¿ªÔÆÌåÓý

ctrl + shift + ? for shortcuts
© 2025 Groups.io
  1. Nanovna-Users
  2. Topics

Has anybody measured electron tubes

 

I read about first TV sets and their tubes, thats history. And began to wonder if anyone has measured tubes with VNAs. Do you get meaningfull results. Connecting something with 200V supply voltage to NanoVna must be interesting.
 

Testing tubes cannot be done with the nanoVNA.

I have tested plenty of tubes, various types etc, using a tube tester designed to provide appropriate voltages to the correct pins to bias the tubes and measure emission.

Should you need to test tubes there are testers that can be found either complete or as kits to build.

Tim VK4TIM.

Sent from my Iphone.

On 17 Jan 2023, at 12:29, Leif <leif.michaelsson@...> wrote:

?I read about first TV sets and their tubes, thats history. And began to wonder if anyone has measured tubes with VNAs. Do you get meaningfull results. Connecting something with 200V supply voltage to NanoVna must be interesting.
 

The Wrong tool for that Job!??
Kent

On Monday, January 16, 2023 at 08:29:46 PM CST, Leif M <leif.michaelsson@...> wrote:

I read about first TV sets and their tubes, thats history. And began to wonder if anyone has measured tubes with VNAs. Do you get meaningfull results. Connecting something with 200V supply voltage to NanoVna must be interesting.
 

Tube tester is a different tool, I agree. But has anybody even measured s-params of a tube. And all without burning the VNA.
 

You don't seem to understand the NanoVNA is in NO Way a tube tester.
Clyde Lambert KC7BJE
 

You would do it the same way you would measure a high voltage MOSFET but
since
you are asking the question, I would not do it considering the voltages
involved.

On Tue, 17 Jan 2023 at 03:29, Leif M <leif.michaelsson@...> wrote:

I read about first TV sets and their tubes, thats history. And began to
wonder if anyone has measured tubes with VNAs. Do you get meaningfull
results. Connecting something with 200V supply voltage to NanoVna must be
interesting.
 

I have never seen any manufacturer publish s-parameters for tubes. And the
design techniques are different. The parts and methods used in biasing and
matching tubes are completely different than those used for FET's and
BJT's. The frequency ranges used for most commonly-available tubes do not
require s-parameters. There are very few tube amps around that operate
above 1296 MHz.

I see no reason to measure a tube on a VNA.

Zack W9SZ

On Mon, Jan 16, 2023 at 8:29 PM Leif M <leif.michaelsson@...> wrote:

I read about first TV sets and their tubes, thats history. And began to
wonder if anyone has measured tubes with VNAs. Do you get meaningfull
results. Connecting something with 200V supply voltage to NanoVna must be
interesting.
 

"The frequency ranges used for most commonly-available tubes do not
require s-parameters. There are very few tube amps around that operate
above 1296 MHz."

I remember back in the day, using the giant round-screen tube oscilloscope
on the rollaround cart to measure high frequencies, because transistors
weren't fast enough. How things have changed. I can't remember if that
was before FETs came on the scene.
 

Most likely. In the really early days, there weren't any transistors that
worked above HF. Then some VHF transistors came along. There were no
high-frequency FET's in the early days. Now we have some FET's that work up
to 40 GHz. All my microwave designs using FET's have been on microstrip
transmission line printed on somewhat exotic substrates like Duroid 5880. I
have never seen a tube design that used microstrip.

Zack W9SZ

On Tue, Jan 17, 2023 at 6:58 AM Allasso <kevinhowjones@...> wrote:

"The frequency ranges used for most commonly-available tubes do not
require s-parameters. There are very few tube amps around that operate
above 1296 MHz."

I remember back in the day, using the giant round-screen tube oscilloscope
on the rollaround cart to measure high frequencies, because transistors
weren't fast enough. How things have changed. I can't remember if that
was before FETs came on the scene.
 

On 1/17/23 2:21 AM, Zack Widup wrote:
I have never seen any manufacturer publish s-parameters for tubes.
Pretty much every manufacturer of Traveling Wave Tubes publishes S-parameters - In particular one tends to be interested in the S12 (reverse gain) at the second harmonic of the intended frequency band.

The other thing one tends to be very interested in is the group delay across the band of operation, which is derived from data gathered with a VNA.

I think you're thinking of gridded tubes like tri-, tetr-, and pent-odes.




And the
design techniques are different. The parts and methods used in biasing and
matching tubes are completely different than those used for FET's and
BJT's. The frequency ranges used for most commonly-available tubes do not
require s-parameters. There are very few tube amps around that operate
above 1296 MHz.
I wouldn't be surprised if there are more microwave tube amps today than HF band tube amps. If you want power, tubes are where it's at.


