Agilent 10076B osc probe for sale

Hi, anyone interested? I will take $125 (Paypal). Bought it years ago for some project and used it twice. Its a passive probe for high voltage (100:1). Specs are on interweb. Can upload pic later.

Regards,
Scot

Sorry for spam but kids swim lessons are not cheap!

 

[John Larkin]

On Tue, 9 Jul 2019 16:12:19 -0700 (PDT), scot@jekdesigngroup.com
wrote:

Hi, anyone interested? I will take $125 (Paypal). Bought it years ago for some project and used it twice. Its a passive probe for high voltage (100:1). Specs are on interweb. Can upload pic later.

Regards,
Scot

Sorry for spam but kids swim lessons are not cheap!

The Rigol 100:1 probe is cheap, but I'm frying them scoping a 1KV, 5
MHz pulse.

https://www.dropbox.com/s/1w83cehwdlafp72/IR_0123.jpg?raw=1

I bet the parts inside are really hot.

I wonder if the Agilent has better behavior at combined high voltage
and high frequency.

One kid costs about as much as ten Porsches. Or 50.





--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[John Larkin]

On 9 Jul 2019 18:12:02 -0700, Winfield Hill <winfieldhill@yahoo.com>
wrote:

John Larkin wrote...

The Rigol 100:1 probe is cheap, but I'm frying
them scoping a 1KV, 5 MHz pulse.

My fast HV pulse stage has a 50-ohm Zout, so
it will accurately back-terminated drive a
segment of 50-ohm coax, w/o reflections, etc.
This means I can connect it to a 100-watt
20dB RF power attenuator, and look at the
output of that, with a final 50-ohm load.
Or another attenuator. No fried probes.

My new pulser drives a capacitive load, a Pockels cell, directly, with
no coax... just very short wires. So I need a hi-Z probe.

I built a 200:1 signal pickoff into the pulser, to drive a 50 ohm
scope, but we need to test that in production, to make sure it's
really 200:1. I guess we'll verify the internal pickoff circuit at
much lower pulse frequency, so's to not fry the probe.

The pickoff was hard to get to work. It works pretty well on the rev C
board. Our mantra is "get it right the first time" and C is not first.

Tek used to make an enormous HV probe that had to be kept full of
liquid Freon.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Winfield Hill]

John Larkin wrote...

The Rigol 100:1 probe is cheap, but I'm frying
them scoping a 1KV, 5 MHz pulse.

My fast HV pulse stage has a 50-ohm Zout, so
it will accurately back-terminated drive a
segment of 50-ohm coax, w/o reflections, etc.
This means I can connect it to a 100-watt
20dB RF power attenuator, and look at the
output of that, with a final 50-ohm load.
Or another attenuator. No fried probes.


--
Thanks,
- Win

 

[Tom Gardner]

On 10/07/19 02:24, John Larkin wrote:

Tek used to make an enormous HV probe that had to be kept full of
liquid Freon.

The 6015 was 13kV without freon, 20kV with.
The 6015A replaces the freon with silicone
Derate voltages above ~100kHz

 

[Winfield Hill]

John Larkin wrote...

My new pulser drives a capacitive load, a Pockels cell, directly,
with no coax... just very short wires. So I need a hi-Z probe.

I built a 200:1 signal pickoff into the pulser, to drive a 50 ohm
scope, but we need to test that in production, to make sure it's
really 200:1. I guess we'll verify the internal pickoff circuit at
much lower pulse frequency, so's to not fry the probe.

The pickoff was hard to get to work. It works pretty well on the rev C
board. Our mantra is "get it right the first time" and C is not first.

My approach to that scene is to use a purely capacitive
divider, which I opamp buffer, followed by an adjustable
CAL trimmer and another opamp. You use a small upper
capacitance, so even with shielding it's hard to account
for minor strays. So I allow 25% or more CAL adjustment.
The CAL operation is easily and accurately done with a
low-Z low-voltage source feeding the HV output.


--
Thanks,
- Win

 

[Winfield Hill]

John Larkin wrote...

