Hi Brooke, ? I found this in the 1959 GenRad catalog for their 722 variable capacitors: ? 
? For the 722-MD variant:- ? For a series resistance of 0.02 ohms @1MHz and C=1,000pF, ? Q = 1/(2*pi*f*R*C) = 8.0*1E(12) (!) ? Or using the dissipation factor: ? Q = 1/D? = 1/0.03*1E(-12) = 33E(12) ? I think I got that right? If so, this would have to be the ultimate capacitor. ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of Brooke Clarke via groups.io
Sent: Friday, 30 September 2022 9:50 AM
To: [email protected]
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter /? Hi George:
Any info on the Q of the GR decade capacitors?
-- Have Fun, ? Brooke Clarke axioms: 1. The extent to which you can fix or improve something will be limited by how well you understand how it works. 2. Everybody, with no exceptions, holds false beliefs. -------- Original Message -------- I have a Genrad 722FS ?0 ¨C 1500pF +/-0.1% that I use as a lab cal standard for low values of capacitance. Not sure about the type of insulation it uses, but it is certainly a very well made piece of eqpt. ? There are many different variants of the 722 available, including some on ePay for $$$. I was lucky ¨C I picked mine up for a few dollars at an old radio auction. ? Cheers, George G VK2KGG ? ? From: [email protected] [mailto:[email protected]] On Behalf Of ebrucehunter via groups.io
Sent: Friday, 30 September 2022 7:11 AM
To: john@...; [email protected]
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / ? For a one of a-kind-project, perhaps a General Radio (later Genrad) capacitor could be considered.? These were available in a wide range of capacitance. typically from 13 to 70 pF to as large as 29 to 1120 pF.? GR's 722-series capacitors from the 1950s had steatite insulation with quartz available at an additional cost.? Later products were the 1420, 1421, and 1422-series.? Some types were available with Rexolite insulation for use at higher frequencies.? These capacitors are likely available on eBay. -----Original Message-----
From: John KN5L <john@...>
To: [email protected]
Sent: Thu, Sep 29, 2022 12:08 pm
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / Hi T. Gerbic,
On 9/28/22 7:16 PM, tgerbic wrote:
> What are your thoughts?
First requirement, for making an absolute voltage ratio Q-Meter, is
ability to accurately measure Q.
I use a VNWA for measuring absolute Q.
Next issue is a high Q resonating capacitor. Hard to come by. Best of my
knowledge, there are no suitable variable capacitors in production.
Next best is vintage variable or fixed low capacitance, 100pF, C0G or
SM. Though should have a way to evaluate capacitor Q.
A voltage ration Q Meter requires; accurate generator, Injection
Transformer (the easy part), High Q resonating capacitor, and high
impedance RF voltmeter. Both generator and voltmeter covering wanted
frequency range.
Analysis steps required to evaluate a voltage ratio Q meter. Using a
nominal Q=250 device.
S11 VNWA Q measurement.
VNWA bandwidth Q measurement using high Q C0G and air variable
resonating capacitors. Q measurement compared to above for validation.
As shown, good correlation. This step validates resonating capacitors
for voltage ratio Q meter testing.
Q measurement using prototype RF voltmeter. A Rigol DG1032 AWG is used
for both RF voltmeter calibration and Q meter generator. Results is
about 10% below above VNWA measurements.
The prototype RF voltmeter does not have bootstrap feedback. Though uses
a 10/120 pF capacitive divider in front of FET input with nominal 10Mohm
input R.
A SimSmith model with 50ohm generator and 10Mohm meter. Resulting Q
error is about 5.
John KN5L
?
|
Hi Brooke, ? It is not a decade capacitor, it is a continuously variable unit designed for bench top use. The capacitance setting is read from a precision dial. It contains a set of precision gears that allow accurate setting of capacitance. It is designed for low drift, high mechanical stability etc. and appears to have had no expense spared, so I imagine it has high Q. ? There is some information and data sheets for the 722 range here: ?