I see no reason to measure a tube on a VNA.
Zack W9SZ
On Mon, Jan 16, 2023 at 8:29 PM Leif M <leif.michaelsson@...> wrote:

I read about first TV sets and their tubes, thats history. And began to
wonder if anyone has measured tubes with VNAs. Do you get meaningfull
results. Connecting something with 200V supply voltage to NanoVna must be
interesting.

That's solved with just a DC block. A bigger issue is not blowing up the VNA with the output power.
(and for the NanoVNA, the fact that the test signal is a square wave)
 

Yes, that's true. But what is the probability that the OP wants to test a
TWT?

Zack W9SZ

On Tue, Jan 17, 2023 at 8:21 AM Jim Lux <jimlux@...> wrote:

On 1/17/23 2:21 AM, Zack Widup wrote:
I have never seen any manufacturer publish s-parameters for tubes.
Pretty much every manufacturer of Traveling Wave Tubes publishes
S-parameters - In particular one tends to be interested in the S12
(reverse gain) at the second harmonic of the intended frequency band.

The other thing one tends to be very interested in is the group delay
across the band of operation, which is derived from data gathered with a
VNA.

I think you're thinking of gridded tubes like tri-, tetr-, and pent-odes.




And the
design techniques are different. The parts and methods used in biasing
and
matching tubes are completely different than those used for FET's and
BJT's. The frequency ranges used for most commonly-available tubes do not
require s-parameters. There are very few tube amps around that operate
above 1296 MHz.
I wouldn't be surprised if there are more microwave tube amps today than
HF band tube amps. If you want power, tubes are where it's at.



I see no reason to measure a tube on a VNA.

Zack W9SZ

On Mon, Jan 16, 2023 at 8:29 PM Leif M <leif.michaelsson@...>
wrote:

I read about first TV sets and their tubes, thats history. And began to
wonder if anyone has measured tubes with VNAs. Do you get meaningfull
results. Connecting something with 200V supply voltage to NanoVna must
be
interesting.

That's solved with just a DC block. A bigger issue is not blowing up
the VNA with the output power.
(and for the NanoVNA, the fact that the test signal is a square wave)
Jim Whartenby
 

Zack
Published S-parameters for a W-band TWT, Figure 5.
See:?
Regards,
Jim

Logic: Method used to arrive at the wrong conclusion, with confidence.? Murphy

-----Original Message-----
From: Zack Widup <w9sz.zack@...>
To: [email protected]
Sent: Tue, Jan 17, 2023 4:21 am
Subject: Re: [nanovna-users] Has anybody measured electron tubes

I have never seen any manufacturer publish s-parameters for tubes. And the
design techniques are different. The parts and methods used in biasing and
matching tubes are completely different than those used for FET's and
BJT's. The frequency ranges used for most commonly-available tubes do not
require s-parameters. There are very few tube amps around that operate
above 1296 MHz.

I see no reason to measure a tube on a VNA.

Zack W9SZ

On Mon, Jan 16, 2023 at 8:29 PM Leif M <leif.michaelsson@...> wrote:

I read about first TV sets and their tubes, thats history. And began to
wonder if anyone has measured tubes with VNAs. Do you get meaningfull
results. Connecting something with 200V supply voltage to NanoVna must be
interesting.
 

Hard to find a TWT that a Nano could test.
He's thinking more 12AX7!
(Been using TWT's for over 40 years, don't think I have ever seen one for under 1 GHz.)
Kent

On Tuesday, January 17, 2023 at 09:00:25 AM CST, Zack Widup <w9sz.zack@...> wrote:

Yes, that's true. But what is the probability that the OP wants to test a
TWT?

Zack W9SZ

On Tue, Jan 17, 2023 at 8:21 AM Jim Lux <jimlux@...> wrote:

On 1/17/23 2:21 AM, Zack Widup wrote:
I have never seen any manufacturer publish s-parameters for tubes.
Pretty much every manufacturer of Traveling Wave Tubes publishes
S-parameters - In particular one tends to be interested in the S12
(reverse gain) at the second harmonic of the intended frequency band.

The other thing one tends to be very interested in is the group delay
across the band of operation, which is derived from data gathered with a
VNA.

I think you're thinking of gridded tubes like tri-, tetr-, and pent-odes.




