Bootstrapping a Tube...

[amdx]

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100MΩ grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100MΩ, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0

                            Thanks, Mikek

 

[John Larkin]

On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.


https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek

The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

 

[amdx]

On 9/23/2022 7:05 PM, John Larkin wrote:

On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

 Just following along on an IO Group in a discussion about designing a
Q meter.
Looking at the Boonton 260A and the HP 4342A for ideas and then
someone ask about bootstrapping the tube in the 260A, so I thought I\'d
ask here.
  Several years ago I made five 6\" coils with 8, 9, 10, 12, and 13
turns per inch, just to see where
the sweet spot was. One peaked close to 1400Q.
   I wound a ferrite rod a couple days ago and it measured 1236Q at
800kHz.
I think someone questioned that number and they wound one on the same
material, and theirs peaked at 1326Q.
The material is very low loss in the AM band, R40C1.
 Also some interesting discussion on 50 to 1 injection transformers
driven by low impedance amps.
Here\'s the beginning, but you need to start at the bottom of the the
page for the first post and page up to follow.

https://groups.io/g/Test-Equipment-Design-Construction/topic/making_a_q_meter/92809585?p=Created%2C%2C%2C20%2C2%2C200%2C0&jump=1

                                 Mikek

 

[John Larkin]

On Fri, 23 Sep 2022 22:53:43 -0500, amdx <amdx@knology.net> wrote:

On 9/23/2022 7:05 PM, John Larkin wrote:
On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

 Just following along on an IO Group in a discussion about designing a
Q meter.
Looking at the Boonton 260A and the HP 4342A for ideas and then
someone ask about bootstrapping the tube in the 260A, so I thought I\'d
ask here.
  Several years ago I made five 6\" coils with 8, 9, 10, 12, and 13
turns per inch, just to see where
the sweet spot was. One peaked close to 1400Q.
   I wound a ferrite rod a couple days ago and it measured 1236Q at
800kHz.
I think someone questioned that number and they wound one on the same
material, and theirs peaked at 1326Q.
The material is very low loss in the AM band, R40C1.
 Also some interesting discussion on 50 to 1 injection transformers
driven by low impedance amps.
Here\'s the beginning, but you need to start at the bottom of the the
page for the first post and page up to follow.
https://groups.io/g/Test-Equipment-Design-Construction/topic/making_a_q_meter/92809585?p=Created%2C%2C%2C20%2C2%2C200%2C0&jump=1


                                 Mikek

When the old Boonton was designed, it was hard to generate and
precisely measure frequencies. Now it\'s easy. So resonant width is a
good way to measure Q.

 

[piglet]

On 24/09/2022 00:33, amdx wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100MΩ grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100MΩ, does
bootstrapping help that?
 Might be barking up the wrong tree.


https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0



                            Thanks, Mikek

Did you see the value of C302? I think V301 is being used in the so
called \"infinite impedance\" detector configuration. The cathode voltage
is not the buffered grid ac input but instead is a dc level representing
something between the peak and average ac input. So AC bootstrapping
that circuit not possible. I guess you could insert a whole new hi-z
buffer between tuned circuit under test and detector but in practice it
would be difficult to get it to be an improvement?

piglet

 

[amdx]

On 9/23/2022 11:20 PM, John Larkin wrote:

On Fri, 23 Sep 2022 22:53:43 -0500, amdx <amdx@knology.net> wrote:

On 9/23/2022 7:05 PM, John Larkin wrote:
On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

