes (1N34 types) and later Tek used point-contact GaAs. The
original Lumatron and Tek (type N) samplers used an avalanche
transistor pulser and a single-diode sampler; nasty.

Lately, everybody uses tiny schottkies with capacitances in the 0.05
pF range. Most modern samplers use two diodes in a half-bridge,
although some of the Tek stuff is a 6-diode traveling-wave sampler.
LeCroy, ever bizarre, has a 1-diode sampler at 60 GHz. I did a
half-bridge for fun a few years ago, using commercial parts
(step-recovery diode, shorted transmission lines as shaper, SOT-23
dual schottky... basicly a crib of a Tek S2 head) and got about 5 GHz
bandwidth. The timebase is more trouble than the sampler.

I saw one magazine article that used an avalanche transistor pulser
and a 4-diode bridge, at about 1 GHz; can't locate it now. The Tek
7S14 works this way.

Alex did a beautiful job of scanning the Tek 1S1 manual... all 24 megs
worth.

http://www.dfpresource.org/pdf/tek1s1_manual.pdf


John
 
R

Robert Baer

Jan 1, 1970
0
John said:
A very good book is "Sampling Oscilloscope Circuits", part of the
Tektronix Concepts Books series. They show up on ebay now and then,
and the bookfinder services like Alibris can usually get them.

The oldest samplers (1950's) used point-contact silicon or germanium
radar diodes (1N34 types) and later Tek used point-contact GaAs. The
original Lumatron and Tek (type N) samplers used an avalanche
transistor pulser and a single-diode sampler; nasty.

Lately, everybody uses tiny schottkies with capacitances in the 0.05
pF range. Most modern samplers use two diodes in a half-bridge,
although some of the Tek stuff is a 6-diode traveling-wave sampler.
LeCroy, ever bizarre, has a 1-diode sampler at 60 GHz. I did a
half-bridge for fun a few years ago, using commercial parts
(step-recovery diode, shorted transmission lines as shaper, SOT-23
dual schottky... basicly a crib of a Tek S2 head) and got about 5 GHz
bandwidth. The timebase is more trouble than the sampler.

I saw one magazine article that used an avalanche transistor pulser
and a 4-diode bridge, at about 1 GHz; can't locate it now. The Tek
7S14 works this way.

Alex did a beautiful job of scanning the Tek 1S1 manual... all 24 megs
worth.

http://www.dfpresource.org/pdf/tek1s1_manual.pdf

John

*GACK!* That is rather nasty on POTS; is ther a way to get the file by
sending for a CD of it?
 
J

John Larkin

Jan 1, 1970
0
*GACK!* That is rather nasty on POTS; is ther a way to get the file by
sending for a CD of it?

You could ask Alex; he posts here as AE.

If I get a minute later today, I'll post some sampler front-end
schematics to a.b.s.e.

John
 
A

A E

Jan 1, 1970
0
George R. Gonzalez said:
There's a chicken-egg problem here... If the thing doesnt work perfectly
from the get-go, what are you going to use to debug it?

Well, I guess you could borrow or rent an even faster sampling scope.....

Nah, you rip the clock synth out of a PC and assume the rising edges are
sub-nano and clean. :)
 
R

Robert Baer

Jan 1, 1970
0
John said:
You could ask Alex; he posts here as AE.

If I get a minute later today, I'll post some sampler front-end
schematics to a.b.s.e.

John

He was kind enough to e-mail me a link to a 15 page PDF circuit
description by Tektronix; took about 5.5 minutes to fetch - much better
than an hour *plus*.
 
R

Robert Baer

Jan 1, 1970
0
George R. Gonzalez said:
There's a chicken-egg problem here... If the thing doesnt work perfectly
from the get-go, what are you going to use to debug it?

Well, I guess you could borrow or rent an even faster sampling scope.....

Regards,

George

Actually, the sampling speed - or more accurately, the bandwidth, is
*easy* to determine.....
Likewise effective signal sensitivity (eg: mV/cm).
 
Top