? search for General Radio 722 . There are a number of different variants listed, some with photos. Unfortunately they do not mention the 722-FS. ? ? ? Regarding fixed capacitors, a long time ago I did some experimenting with various types of fixed capacitors. I did not have a Q meter, but using a grid dip meter with a coil to form a resonant circuit, I came to the conclusion that polystyrene capacitors (Styroseal) had a fairly high Q. ? Just wondering - has anyone done quantitative Q measurements on different types of capacitors? ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of Brooke Clarke via groups.io
Sent: Friday, 30 September 2022 9:50 AM
To: [email protected]
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter /? Hi George:
Any info on the Q of the GR decade capacitors? -- Have Fun, ? Brooke Clarke axioms: 1. The extent to which you can fix or improve something will be limited by how well you understand how it works. 2. Everybody, with no exceptions, holds false beliefs. -------- Original Message -------- I have a Genrad 722FS ?0 ¨C 1500pF +/-0.1% that I use as a lab cal standard for low values of capacitance. Not sure about the type of insulation it uses, but it is certainly a very well made piece of eqpt. ? There are many different variants of the 722 available, including some on ePay for $$$. I was lucky ¨C I picked mine up for a few dollars at an old radio auction. ? Cheers, George G VK2KGG ? ? From: [email protected] [mailto:[email protected]] On Behalf Of ebrucehunter via groups.io
Sent: Friday, 30 September 2022 7:11 AM
To: john@...; [email protected]
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / ? For a one of a-kind-project, perhaps a General Radio (later Genrad) capacitor could be considered.? These were available in a wide range of capacitance. typically from 13 to 70 pF to as large as 29 to 1120 pF.? GR's 722-series capacitors from the 1950s had steatite insulation with quartz available at an additional cost.? Later products were the 1420, 1421, and 1422-series.? Some types were available with Rexolite insulation for use at higher frequencies.? These capacitors are likely available on eBay. -----Original Message-----
From: John KN5L <john@...>
To: [email protected]
Sent: Thu, Sep 29, 2022 12:08 pm
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / Hi T. Gerbic,
On 9/28/22 7:16 PM, tgerbic wrote:
> What are your thoughts?
First requirement, for making an absolute voltage ratio Q-Meter, is
ability to accurately measure Q.
I use a VNWA for measuring absolute Q.
Next issue is a high Q resonating capacitor. Hard to come by. Best of my
knowledge, there are no suitable variable capacitors in production.
Next best is vintage variable or fixed low capacitance, 100pF, C0G or
SM. Though should have a way to evaluate capacitor Q.
A voltage ration Q Meter requires; accurate generator, Injection
Transformer (the easy part), High Q resonating capacitor, and high
impedance RF voltmeter. Both generator and voltmeter covering wanted
frequency range.
Analysis steps required to evaluate a voltage ratio Q meter. Using a
nominal Q=250 device.
S11 VNWA Q measurement.
VNWA bandwidth Q measurement using high Q C0G and air variable
resonating capacitors. Q measurement compared to above for validation.
As shown, good correlation. This step validates resonating capacitors
for voltage ratio Q meter testing.
Q measurement using prototype RF voltmeter. A Rigol DG1032 AWG is used
for both RF voltmeter calibration and Q meter generator. Results is
about 10% below above VNWA measurements.
The prototype RF voltmeter does not have bootstrap feedback. Though uses
a 10/120 pF capacitive divider in front of FET input with nominal 10Mohm
input R.
A SimSmith model with 50ohm generator and 10Mohm meter. Resulting Q
error is about 5.
John KN5L
?
|
Hi George:
Any info on the Q of the GR decade capacitors?
--
Have Fun,
Brooke Clarke
axioms:
1. The extent to which you can fix or improve something will be limited by how well you understand how it works.
2. Everybody, with no exceptions, holds false beliefs.
-------- Original Message --------
toggle quoted message
Show quoted text
I
have a Genrad 722FS ?0 ¨C 1500pF +/-0.1% that I use as a lab
cal standard for low values of capacitance. Not sure about
the type of insulation it uses, but it is certainly a very
well made piece of eqpt.
?
There
are many different variants of the 722 available, including
some on ePay for $$$. I was lucky ¨C I picked mine up for a
few dollars at an old radio auction.
?
Cheers,
George
G
VK2KGG
?
?