 Just following along on an IO Group in a discussion about designing a
Q meter.
Looking at the Boonton 260A and the HP 4342A for ideas and then
someone ask about bootstrapping the tube in the 260A, so I thought I\'d
ask here.
  Several years ago I made five 6\" coils with 8, 9, 10, 12, and 13
turns per inch, just to see where
the sweet spot was. One peaked close to 1400Q.
   I wound a ferrite rod a couple days ago and it measured 1236Q at
800kHz.
I think someone questioned that number and they wound one on the same
material, and theirs peaked at 1326Q.
The material is very low loss in the AM band, R40C1.
 Also some interesting discussion on 50 to 1 injection transformers
driven by low impedance amps.
Here\'s the beginning, but you need to start at the bottom of the the
page for the first post and page up to follow.
https://groups.io/g/Test-Equipment-Design-Construction/topic/making_a_q_meter/92809585?p=Created%2C%2C%2C20%2C2%2C200%2C0&jump=1

                                 Mikek
When the old Boonton was designed, it was hard to generate and
precisely measure frequencies. Now it\'s easy. So resonant width is a
good way to measure Q.

  Yes. I have done a lot of that, but all I had was a scope for amplitude.
So, after too many BW tests, I came up with 5 / 7 = 0.714, that is close
enough
for relative measurements using a scope. So, I would set Fr at 7 units,
adjusting either
the freq gen amplitude or scope CAL, then raise/lower the frequency
until I had 5 units.
At that the time about 30 years ago, I was sorting piezo discs with Qs
over 1200 for the highest Q.
We only used the highest Q discs to bond for our transducers.
 I ask this on another group and never got an answer,
\"Is it possible to come up with a formula or spreadsheet that would
calculate Q from 3db BW,
with the caveat that instead of using 0.707, use the number 0.7 and
normalize it to 3db?
That would be great for needle type meters. For oscilloscopes, I\'d like
to see
0.714 normalized to 3db, why because I like to peak at 7 units and
raise/lower the frequency to 5 units.
5 / 7 = 0.714.
       Mikek

 

[amdx]

On 9/24/2022 1:11 AM, piglet wrote:

On 24/09/2022 00:33, amdx wrote:
  On one of the groups I monitor, it was ask if the tube circuit
shown below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low.
(purpose is to reduce loading on the LC being measured--Q meter)
  Can you give a little detail on how you would implement it? And
what happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100MΩ grid resistor, Is that a detriment to adding bootstrapping?
  I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100MΩ, does
bootstrapping help that?
  Might be barking up the wrong tree.

 
https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0



                             Thanks, Mikek


Did you see the value of C302? I think V301 is being used in the so
called \"infinite impedance\" detector configuration. The cathode
voltage is not the buffered grid ac input but instead is a dc level
representing something between the peak and average ac input. So AC
bootstrapping that circuit not possible. I guess you could insert a
whole new hi-z buffer between tuned circuit under test and detector
but in practice it would be difficult to get it to be an improvement?

piglet

 I thought I\'d seen it referred to as an \"infinite impedance detector\",
but don\'t find it in the manual.
I do see this line in the manual and don\'t understand how transit time
loss in the voltmeter
tube causes a shunt resistance across the measuring circuit at High
Frequencies.
Can someone provide a simple explanation?

\"QVOLTMETERCONDUCTANCE.
Anotherinternalparameterwhich causestheindi-
catedQtodeviatefromeffectiveQ,atbothvery low
andveryhighfrequencies, is theinputconductanceof
theQvoltmeter circuit.Atvery low frequenciesthis
conductance consistsofa 100megohmgridleak resistor
inparallelwiththeinternallossesofthevacuum tube.
Atveryhighfrequenciesthetransittime lossinthevolt-
metertubeshuntstheresonatingcapacitorandintro-
duces ashuntresistance acrossthemeasuring circuit\"

Someone has developed a single fet circuit that does an OK job.
I always wanted to try this circuit,
> https://www.dropbox.com/s/wvwkw2iapf4ufd7/High%20impedance%20input.jpg?dl=0
 But have not got to it yet, to much computer time and not enough bench
time.
                                      Mikek

 

[piglet]

On 24/09/2022 11:58, amdx wrote:

On 9/24/2022 1:11 AM, piglet wrote:
On 24/09/2022 00:33, amdx wrote:
  On one of the groups I monitor, it was ask if the tube circuit
shown below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low.
(purpose is to reduce loading on the LC being measured--Q meter)
  Can you give a little detail on how you would implement it? And
what happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100MΩ grid resistor, Is that a detriment to adding bootstrapping?
  I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100MΩ, does
bootstrapping help that?
  Might be barking up the wrong tree.

 
https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0



                             Thanks, Mikek


Did you see the value of C302? I think V301 is being used in the so
called \"infinite impedance\" detector configuration. The cathode
voltage is not the buffered grid ac input but instead is a dc level
representing something between the peak and average ac input. So AC
bootstrapping that circuit not possible. I guess you could insert a
whole new hi-z buffer between tuned circuit under test and detector
but in practice it would be difficult to get it to be an improvement?

piglet


 I thought I\'d seen it referred to as an \"infinite impedance detector\",
but don\'t find it in the manual.
I do see this line in the manual and don\'t understand how transit time
loss in the voltmeter
tube causes a shunt resistance across the measuring circuit at High
Frequencies.
Can someone provide a simple explanation?

\"QVOLTMETERCONDUCTANCE.
Anotherinternalparameterwhich causestheindi-
catedQtodeviatefromeffectiveQ,atbothvery low
andveryhighfrequencies, is theinputconductanceof
theQvoltmeter circuit.Atvery low frequenciesthis
conductance consistsofa 100megohmgridleak resistor
inparallelwiththeinternallossesofthevacuum tube.
Atveryhighfrequenciesthetransittime lossinthevolt-
metertubeshuntstheresonatingcapacitorandintro-
duces ashuntresistance acrossthemeasuring circuit\"

Someone has developed a single fet circuit that does an OK job.
I always wanted to try this circuit,
https://www.dropbox.com/s/wvwkw2iapf4ufd7/High%20impedance%20input.jpg?dl=0

 But have not got to it yet, to much computer time and not enough bench
time.
                                      Mikek

The transit time part comes into it like this: due to the electron
travel time there is a short delay before grid voltage changes during
the positive peaks get to influence the cathode voltage. At very high
frequencies this increased loading at the peaks gets more significant.

The 535A tube is quite large and looks like it was built for low
inter-electrode capacity and low leakage. For VHF use a smaller tube
like an acorn device may have had shorter delay times.

Looks to me like that bootstrap all nodes fet buffer came straight out
of a late 1960s-early 70s NatSemi app note.

piglet



at very high frequencies the electron time delay

 

[piglet]

On 24/09/2022 11:33, amdx wrote:

On 9/23/2022 11:20 PM, John Larkin wrote:
On Fri, 23 Sep 2022 22:53:43 -0500, amdx <amdx@knology.net> wrote:

On 9/23/2022 7:05 PM, John Larkin wrote:
On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

   On one of the groups I monitor, it was ask if the tube circuit
shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low.
(purpose
is to reduce loading on the LC being measured--Q meter)
   Can you give a little detail on how you would implement it? And
what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
   I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
   Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0



                              Thanks, Mikek
The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

  Just following along on an IO Group in a discussion about designing a
Q meter.
Looking at the Boonton 260A and the HP 4342A for ideas and then
someone ask about bootstrapping the tube in the 260A, so I thought I\'d
ask here.
   Several years ago I made five 6\" coils with 8, 9, 10, 12, and 13
turns per inch, just to see where
the sweet spot was. One peaked close to 1400Q.
    I wound a ferrite rod a couple days ago and it measured 1236Q at
800kHz.
I think someone questioned that number and they wound one on the same
material, and theirs peaked at 1326Q.
The material is very low loss in the AM band, R40C1.
  Also some interesting discussion on 50 to 1 injection transformers
driven by low impedance amps.
Here\'s the beginning, but you need to start at the bottom of the the
page for the first post and page up to follow.
https://groups.io/g/Test-Equipment-Design-Construction/topic/making_a_q_meter/92809585?p=Created%2C%2C%2C20%2C2%2C200%2C0&jump=1


                                  Mikek
When the old Boonton was designed, it was hard to generate and
precisely measure frequencies. Now it\'s easy. So resonant width is a
good way to measure Q.