Here's my pickoff.

https://www.dropbox.com/s/es2unxg5wllrte0/T850_Pickoff.jpg?raw=1

What 1pF part do you use for C43?
Also, how is it 200:1 with C40 = 68pF?


--
Thanks,
- Win

 

[Phil Hobbs]

On 7/10/19 10:09 AM, John Larkin wrote:

On 10 Jul 2019 06:17:34 -0700, Winfield Hill <winfieldhill@yahoo.com
wrote:

John Larkin wrote...

My new pulser drives a capacitive load, a Pockels cell, directly,
with no coax... just very short wires. So I need a hi-Z probe.

I built a 200:1 signal pickoff into the pulser, to drive a 50 ohm
scope, but we need to test that in production, to make sure it's
really 200:1. I guess we'll verify the internal pickoff circuit at
much lower pulse frequency, so's to not fry the probe.

The pickoff was hard to get to work. It works pretty well on the rev C
board. Our mantra is "get it right the first time" and C is not first.

My approach to that scene is to use a purely capacitive
divider, which I opamp buffer, followed by an adjustable
CAL trimmer and another opamp. You use a small upper
capacitance, so even with shielding it's hard to account
for minor strays. So I allow 25% or more CAL adjustment.
The CAL operation is easily and accurately done with a
low-Z low-voltage source feeding the HV output.

Here's my pickoff.

https://www.dropbox.com/s/es2unxg5wllrte0/T850_Pickoff.jpg?raw=1

It's reasonably accurate but I could add a cal pot or trim cap for rev
D. In real life, my customers will trim the HV pulse amplitude for
optical effect, so the voltage monitor doesn't need to be very
accurate.

Another useful thing for Rev D would be a DC bias adjustment to control
for the Pockels cell's large static birefringence. That way you can go
from linear polarization to linear polarization.

J1-J2-J3 is a 3-pole Phoenix barrier strip. It had so much dielectric
loss that it was getting hot and costing me several watts. The fix is
to rip all the metal bits out of the unused J2 section.

https://www.dropbox.com/s/x0mw9j68eywoatz/T850_Phoenix.JPG?raw=1

The parts to the right are the beginnings of the pickoff.

I once had a circuit oscillate hard enough that a ferrite bead melted
its way through a minigrabber clip. (I don't build manly stuff like
your drivers.)

Cheers

Phil Hobbs



--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[John Larkin]

On 10 Jul 2019 06:17:34 -0700, Winfield Hill <winfieldhill@yahoo.com>
wrote:

John Larkin wrote...

My new pulser drives a capacitive load, a Pockels cell, directly,
with no coax... just very short wires. So I need a hi-Z probe.

I built a 200:1 signal pickoff into the pulser, to drive a 50 ohm
scope, but we need to test that in production, to make sure it's
really 200:1. I guess we'll verify the internal pickoff circuit at
much lower pulse frequency, so's to not fry the probe.

The pickoff was hard to get to work. It works pretty well on the rev C
board. Our mantra is "get it right the first time" and C is not first.

My approach to that scene is to use a purely capacitive
divider, which I opamp buffer, followed by an adjustable
CAL trimmer and another opamp. You use a small upper
capacitance, so even with shielding it's hard to account
for minor strays. So I allow 25% or more CAL adjustment.
The CAL operation is easily and accurately done with a
low-Z low-voltage source feeding the HV output.

Here's my pickoff.

https://www.dropbox.com/s/es2unxg5wllrte0/T850_Pickoff.jpg?raw=1

It's reasonably accurate but I could add a cal pot or trim cap for rev
D. In real life, my customers will trim the HV pulse amplitude for
optical effect, so the voltage monitor doesn't need to be very
accurate.

J1-J2-J3 is a 3-pole Phoenix barrier strip. It had so much dielectric
loss that it was getting hot and costing me several watts. The fix is
to rip all the metal bits out of the unused J2 section.

https://www.dropbox.com/s/x0mw9j68eywoatz/T850_Phoenix.JPG?raw=1

The parts to the right are the beginnings of the pickoff.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Joerg]

On 2019-07-09 16:58, John Larkin wrote:

On Tue, 9 Jul 2019 16:12:19 -0700 (PDT), scot@jekdesigngroup.com
wrote:

Hi, anyone interested? I will take $125 (Paypal). Bought it years ago for some project and used it twice. Its a passive probe for high voltage (100:1). Specs are on interweb. Can upload pic later.