From: [email protected]
[mailto:[email protected]] On
Behalf Of ebrucehunter via groups.io
Sent: Friday, 30 September 2022 7:11 AM
To: john@...;
[email protected]
Subject: Re: [Test Equipment Design &
Construction] Making a Q-meter /
?
For
a one of a-kind-project, perhaps a General Radio (later
Genrad) capacitor could be considered.? These were
available in a wide range of capacitance. typically from
13 to 70 pF to as large as 29 to 1120 pF.? GR's
722-series capacitors from the 1950s had steatite
insulation with quartz available at an additional cost.?
Later products were the 1420, 1421, and 1422-series.?
Some types were available with Rexolite insulation for
use at higher frequencies.? These capacitors are likely
available on eBay.
-----Original
Message-----
From: John KN5L <john@...>
To: [email protected]
Sent: Thu, Sep 29, 2022 12:08 pm
Subject: Re: [Test Equipment Design & Construction]
Making a Q-meter /
Hi
T. Gerbic,
On 9/28/22 7:16 PM, tgerbic wrote:
> What are your thoughts?
First requirement, for making an absolute voltage
ratio Q-Meter, is
ability to accurately measure Q.
I use a VNWA for measuring absolute Q.
Next issue is a high Q resonating capacitor. Hard to
come by. Best of my
knowledge, there are no suitable variable capacitors
in production.
Next best is vintage variable or fixed low
capacitance, 100pF, C0G or
SM. Though should have a way to evaluate capacitor Q.
A voltage ration Q Meter requires; accurate generator,
Injection
Transformer (the easy part), High Q resonating
capacitor, and high
impedance RF voltmeter. Both generator and voltmeter
covering wanted
frequency range.
Analysis steps required to evaluate a voltage ratio Q
meter. Using a
nominal Q=250 device.
S11 VNWA Q measurement.
VNWA bandwidth Q measurement using high Q C0G and air
variable
resonating capacitors. Q measurement compared to above
for validation.
As shown, good correlation. This step validates
resonating capacitors
for voltage ratio Q meter testing.
Q measurement using prototype RF voltmeter. A Rigol
DG1032 AWG is used
for both RF voltmeter calibration and Q meter
generator. Results is
about 10% below above VNWA measurements.
The prototype RF voltmeter does not have bootstrap
feedback. Though uses
a 10/120 pF capacitive divider in front of FET input
with nominal 10Mohm
input R.
A SimSmith model with 50ohm generator and 10Mohm
meter. Resulting Q
error is about 5.
John KN5L
|
I have a Genrad 722FS ?0 ¨C 1500pF +/-0.1% that I use as a lab cal standard for low values of capacitance. Not sure about the type of insulation it uses, but it is certainly a very well made piece of eqpt. ? There are many different variants of the 722 available, including some on ePay for $$$. I was lucky ¨C I picked mine up for a few dollars at an old radio auction. ? Cheers, George G VK2KGG ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of ebrucehunter via groups.io
Sent: Friday, 30 September 2022 7:11 AM
To: john@...; [email protected]
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / ? For a one of a-kind-project, perhaps a General Radio (later Genrad) capacitor could be considered.? These were available in a wide range of capacitance. typically from 13 to 70 pF to as large as 29 to 1120 pF.? GR's 722-series capacitors from the 1950s had steatite insulation with quartz available at an additional cost.? Later products were the 1420, 1421, and 1422-series.? Some types were available with Rexolite insulation for use at higher frequencies.? These capacitors are likely available on eBay. -----Original Message-----
From: John KN5L <john@...>
To: [email protected]
Sent: Thu, Sep 29, 2022 12:08 pm
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / Hi T. Gerbic,
On 9/28/22 7:16 PM, tgerbic wrote:
> What are your thoughts?
First requirement, for making an absolute voltage ratio Q-Meter, is
ability to accurately measure Q.
I use a VNWA for measuring absolute Q.
Next issue is a high Q resonating capacitor. Hard to come by. Best of my
knowledge, there are no suitable variable capacitors in production.
Next best is vintage variable or fixed low capacitance, 100pF, C0G or
SM. Though should have a way to evaluate capacitor Q.