  Yes. I have done a lot of that, but all I had was a scope for amplitude.
So, after too many BW tests, I came up with 5 / 7 = 0.714, that is close
enough
for relative measurements using a scope. So, I would set Fr at 7 units,
adjusting either
the freq gen amplitude or scope CAL, then raise/lower the frequency
until I had 5 units.
At that the time about 30 years ago, I was sorting piezo discs with Qs
over 1200 for the highest Q.
We only used the highest Q discs to bond for our transducers.
 I ask this on another group and never got an answer,
\"Is it possible to come up with a formula or spreadsheet that would
calculate Q from 3db BW,
with the caveat that instead of using 0.707, use the number 0.7 and
normalize it to 3db?
That would be great for needle type meters. For oscilloscopes, I\'d like
to see
0.714 normalized to 3db, why because I like to peak at 7 units and
raise/lower the frequency to 5 units.
5 / 7 = 0.714.
       Mikek

If that were me I\'d use either 0.7 or 0.714 but be consistent and just
the same method everytime. Good enough for comparative evaluation of
coils and both are within 1% of 0.707 and anyhow Q is such a tricky
parameter to nail down consistently.

piglet

 

[amdx]

On 9/24/2022 6:47 AM, piglet wrote:

On 24/09/2022 11:58, amdx wrote:
On 9/24/2022 1:11 AM, piglet wrote:
On 24/09/2022 00:33, amdx wrote:
  On one of the groups I monitor, it was ask if the tube circuit
shown below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low.
(purpose is to reduce loading on the LC being measured--Q meter)
  Can you give a little detail on how you would implement it? And
what happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s,
but when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use
a 100MΩ grid resistor, Is that a detriment to adding bootstrapping?
  I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100MΩ, does
bootstrapping help that?
  Might be barking up the wrong tree.

 
https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0



                             Thanks, Mikek


Did you see the value of C302? I think V301 is being used in the so
called \"infinite impedance\" detector configuration. The cathode
voltage is not the buffered grid ac input but instead is a dc level
representing something between the peak and average ac input. So AC
bootstrapping that circuit not possible. I guess you could insert a
whole new hi-z buffer between tuned circuit under test and detector
but in practice it would be difficult to get it to be an improvement?

piglet


  I thought I\'d seen it referred to as an \"infinite impedance
detector\", but don\'t find it in the manual.
I do see this line in the manual and don\'t understand how transit
time loss in the voltmeter
tube causes a shunt resistance across the measuring circuit at High
Frequencies.
Can someone provide a simple explanation?

\"QVOLTMETERCONDUCTANCE.
Anotherinternalparameterwhich causestheindi-
catedQtodeviatefromeffectiveQ,atbothvery low
andveryhighfrequencies, is theinputconductanceof
theQvoltmeter circuit.Atvery low frequenciesthis
conductance consistsofa 100megohmgridleak resistor
inparallelwiththeinternallossesofthevacuum tube.
Atveryhighfrequenciesthetransittime lossinthevolt-
metertubeshuntstheresonatingcapacitorandintro-
duces ashuntresistance acrossthemeasuring circuit\"

Someone has developed a single fet circuit that does an OK job.
I always wanted to try this circuit,
https://www.dropbox.com/s/wvwkw2iapf4ufd7/High%20impedance%20input.jpg?dl=0

  But have not got to it yet, to much computer time and not enough
bench time.
                                       Mikek




The transit time part comes into it like this: due to the electron
travel time there is a short delay before grid voltage changes during
the positive peaks get to influence the cathode voltage. At very high
frequencies this increased loading at the peaks gets more significant.

The 535A tube is quite large and looks like it was built for low
inter-electrode capacity and low leakage. For VHF use a smaller tube
like an acorn device may have had shorter delay times.