Regards,
Scot

Sorry for spam but kids swim lessons are not cheap!

The Rigol 100:1 probe is cheap, but I'm frying them scoping a 1KV, 5
MHz pulse.

https://www.dropbox.com/s/1w83cehwdlafp72/IR_0123.jpg?raw=1

I bet the parts inside are really hot.

I wonder if the Agilent has better behavior at combined high voltage
and high frequency.

One kid costs about as much as ten Porsches. Or 50.

Or a Jeep Rubicon. And then builds up a savings account

--
Regards, Joerg

http://www.analogconsultants.com/

 

[John Larkin]

On Wed, 10 Jul 2019 10:34:37 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

On 7/10/19 10:09 AM, John Larkin wrote:
On 10 Jul 2019 06:17:34 -0700, Winfield Hill <winfieldhill@yahoo.com
wrote:

John Larkin wrote...

My new pulser drives a capacitive load, a Pockels cell, directly,
with no coax... just very short wires. So I need a hi-Z probe.

I built a 200:1 signal pickoff into the pulser, to drive a 50 ohm
scope, but we need to test that in production, to make sure it's
really 200:1. I guess we'll verify the internal pickoff circuit at
much lower pulse frequency, so's to not fry the probe.

The pickoff was hard to get to work. It works pretty well on the rev C
board. Our mantra is "get it right the first time" and C is not first.

My approach to that scene is to use a purely capacitive
divider, which I opamp buffer, followed by an adjustable
CAL trimmer and another opamp. You use a small upper
capacitance, so even with shielding it's hard to account
for minor strays. So I allow 25% or more CAL adjustment.
The CAL operation is easily and accurately done with a
low-Z low-voltage source feeding the HV output.

Here's my pickoff.

https://www.dropbox.com/s/es2unxg5wllrte0/T850_Pickoff.jpg?raw=1

It's reasonably accurate but I could add a cal pot or trim cap for rev
D. In real life, my customers will trim the HV pulse amplitude for
optical effect, so the voltage monitor doesn't need to be very
accurate.

Another useful thing for Rev D would be a DC bias adjustment to control
for the Pockels cell's large static birefringence. That way you can go
from linear polarization to linear polarization.

The pulser is AC coupled into the Pockels. I hate that because we get
a negative baseline shift at higher pulse rates, which costs me a bit
of peak voltage. My prime customer insists on AC coupling so that
there is no long-term DC voltage across the crystal.

It's a femtosecond pulse picker. The sine-squared transfer function of
the Pockels hides all sorts of sins since we only care about the flat
bottoms and tops of the pulses.

No offense to an acknowledged great circuit designer, but it's a
continuous struggle to get the physicists to understand concepts like
inductance and grounding. The horrors I have seen!

J1-J2-J3 is a 3-pole Phoenix barrier strip. It had so much dielectric
loss that it was getting hot and costing me several watts. The fix is
to rip all the metal bits out of the unused J2 section.

https://www.dropbox.com/s/x0mw9j68eywoatz/T850_Phoenix.JPG?raw=1

The parts to the right are the beginnings of the pickoff.

I once had a circuit oscillate hard enough that a ferrite bead melted
its way through a minigrabber clip. (I don't build manly stuff like
your drivers.)

I recently read a book about Tesla, which inspired me to grab the 1KV
output and see if I'd get shocked. I confirmed what Nikoli said, no
shock but a burn is possible.


--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[John Larkin]

On 10 Jul 2019 07:27:14 -0700, Winfield Hill <winfieldhill@yahoo.com>
wrote:

John Larkin wrote...

Here's my pickoff.

https://www.dropbox.com/s/es2unxg5wllrte0/T850_Pickoff.jpg?raw=1

What 1pF part do you use for C43?

MFR1 VISHAY VJ1111D1R0BXRAJ
MFR2 DIGIKEY 720-1436-1-ND

It's an 1111. I never knew there were 1111 caps!