A voltage ration Q Meter requires; accurate generator, Injection
Transformer (the easy part), High Q resonating capacitor, and high
impedance RF voltmeter. Both generator and voltmeter covering wanted
frequency range.
Analysis steps required to evaluate a voltage ratio Q meter. Using a
nominal Q=250 device.
S11 VNWA Q measurement.
VNWA bandwidth Q measurement using high Q C0G and air variable
resonating capacitors. Q measurement compared to above for validation.
As shown, good correlation. This step validates resonating capacitors
for voltage ratio Q meter testing.
Q measurement using prototype RF voltmeter. A Rigol DG1032 AWG is used
for both RF voltmeter calibration and Q meter generator. Results is
about 10% below above VNWA measurements.
The prototype RF voltmeter does not have bootstrap feedback. Though uses
a 10/120 pF capacitive divider in front of FET input with nominal 10Mohm
input R.
A SimSmith model with 50ohm generator and 10Mohm meter. Resulting Q
error is about 5.
John KN5L
|
Hi John:
A possible answer to the large range variable cap is to combine a
series of capacitors using a step switch with a variable cap.? This
is what HP did.
An AM broadcast variable capacitor was 365 pF the step size of the
HP 16462 is 300 pF with switch positions of 0, 300, 600, 900, 1200,
1500, 1800, 2100, 2400 and 2700pF.
The caps are marked:
JFD, UY25, 300V and in values of: 152F, 301F, 271F
There's also a small air variable cap to trim the overall unit.
--
Have Fun,
Brooke Clarke, N6GCE
axioms:
1. The extent to which you can fix or improve something will be limited by how well you understand how it works.
2. Everybody, with no exceptions, holds false beliefs.
-------- Original Message --------
toggle quoted message
Show quoted text
Hi T. Gerbic,
On 9/28/22 7:16 PM, tgerbic wrote:
What are your thoughts?
First requirement, for making an absolute voltage ratio Q-Meter, is
ability to accurately measure Q.
I use a VNWA for measuring absolute Q.
Next issue is a high Q resonating capacitor. Hard to come by. Best of my
knowledge, there are no suitable variable capacitors in production.
Next best is vintage variable or fixed low capacitance, 100pF, C0G or
SM. Though should have a way to evaluate capacitor Q.
A voltage ration Q Meter requires; accurate generator, Injection
Transformer (the easy part), High Q resonating capacitor, and high
impedance RF voltmeter. Both generator and voltmeter covering wanted
frequency range.
Analysis steps required to evaluate a voltage ratio Q meter. Using a
nominal Q=250 device.
S11 VNWA Q measurement.
VNWA bandwidth Q measurement using high Q C0G and air variable
resonating capacitors. Q measurement compared to above for validation.
As shown, good correlation. This step validates resonating capacitors
for voltage ratio Q meter testing.
Q measurement using prototype RF voltmeter. A Rigol DG1032 AWG is used
for both RF voltmeter calibration and Q meter generator. Results is
about 10% below above VNWA measurements.
The prototype RF voltmeter does not have bootstrap feedback. Though uses
a 10/120 pF capacitive divider in front of FET input with nominal 10Mohm
input R.
A SimSmith model with 50ohm generator and 10Mohm meter. Resulting Q
error is about 5.
John KN5L
|
Well, at risk of riling the masses, e-Bay sources can include
test instruments such as selective voltmeters, discarded
Q Meters with blown thermocouples, and bridges, most of
which used? high end variables.
Another lashup which may work is to use high grade caps
such as the ATC-100 Big Block series of caps in a binary
arrangement. 100 pF, 200 pF, 400 pF and 800 pF switched
with high grade relays or high isolation FET's. For finer
granularity use a more easily found air variable covering
around 10 to 100 pF.
Chuck
On Sep 29, 2022, at 5:10 PM, ebrucehunter via groups.io <Brucekareen@...> wrote:
John,
For a one of a-kind-project, perhaps a General Radio (later Genrad) capacitor could be considered.? These were available in a wide range of capacitance. typically from 13 to 70 pF to as large as 29 to 1120 pF.? GR's 722-series capacitors from the 1950s had steatite insulation with quartz available at an additional cost.? Later products were the 1420, 1421, and 1422-series.? Some types were available with Rexolite insulation for use at higher frequencies.? These capacitors are likely available on eBay.