Looks to me like that bootstrap all nodes fet buffer came straight out
of a late 1960s-early 70s NatSemi app note.

piglet

  Yes, it did, but, the question is will it work in place of the tube?
                                             Thanks, Mikek

 

[piglet]

On 24/09/2022 13:44, amdx wrote:

On 9/24/2022 6:47 AM, piglet wrote:
On 24/09/2022 11:58, amdx wrote:
On 9/24/2022 1:11 AM, piglet wrote:
On 24/09/2022 00:33, amdx wrote:
  On one of the groups I monitor, it was ask if the tube circuit
shown below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low.
(purpose is to reduce loading on the LC being measured--Q meter)
  Can you give a little detail on how you would implement it? And
what happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s,
but when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use
a 100MΩ grid resistor, Is that a detriment to adding bootstrapping?
  I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100MΩ, does
bootstrapping help that?
  Might be barking up the wrong tree.

 
https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0



                             Thanks, Mikek


Did you see the value of C302? I think V301 is being used in the so
called \"infinite impedance\" detector configuration. The cathode
voltage is not the buffered grid ac input but instead is a dc level
representing something between the peak and average ac input. So AC
bootstrapping that circuit not possible. I guess you could insert a
whole new hi-z buffer between tuned circuit under test and detector
but in practice it would be difficult to get it to be an improvement?

piglet


  I thought I\'d seen it referred to as an \"infinite impedance
detector\", but don\'t find it in the manual.
I do see this line in the manual and don\'t understand how transit
time loss in the voltmeter
tube causes a shunt resistance across the measuring circuit at High
Frequencies.
Can someone provide a simple explanation?

\"QVOLTMETERCONDUCTANCE.
Anotherinternalparameterwhich causestheindi-
catedQtodeviatefromeffectiveQ,atbothvery low
andveryhighfrequencies, is theinputconductanceof
theQvoltmeter circuit.Atvery low frequenciesthis
conductance consistsofa 100megohmgridleak resistor
inparallelwiththeinternallossesofthevacuum tube.
Atveryhighfrequenciesthetransittime lossinthevolt-
metertubeshuntstheresonatingcapacitorandintro-
duces ashuntresistance acrossthemeasuring circuit\"

Someone has developed a single fet circuit that does an OK job.
I always wanted to try this circuit,
https://www.dropbox.com/s/wvwkw2iapf4ufd7/High%20impedance%20input.jpg?dl=0

  But have not got to it yet, to much computer time and not enough
bench time.
                                       Mikek




The transit time part comes into it like this: due to the electron
travel time there is a short delay before grid voltage changes during
the positive peaks get to influence the cathode voltage. At very high
frequencies this increased loading at the peaks gets more significant.

The 535A tube is quite large and looks like it was built for low
inter-electrode capacity and low leakage. For VHF use a smaller tube
like an acorn device may have had shorter delay times.

Looks to me like that bootstrap all nodes fet buffer came straight out
of a late 1960s-early 70s NatSemi app note.

piglet

  Yes, it did, but, the question is will it work in place of the tube?
                                             Thanks, Mikek

It will work as a buffer ahead of the tube. Or indeed in front of any
other detector/rectifier. You cannot make an infinite impedance detector
with bootstrapped input in one go - gotta have something to drive the
bootstrap!

piglet

 

[Jan Panteltje]

\"Is it possible to come up with a formula or spreadsheet that would
calculate Q from 3db BW,

B = f0 / Q
where
B is the 3 dB bandwidth
f0 is the center frequency

So Q = f0 / B

So for example 100 kHz bandwidth at 1 MHz makes a Q of 10.

 

[amdx]

On 9/24/2022 10:30 AM, Jan Panteltje wrote:

\"Is it possible to come up with a formula or spreadsheet that would
calculate Q from 3db BW,
B = f0 / Q
where
B is the 3 dB bandwidth
f0 is the center frequency

So Q = f0 / B

So for example 100 kHz bandwidth at 1 MHz makes a Q of 10.