> Also, how is it 200:1 with C40 = 68pF?

Beats me; that's what works.

It's 200:1 when loaded with a 50 ohm scope. There are capacitive
strays, and the source of the mosfet is not zero ohms.



--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[Phil Hobbs]

On 7/10/19 11:41 AM, John Larkin wrote:

On Wed, 10 Jul 2019 10:34:37 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

On 7/10/19 10:09 AM, John Larkin wrote:
On 10 Jul 2019 06:17:34 -0700, Winfield Hill <winfieldhill@yahoo.com
wrote:

John Larkin wrote...

My new pulser drives a capacitive load, a Pockels cell, directly,
with no coax... just very short wires. So I need a hi-Z probe.

I built a 200:1 signal pickoff into the pulser, to drive a 50 ohm
scope, but we need to test that in production, to make sure it's
really 200:1. I guess we'll verify the internal pickoff circuit at
much lower pulse frequency, so's to not fry the probe.

The pickoff was hard to get to work. It works pretty well on the rev C
board. Our mantra is "get it right the first time" and C is not first.

My approach to that scene is to use a purely capacitive
divider, which I opamp buffer, followed by an adjustable
CAL trimmer and another opamp. You use a small upper
capacitance, so even with shielding it's hard to account
for minor strays. So I allow 25% or more CAL adjustment.
The CAL operation is easily and accurately done with a
low-Z low-voltage source feeding the HV output.

Here's my pickoff.

https://www.dropbox.com/s/es2unxg5wllrte0/T850_Pickoff.jpg?raw=1

It's reasonably accurate but I could add a cal pot or trim cap for rev
D. In real life, my customers will trim the HV pulse amplitude for
optical effect, so the voltage monitor doesn't need to be very
accurate.

Another useful thing for Rev D would be a DC bias adjustment to control
for the Pockels cell's large static birefringence. That way you can go
from linear polarization to linear polarization.

The pulser is AC coupled into the Pockels. I hate that because we get
a negative baseline shift at higher pulse rates, which costs me a bit
of peak voltage. My prime customer insists on AC coupling so that
there is no long-term DC voltage across the crystal.

Okay, so they must have a polarization compensator in there someplace.
It's probably a fibre/grating pulse compressor, so you can do that with
the usual three-paddle thing in SMF based on bend birefringence, e.g.
<https://www.thorlabs.com/newgrouppage9.cfm?objectgroup_ID=343>.

You can make any polarization state from any other using two
quarter-wave plates, but they have to be accurate. With the
three-paddle model, you take one turn round the outside ring and two
round the middle one, making two quarter- and one half-wave retarder.
That doesn't have to be as accurate, and besides, once you have the
polarization linear you can rotate it using the middle paddle. That's
much harder with a two-paddle controller.

It's a femtosecond pulse picker. The sine-squared transfer function of
the Pockels hides all sorts of sins since we only care about the flat
bottoms and tops of the pulses.

Fun.

No offense to an acknowledged great circuit designer, but it's a
continuous struggle to get the physicists to understand concepts like
inductance and grounding. The horrors I have seen!

Don't talk about my customers like that.

Physicists also think that lowpass filters chop off everything above the
3-dB point.

Of course they can also bring a useful sort of first-principles approach
sometimes.

J1-J2-J3 is a 3-pole Phoenix barrier strip. It had so much dielectric
loss that it was getting hot and costing me several watts. The fix is
to rip all the metal bits out of the unused J2 section.

https://www.dropbox.com/s/x0mw9j68eywoatz/T850_Phoenix.JPG?raw=1

The parts to the right are the beginnings of the pickoff.

I once had a circuit oscillate hard enough that a ferrite bead melted
its way through a minigrabber clip. (I don't build manly stuff like
your drivers.)

I recently read a book about Tesla, which inspired me to grab the 1KV
output and see if I'd get shocked. I confirmed what Nikoli said, no
shock but a burn is possible.

RF burns are super nasty, too. (Next time try it on a water balloon
with an ammeter in series to ground.)