Bruce?
-----Original Message-----
From: John KN5L <john@...>
Sent: Thu, Sep 29, 2022 12:08 pm
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter /
Hi T. Gerbic,
On 9/28/22 7:16 PM, tgerbic wrote:
> What are your thoughts?
First requirement, for making an absolute voltage ratio Q-Meter, is
ability to accurately measure Q.
I use a VNWA for measuring absolute Q.
Next issue is a high Q resonating capacitor. Hard to come by. Best of my
knowledge, there are no suitable variable capacitors in production.
Next best is vintage variable or fixed low capacitance, 100pF, C0G or
SM. Though should have a way to evaluate capacitor Q.
A voltage ration Q Meter requires; accurate generator, Injection
Transformer (the easy part), High Q resonating capacitor, and high
impedance RF voltmeter. Both generator and voltmeter covering wanted
frequency range.
Analysis steps required to evaluate a voltage ratio Q meter. Using a
nominal Q=250 device.
S11 VNWA Q measurement.
VNWA bandwidth Q measurement using high Q C0G and air variable
resonating capacitors. Q measurement compared to above for validation.
As shown, good correlation. This step validates resonating capacitors
for voltage ratio Q meter testing.
Q measurement using prototype RF voltmeter. A Rigol DG1032 AWG is used
for both RF voltmeter calibration and Q meter generator. Results is
about 10% below above VNWA measurements.
The prototype RF voltmeter does not have bootstrap feedback. Though uses
a 10/120 pF capacitive divider in front of FET input with nominal 10Mohm
input R.
A SimSmith model with 50ohm generator and 10Mohm meter. Resulting Q
error is about 5.
John KN5L
|
John,
For a one of a-kind-project, perhaps a General Radio (later Genrad) capacitor could be considered.? These were available in a wide range of capacitance. typically from 13 to 70 pF to as large as 29 to 1120 pF.? GR's 722-series capacitors from the 1950s had steatite insulation with quartz available at an additional cost.? Later products were the 1420, 1421, and 1422-series.? Some types were available with Rexolite insulation for use at higher frequencies.? These capacitors are likely available on eBay.
Bruce?
-----Original Message-----
From: John KN5L <john@...>
To: [email protected]
Sent: Thu, Sep 29, 2022 12:08 pm
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter /
Hi T. Gerbic,
On 9/28/22 7:16 PM, tgerbic wrote:
> What are your thoughts?
First requirement, for making an absolute voltage ratio Q-Meter, is
ability to accurately measure Q.
I use a VNWA for measuring absolute Q.
Next issue is a high Q resonating capacitor. Hard to come by. Best of my
knowledge, there are no suitable variable capacitors in production.
Next best is vintage variable or fixed low capacitance, 100pF, C0G or
SM. Though should have a way to evaluate capacitor Q.
A voltage ration Q Meter requires; accurate generator, Injection
Transformer (the easy part), High Q resonating capacitor, and high
impedance RF voltmeter. Both generator and voltmeter covering wanted
frequency range.
Analysis steps required to evaluate a voltage ratio Q meter. Using a
nominal Q=250 device.
S11 VNWA Q measurement.
VNWA bandwidth Q measurement using high Q C0G and air variable
resonating capacitors. Q measurement compared to above for validation.
As shown, good correlation. This step validates resonating capacitors
for voltage ratio Q meter testing.
Q measurement using prototype RF voltmeter. A Rigol DG1032 AWG is used
for both RF voltmeter calibration and Q meter generator. Results is
about 10% below above VNWA measurements.
The prototype RF voltmeter does not have bootstrap feedback. Though uses
a 10/120 pF capacitive divider in front of FET input with nominal 10Mohm
input R.
A SimSmith model with 50ohm generator and 10Mohm meter. Resulting Q
error is about 5.