 Yes, I got that part, what I wanted is to normalize 0.7 and 0.714
voltage levels
to 3db. But as piglet said, it is a very small error, so fagetaboutit.
Unless you want to do the math and tell me multiply by xxx to get the
3db BW.

The question as posed,

\"Is it possible to come up with a formula or spreadsheet that would
calculate Q from 3db BW,  edit: hmm, poorly worded
with the caveat that instead of using 0.707, use the number 0.7 and
normalize it to 3db?
That would be great for needle type meters. For oscilloscopes, I\'d like
to see
0.714 normalized to 3db, why because I like to peak at 7 units and
raise/lower the frequency to 5 units.
5 / 7 = 0.714.                                Mikek

 

[John Larkin]

On Sat, 24 Sep 2022 11:27:42 -0500, amdx <amdx@knology.net> wrote:

On 9/24/2022 10:30 AM, Jan Panteltje wrote:
\"Is it possible to come up with a formula or spreadsheet that would
calculate Q from 3db BW,
B = f0 / Q
where
B is the 3 dB bandwidth
f0 is the center frequency

So Q = f0 / B

So for example 100 kHz bandwidth at 1 MHz makes a Q of 10.

 Yes, I got that part, what I wanted is to normalize 0.7 and 0.714
voltage levels
to 3db. But as piglet said, it is a very small error, so fagetaboutit.
Unless you want to do the math and tell me multiply by xxx to get the
3db BW.

There is a simple fudge factor. Spice will show what it is.

You can also fudge for a finite source resistance in the signal
source, which matters only for very high Qs.

The other way to measure Q is to drive a series LC from a 50 ohm
signal generator, series resonate, and scope the input voltage. Be
careful about harmonics.

 

[Phil Hobbs]

John Larkin wrote:

On Fri, 23 Sep 2022 22:53:43 -0500, amdx <amdx@knology.net> wrote:

On 9/23/2022 7:05 PM, John Larkin wrote:
On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

 Just following along on an IO Group in a discussion about designing a
Q meter.
Looking at the Boonton 260A and the HP 4342A for ideas and then
someone ask about bootstrapping the tube in the 260A, so I thought I\'d
ask here.
  Several years ago I made five 6\" coils with 8, 9, 10, 12, and 13
turns per inch, just to see where
the sweet spot was. One peaked close to 1400Q.
   I wound a ferrite rod a couple days ago and it measured 1236Q at
800kHz.
I think someone questioned that number and they wound one on the same
material, and theirs peaked at 1326Q.
The material is very low loss in the AM band, R40C1.
 Also some interesting discussion on 50 to 1 injection transformers
driven by low impedance amps.
Here\'s the beginning, but you need to start at the bottom of the the
page for the first post and page up to follow.
https://groups.io/g/Test-Equipment-Design-Construction/topic/making_a_q_meter/92809585?p=Created%2C%2C%2C20%2C2%2C200%2C0&jump=1


                                 Mikek

When the old Boonton was designed, it was hard to generate and
precisely measure frequencies. Now it\'s easy. So resonant width is a
good way to measure Q.

All the usual definitions of Q and resonant frequency give the same
results at high Q, but not at low Q. You have to pay attention for Q <~ 10.

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 Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.


https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek

Here\'s a High-Q resonator we created at lunch today:

https://www.dropbox.com/s/eeocp4rvgaz2dis/Hi-Q_resonator.MOV?dl=0

entirely by accident. There may be a gain mechanism involved.