Cheers

Phil Hobbs

(who started building circuits well before taking his first physics course)


--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

 

[John Larkin]

On Wed, 10 Jul 2019 07:12:55 -0700, Joerg <news@analogconsultants.com>
wrote:

On 2019-07-09 16:58, John Larkin wrote:
On Tue, 9 Jul 2019 16:12:19 -0700 (PDT), scot@jekdesigngroup.com
wrote:

Hi, anyone interested? I will take $125 (Paypal). Bought it years ago for some project and used it twice. Its a passive probe for high voltage (100:1). Specs are on interweb. Can upload pic later.

Regards,
Scot

Sorry for spam but kids swim lessons are not cheap!

The Rigol 100:1 probe is cheap, but I'm frying them scoping a 1KV, 5
MHz pulse.

https://www.dropbox.com/s/1w83cehwdlafp72/IR_0123.jpg?raw=1

I bet the parts inside are really hot.

I wonder if the Agilent has better behavior at combined high voltage
and high frequency.

One kid costs about as much as ten Porsches. Or 50.


Or a Jeep Rubicon. And then builds up a savings account

Cornell alone cost as much as a fleet of Rubicons.

She actually bought a Rubicon. Great in snow but otherwise an
astonishing piece of junk. Chrysler must purchase pre-stripped
fasteners and mate-once connectors.

It was hard to keep it inside one lane on the freeway. It at least
kept the driver awake.






--

John Larkin Highland Technology, Inc

lunatic fringe electronics

 

[John Larkin]

On Wed, 10 Jul 2019 12:03:32 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

RF burns are super nasty, too. (Next time try it on a water balloon
with an ammeter in series to ground.)

As if. I don't want water all over my circuit and my bench and my
jeans.


--

John Larkin Highland Technology, Inc
picosecond timing precision measurement

jlarkin att highlandtechnology dott com
http://www.highlandtechnology.com

 

[Winfield Hill]

John Larkin wrote...

I recently read a book about Tesla, which inspired me
to grab the 1KV output and see if I'd get shocked.
I confirmed what Nikoli said, no shock but a burn is
possible.

What's more you can get bad RF burns though insulation,
or even by just getting your hand close. And the burns
can go deep into the dermal layers, ouch! But at least
they come pre-sterilized.


--
Thanks,
- Win

 

[Joerg]

On 2019-07-10 08:48, John Larkin wrote:

On Wed, 10 Jul 2019 07:12:55 -0700, Joerg <news@analogconsultants.com
wrote:

On 2019-07-09 16:58, John Larkin wrote:
On Tue, 9 Jul 2019 16:12:19 -0700 (PDT), scot@jekdesigngroup.com
wrote:

Hi, anyone interested? I will take $125 (Paypal). Bought it years ago for some project and used it twice. Its a passive probe for high voltage (100:1). Specs are on interweb. Can upload pic later.

Regards,
Scot

Sorry for spam but kids swim lessons are not cheap!

The Rigol 100:1 probe is cheap, but I'm frying them scoping a 1KV, 5
MHz pulse.

https://www.dropbox.com/s/1w83cehwdlafp72/IR_0123.jpg?raw=1

I bet the parts inside are really hot.

I wonder if the Agilent has better behavior at combined high voltage
and high frequency.

One kid costs about as much as ten Porsches. Or 50.


Or a Jeep Rubicon. And then builds up a savings account

Cornell alone cost as much as a fleet of Rubicons.

She actually bought a Rubicon. Great in snow but otherwise an
astonishing piece of junk. Chrysler must purchase pre-stripped
fasteners and mate-once connectors.

It was hard to keep it inside one lane on the freeway. It at least
kept the driver awake.

After my dad's Chrysler and having my own later our family never bought
another one. I never will. The quality simply wasn't there. My Chrysler
rusted out so badly after only six years that it was unable to pass
German roadworthiness checks. Plus lots of other issues. The Audi I
bought afterwards never had a speck of rust or other issues in 30 years.
It probably still runs fine but that was the time where the man I sold
it to had to give up driving because of his advanced age and health
problems. Also, it is much bigger than the Chrysler, can haul half a ton
of equipment, is faster _and_ has better gas mileage.

--
Regards, Joerg

http://www.analogconsultants.com/