John KN5L
|
Hi T. Gerbic, On 9/28/22 7:16 PM, tgerbic wrote: What are your thoughts? First requirement, for making an absolute voltage ratio Q-Meter, is ability to accurately measure Q. I use a VNWA for measuring absolute Q. Next issue is a high Q resonating capacitor. Hard to come by. Best of my knowledge, there are no suitable variable capacitors in production. Next best is vintage variable or fixed low capacitance, 100pF, C0G or SM. Though should have a way to evaluate capacitor Q. A voltage ration Q Meter requires; accurate generator, Injection Transformer (the easy part), High Q resonating capacitor, and high impedance RF voltmeter. Both generator and voltmeter covering wanted frequency range. Analysis steps required to evaluate a voltage ratio Q meter. Using a nominal Q=250 device. S11 VNWA Q measurement. VNWA bandwidth Q measurement using high Q C0G and air variable resonating capacitors. Q measurement compared to above for validation. As shown, good correlation. This step validates resonating capacitors for voltage ratio Q meter testing. Q measurement using prototype RF voltmeter. A Rigol DG1032 AWG is used for both RF voltmeter calibration and Q meter generator. Results is about 10% below above VNWA measurements. The prototype RF voltmeter does not have bootstrap feedback. Though uses a 10/120 pF capacitive divider in front of FET input with nominal 10Mohm input R. A SimSmith model with 50ohm generator and 10Mohm meter. Resulting Q error is about 5. John KN5L
|
Dave,
I too would like to finally have a wide frequency, accurate Q meter and so am following this thread closely and want to see something useful come of it.
I would like to bring back up a couple of subjects mentioned early in a couple of Q meter discussions. This stuff has bothered me since this thread began.
You have stated a desired frequency range of 1.5Mhz to 150Mhz. What is your desired accuracy for the meter? 1%, 5%, 10% or something else, and how much uncertainty is acceptable?? Is it ok for one range to have less accuracy than another, such as 1Mhz to 10Mhz has 1% accuracy, 10Mhz to 100Mhz has 5% accuracy while 100Mhz to 150Mhz has 10% accuracy, or the opposite. Does the accuracy of the Q measurement ranges need to be the same or can the lower Q measurement accuracy be different from the higher Q measurements?? Looking at a bunch of Q meter specs, I see tolerances all over the place based on frequency and range. Some manuals don't mention uncertainty, expectations for uncertainty ranges, or possible ways to calculate it. So what are we trying to work toward, and when will it be enough accuracy to nail down a design?
There is a lot of discussion going on here that seems to be designing for close to perfect measurements across a 150Mhz range of frequencies using either capacitor, resistor or coil driver circuits and some type of non-loading AC voltmeter to read the Q value. I don't think one size will fit all and it will be hard to get good accuracy across the range without some type of automatic compensation function. Just a thought but perhaps the output to the meter needs to feed through a small processor that provides some automatic gain correction based on frequency related amplitude losses at certain calibrated points. This removes the need to over design the circuitry to compensate for such a wide range of frequency in an instrument that does not have perfect components.? In this case more focus is on reading the actual voltage based on the less than perfect components and adding correction to bring the readings into some higher percentage of accuracy.
I have a few Q meters and have also done measurements using 3dB and ring-down methods, as well as measuring Q on a VNA. Each gives a different value, based on the physical environment, and even the mounting of the coil to the measuring device gives different Q readings. I just measured a coil today with 3/4" leads and changing the attachment points along the leads has shown a coil Q change from 170 to 44. The results are almost the exact readings, per attachment point, for two different meters.? The meters are reading ok but he physical variables have a large effect. The mechanical design and physical attachment is as important as anything electrical in the design. You can see this when measuring device parameters with a VNA. You get better accuracy if you include the physical distance between the calibration plane and the device body. A VNA will typically correct for this distance if known.