 

[John Larkin]

On Sun, 25 Sep 2022 13:55:44 -0400, Phil Hobbs
<pcdhSpamMeSenseless@electrooptical.net> wrote:

John Larkin wrote:
On Fri, 23 Sep 2022 22:53:43 -0500, amdx <amdx@knology.net> wrote:

On 9/23/2022 7:05 PM, John Larkin wrote:
On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

 Just following along on an IO Group in a discussion about designing a
Q meter.
Looking at the Boonton 260A and the HP 4342A for ideas and then
someone ask about bootstrapping the tube in the 260A, so I thought I\'d
ask here.
  Several years ago I made five 6\" coils with 8, 9, 10, 12, and 13
turns per inch, just to see where
the sweet spot was. One peaked close to 1400Q.
   I wound a ferrite rod a couple days ago and it measured 1236Q at
800kHz.
I think someone questioned that number and they wound one on the same
material, and theirs peaked at 1326Q.
The material is very low loss in the AM band, R40C1.
 Also some interesting discussion on 50 to 1 injection transformers
driven by low impedance amps.
Here\'s the beginning, but you need to start at the bottom of the the
page for the first post and page up to follow.
https://groups.io/g/Test-Equipment-Design-Construction/topic/making_a_q_meter/92809585?p=Created%2C%2C%2C20%2C2%2C200%2C0&jump=1


                                 Mikek

When the old Boonton was designed, it was hard to generate and
precisely measure frequencies. Now it\'s easy. So resonant width is a
good way to measure Q.


All the usual definitions of Q and resonant frequency give the same
results at high Q, but not at low Q. You have to pay attention for Q <~ 10.

Cheers

Phil Hobbs

An R-L-C goes to unity gain at low frequencies, which suggests Q=1 by
the simple voltage multiplier formula, for any R.

Or Q goes to zero if the L is the grounded part.

Q is a crude way to collapse the actual impedances. I use it to select
parts for LC oscillators, where it does correlate to spectral purity
and jitter.

Coaxial ceramic resonators get into the thousands, 10\'s of thousands
sometimes, but are not well suited to my oscillators; equivalent
txline impedances are too low.

 

[Phil Hobbs]

John Larkin wrote:

On Sun, 25 Sep 2022 13:55:44 -0400, Phil Hobbs
pcdhSpamMeSenseless@electrooptical.net> wrote:

John Larkin wrote:
On Fri, 23 Sep 2022 22:53:43 -0500, amdx <amdx@knology.net> wrote:

On 9/23/2022 7:05 PM, John Larkin wrote:
On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
The bootstrap would reduce the R201 loading a bit but I doubt it would
improve Q measurement much.

The loading of V301 probably dominates.

Are you trying to measure super high Qs?

I built a Q-meter

https://www.dropbox.com/s/ycd8ya9kwit8o0y/Q-meter.jpg?raw=1

https://www.dropbox.com/s/b4xwf1g2ldk0bfl/Q.JPG?raw=1

I used resonance width, rather than voltage multiplication.

I didn\'t have any black wrinkle paint to finish it off.

 Just following along on an IO Group in a discussion about designing a
Q meter.
Looking at the Boonton 260A and the HP 4342A for ideas and then
someone ask about bootstrapping the tube in the 260A, so I thought I\'d
ask here.
  Several years ago I made five 6\" coils with 8, 9, 10, 12, and 13
turns per inch, just to see where
the sweet spot was. One peaked close to 1400Q.
   I wound a ferrite rod a couple days ago and it measured 1236Q at
800kHz.
I think someone questioned that number and they wound one on the same
material, and theirs peaked at 1326Q.
The material is very low loss in the AM band, R40C1.
 Also some interesting discussion on 50 to 1 injection transformers
driven by low impedance amps.
Here\'s the beginning, but you need to start at the bottom of the the
page for the first post and page up to follow.
https://groups.io/g/Test-Equipment-Design-Construction/topic/making_a_q_meter/92809585?p=Created%2C%2C%2C20%2C2%2C200%2C0&jump=1


                                 Mikek

When the old Boonton was designed, it was hard to generate and
precisely measure frequencies. Now it\'s easy. So resonant width is a
good way to measure Q.


All the usual definitions of Q and resonant frequency give the same
results at high Q, but not at low Q. You have to pay attention for Q <~ 10.