And the
design techniques are different. The parts and methods used in biasing
and
matching tubes are completely different than those used for FET's and
BJT's. The frequency ranges used for most commonly-available tubes do not
require s-parameters. There are very few tube amps around that operate
above 1296 MHz.
I wouldn't be surprised if there are more microwave tube amps today than
HF band tube amps.? If you want power, tubes are where it's at.



I see no reason to measure a tube on a VNA.

Zack W9SZ

On Mon, Jan 16, 2023 at 8:29 PM Leif M <leif.michaelsson@...>
wrote:

I read about first TV sets and their tubes, thats history. And began to
wonder if anyone has measured tubes with VNAs. Do you get meaningfull
results. Connecting something with 200V supply voltage to NanoVna must
be
interesting.

That's solved with just a DC block.? A bigger issue is not blowing up
the VNA with the output power.
(and for the NanoVNA, the fact that the test signal is a square wave)
 

I learned that S Parameters came into use when devices designed for higher frequencies (UHF and Microwave) had to be characterized. Also, test equipment have to be developed which could measure S Parameters.

Here is a historical perspective:

<

Mike N2MS

On 01/17/2023 11:07 AM KENT BRITAIN <wa5vjb@...> wrote:


Hard to find a TWT that a Nano could test.
He's thinking more 12AX7!
(Been using TWT's for over 40 years, don't think I have ever seen one for under 1 GHz.)
Kent
On Tuesday, January 17, 2023 at 09:00:25 AM CST, Zack Widup <w9sz.zack@...> wrote:

Yes, that's true. But what is the probability that the OP wants to test a
TWT?

Zack W9SZ
 

On 1/17/23 8:07 AM, KENT BRITAIN wrote:
Hard to find a TWT that a Nano could test.
He's thinking more 12AX7!
(Been using TWT's for over 40 years, don't think I have ever seen one for under 1 GHz.)
Kent
Plenty of L-band TWTs around (big honkin' things) and some Nanos go up to 1.5 GHz. But mostly in space applications, and if you're playing in that ballpark, you can probably afford suitable test equipment.. (Inmarsat uses L-band, for instance)


I don't think I've seen one as a low noise preamp, though, which is more level-appropriate for a Nano.


Yes, I'm sure he's thinking classic gridded power tubes. But people *do* design tube circuits these days using S-parameters.
 

A vacuum tube is a multi port device. it is no different than any other device that would be characterized by matrix methods, like Z, Y, ABCD and S. In fact if you search
the IEEE data base, you will find a number of recent papers on scattering parameter analysis and measurements of tubes. Most are large tubes for unique systems. For example, see the attached. So, if you really wanted to and desired to build the appropriate added testing elements, appropriate bias tees and so on, go for it!

tubes.JPG

 

Firstly, of course a NanoVNA can be used directly to measure some tube characteristics, such as interelectrode capacitances, without powering up the tube.

And then, I don't see why it should be impossible to use a NanoVNA to measure two-port characteristics of a tube amplifier, just like it can be used to measure them on a solid state amplifier! Appropriate tube biasing, matching, and attenuation would be required, and the measured results would be for the tube in that specific test amplifier setup, not for the bare-bones tube, but that's how such parameters are usually measured in transistors too.

My main worry would be to make sure that in the event the tube breaks into oscillation, there is enough attenuation or limiting of some sort between the tube and the NanoVNA, to avoid damage. That's no different from measuring solid state amplifiers that could develop enough power to kill the NanoVNA.
 

On 1/17/23 10:10 AM, Manfred Mornhinweg wrote:
Firstly, of course a NanoVNA can be used directly to measure some tube characteristics, such as interelectrode capacitances, without powering up the tube.
And then, I don't see why it should be impossible to use a NanoVNA to measure two-port characteristics of a tube amplifier, just like it can be used to measure them on a solid state amplifier! Appropriate tube biasing, matching, and attenuation would be required, and the measured results would be for the tube in that specific test amplifier setup, not for the bare-bones tube, but that's how such parameters are usually measured in transistors too.
My main worry would be to make sure that in the event the tube breaks into oscillation, there is enough attenuation or limiting of some sort between the tube and the NanoVNA, to avoid damage. That's no different from measuring solid state amplifiers that could develop enough power to kill the NanoVNA.
To me, the square wave test signal is the big problem with NanoVNA and wideband power amplifiers. But you know, if your application is "narrow band-ish" (< octave BW), then a band pass filter would solve the problem. (and that's something people already have to be careful about, new fancy power meters that measure "power in the signal" with a tracking filter vs "total power")

And yeah, blowing up the VNA when the amplifier oscillates.. But I'd rather spend $100 on a new NanoVNA than $100k on a new Keysight <grin>

1 - 18 of 18