Another thing that bothers me is what to do with the measurement once you have it. If you are measuring coils to use in a circuit and get a Q reading on a meter, eventually you have to put the coil in a circuit, often inside a physical device. Unfortunately the lead dressing, connection method and closeness to other components affect the coil and associated caps, changing the Q of the actual tuned circuit.? Outside of a general measurement of the Q of the coil and knowing if changes to the coil tend to increase or decrease the Q, it is uncertain what the Q actually is in a circuit.? I would expect that if I needed a high Q for making a filter, I would design or buy the highest Q I practically can. So if I measured a Q of 2000 and could adjust the coil to get 2500 I probably would.? However in reality, the actual Q in the circuit will most likely be something less. So this brings me back to the accuracy needed. Outside of using the meter to accurately measure the Q of a standard on one meter and using it to cal another meter of the same model, extreme accuracy is really not needed and is probably not practically attainable across a 150Mhz range of frequency. Probably a more constant accuracy across the frequency range, and Q scale, with a bit of tolerance for attachment method, might be a better design goal.
Again, I would like to see a wide frequency Q meter that could be built by people on this group that has a reasonable accuracy across the band. I certainly would build one.
What are your thoughts?
Thanks to all that have contributed so far. Lots of good info and measurements so far.
|
I have four of the HP 3586C SLMs. So far I 1.95 in good working order. Remaining issue with the HPIB address setting not being recognized. Intermittent as I replug the HPIB controller board on an extension.? One with a bad mother board having a short after a rectifier. ? Let me know if I can be of help, particularly with testing boards in a known good unit. ? Lester B Veenstra? K1YCM? M?YCM? W8YCM?? 6Y6Y lester@... ? 452 Stable Ln (HC84 RFD USPS Mail) Keyser WV 26726 ? GPS: 39.336826 N? 78.982287 W (Google) GPS: 39.33682 N? 78.9823741 W (GPSDO) ? ? Telephones: Home:?????????????????? ? +1-304-289-6057 US cell??????????????????? +1-304-790-9192 Jamaica cell:?????????? +1-876-456-8898 ? ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of bruno ruche
Sent: Wednesday, September 28, 2022 3:15 AM
To: [email protected]
Subject: Re: [Test Equipment Design & Construction] Making a Q-meter /? This topic has provided me a renewed incentive to complete the restoration of a HP3586c selective voltmeter.? For background on the topic of Q measurement, and a thorough and comprehensive description of the concept--(and simplified too), have a look at this excellent presentation by Max Robinson:? ? The aforementioned link is non-commercial, and also includes historical schematics of various Q measurement devices.? 73 de kr8s
|
Hi Mikek,
Attached is an example 3dB bandwidth and notch dB for a Series Shunt
measurement. Using a 120pF C0G 5% for computing reactance at resonant
frequency.
1.2943MHz, 120pF; X = 1024.7 ohms
Simplified EMRFD Eq 7.4; Q = 2 * X/Z * (10^(dB/20) - 1)
VNWA measured notch = -22.54dB
Applying above formula Q = 2 * 1024.7/50 * (10^(22.54/20) - 1) = 508
VNWA 3dB BW Q = 510
John KN5L
toggle quoted message
Show quoted text
On 9/27/22 11:28 AM, Mikek wrote: Sooo... the formula form EMRAD is, Qs = 4pifLu/Z x (10^A/20 -1). L is in
uH, frequency in MHz, Z is 50¦¸, the system impedance.
4 x 3.14159 x 65 x 0.46 / 50 X (10^(36/20) -1) =
12.56 x 65 x 0.46 / 50 x (10^1.8 -1) =
375.5 (63.0957 -1)
375.544/50 x (63.1 -1) =
751.088 x 62.09 = 466.35 The answer they presented was Q=469. I put in
all the precise numbers because with truncated numbers I didn't get 469.
That didn't help. Looks like they probably extrapolated from the chart
to get notch depth and it was not precisely a 36db notch.
????????????????????????????? Mikek
|
Hi John,
The FET impedance converter is the bandwidth limiting section.
John KN5L
toggle quoted message
Show quoted text
On 9/28/22 1:37 AM, John Kolb wrote: Haven't seen mention of using a Boonton RF Voltmeter as the measuring
device. The model 92 is basically 1% of full scale and 1% of reading
from 50 KHz to 150 MHz, 200 uV to 3V and usable from 10KHz to 1.2 GHz.
Input resistance ranges from 100K to 1.5M below 2 MHz, and falls off
below that with increasing frequency. Not a problem if using the
attenuation method or following a wideband impedance converter.