Cheers

Phil Hobbs

An R-L-C goes to unity gain at low frequencies, which suggests Q=1 by
the simple voltage multiplier formula, for any R.

Or Q goes to zero if the L is the grounded part.

Q is a crude way to collapse the actual impedances. I use it to select
parts for LC oscillators, where it does correlate to spectral purity
and jitter.

Coaxial ceramic resonators get into the thousands, 10\'s of thousands
sometimes, but are not well suited to my oscillators; equivalent
txline impedances are too low.

\'Q\' is a bit like \'baseband\'--useful but not too clearly defined.

For low-loss systems, the calculated

f0 = 1/(2 pi sqrt(LC))

gives, to decent accuracy,

(1) the position of the peak of the Z(f) curve,
(2) the frequency where the impedance of a tank circuit is purely
resistive, and
(3) the average frequency of the free ringdown waveform.

Similarly, and for much the same reason,

Q = 1/(2 pi f_0 R_s C)
= 2 pi f_0 L/R_s
= 1/logarithmic decrement of the free ringing, per cycle
= FWHM / f_0.

For a tank circuit, series and parallel resistances produce equivalent
loss when

R_p = (1 + Q**2) R_s.

All of these hold for high Q, but none of them at low Q. The
discrepancies tend to go like 1/Q**2, so you have to start paying
attention as Q goes below about 10.

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

 

[amdx]

On 9/25/2022 3:22 PM, John Larkin wrote:

On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
Here\'s a High-Q resonator we created at lunch today:

https://www.dropbox.com/s/eeocp4rvgaz2dis/Hi-Q_resonator.MOV?dl=0

entirely by accident. There may be a gain mechanism involved.

Air currents from the fan sloshing around in the bowl adding gain?

I had comments on another group about bootstrapping the tube.

\"The V301 valve in the Boonton Q meter is NOT a cathode follower
but an \" infinite impedance detector\" due to the capacitor C302.
So you cannot bootstrap at AC.\"

 That\'s beyond my level.

 

[John Larkin]

On Sun, 25 Sep 2022 21:37:54 -0500, amdx <amdx@knology.net> wrote:

On 9/25/2022 3:22 PM, John Larkin wrote:
On Fri, 23 Sep 2022 18:33:05 -0500, amdx <amdx@knology.net> wrote:

 On one of the groups I monitor, it was ask if the tube circuit shown
below could be bootstrapped to raise the input impedance,
especially at the higher frequencies where it drops pretty low. (purpose
is to reduce loading on the LC being measured--Q meter)
 Can you give a little detail on how you would implement it? And what
happens to gain and frequency response?
When was bootstrapping introduced, the word started in the 1880s, but
when was it first used in an electrical circuit?
I did manage to find a bootstrapped tube circuit, but it didn\'t use a
100M? grid resistor, Is that a detriment to adding bootstrapping?
 I made my proposed idea on the right side.
Hmm, the problem is more the tube input than the 100M?, does
bootstrapping help that?
 Might be barking up the wrong tree.

https://www.dropbox.com/s/in6no3nvw0qnie7/Boonton%20260A%20Tube%20bootstrap%20question.jpg?dl=0


                            Thanks, Mikek
Here\'s a High-Q resonator we created at lunch today:

https://www.dropbox.com/s/eeocp4rvgaz2dis/Hi-Q_resonator.MOV?dl=0

entirely by accident. There may be a gain mechanism involved.

Air currents from the fan sloshing around in the bowl adding gain?

I had comments on another group about bootstrapping the tube.

\"The V301 valve in the Boonton Q meter is NOT a cathode follower
but an \" infinite impedance detector\" due to the capacitor C302.
So you cannot bootstrap at AC.\"

 That\'s beyond my level.

The triode is acting like a rectifier into C201 so it can drive a DC
meter.

But a little grid capacitance can be, probably already is, calibrated
out of the circuit.

A triode grid has a resistive component that reduces Q a bit. I don\'t
know the magnitude of that.