Attached is a schematic for a full wave rectifier. I wonder what the
frequency response would be if the op amps were AD8055 or similar high
speed devices.
|
This topic has provided me a renewed incentive to complete the restoration of a HP3586c selective voltmeter.? For background on the topic of Q measurement, and a thorough and comprehensive description of the concept--(and simplified too), have a look at this excellent presentation by Max Robinson:? ? The aforementioned link is non-commercial, and also includes historical schematics of various Q measurement devices.? 73 de kr8s
|
Haven't seen mention of using a Boonton RF Voltmeter as the measuring device. The model 92 is basically 1% of full scale and 1% of reading from 50 KHz to 150 MHz, 200 uV to 3V and usable from 10KHz to 1.2 GHz.
Input resistance ranges from 100K to 1.5M below 2 MHz, and falls off below that with increasing frequency. Not a problem if using the attenuation method or following a wideband impedance converter.
Attached is a schematic for a full wave rectifier. I wonder what the frequency response would be if the op amps were AD8055 or similar high speed devices.
John KK6IL
toggle quoted message
Show quoted text
On 9/27/2022 7:48 AM, John KN5L wrote: Hi Mikek,
Making a stand alone voltage ratio Q Meter with range up to 150MHz does
not seem realistic. Limitation is volt meter frequency range. HP-400E
and home-brew voltmeter is limited to about 10MHz.
I'm not sure I have the ability to home-brew a RF voltmeter with
frequency range much above 10MHz. Can not use a HP400E as a back end
detector and meter display because of it's 10MHz BW limit. Therefore,
whatever is designed requires front end to RF detector.
I am testing high Q inductors and will post to the list later today.
Today is a peak Q=600 device. I also have a nominal peak Q=1200 device.
John KN5L
|
Attached is a concept drawing I did to use the binding post mounting screw and a standoff extension as the 1 turn secondary for low inductance/resistance using a T82 size toroid. Assume one would want the same terminal arrangement as on the 260/4342. To scale for a T50 size toroid would either have a large height to allow room for the standoff or a binding post with a very long mounting screw to have a short height. How much space is desirable between the transformer and the shield?
Another concept would have both the inductor and capacitor LO side posts grounded, and the transformer on a heavy bar between the HI posts as the single turn secondary.
John KK6IL
toggle quoted message
Show quoted text
On 9/27/2022 6:15 AM, Mikek wrote: John KN5L,
?So far it looks like you are the only one to have actually run with the ball.
?Do have any thoughts about actually building this in an instrument?
As always once you get to 1MHz and up strays become
important to minimize, but, apparently a Q meter can be built that does this.
? I ask because, I would want a variable capacitor, I would also have a difficult time
measuring a 6 inch diameter coil on the small pcb you have.
?I made the suggestion that unless we can find a better way to make the 50 to 1 transformer so it is wideband,
?we may have to have 3 switchable transformers for low, medium, and high frequencies. The OPs design goal was
1.5MHz to 150MHz. I received no response on the idea. I have enjoyed this thread and your posts.
?????????????????????????????????????????????????????? Mikek
_._,_._,_
------------------------------------------------------------------------
|
A major Q Meter component is a capacitor with very high Q. I've tested
FT82-61 40T using 400pF variable shown in:
at 1200kHz, using VNWA BW method, Q=453. As shown in:
Q should = 525.
I'll need to find another variable capacitor with higher Q!
John KN5L
toggle quoted message
Show quoted text
On 9/27/22 12:27 PM, John KN5L wrote: Hi Mikek,
After making VNWA BW measurements this morning, my VNWA has the dreaded,
"USB device not licensed or wrong activation code" error. I'll need to
wait for a fix from Tom to continue with measurements.
My measurements from this morning:
Using an air variable capacitor and VNWA bandwidth Q measurements are
rather close to HP Q Meter results. Adding a 120pF C0G de-Qs a little.
330pF C0G has excessive Q.
John KN5L
On 9/27/22 9:48 AM, John KN5L wrote:
Hi Mikek,
Making a stand alone voltage ratio Q Meter with range up to 150MHz does
not seem realistic. Limitation is volt meter frequency range. HP-400E
and home-brew voltmeter is limited to about 10MHz.
I'm not sure I have the ability to home-brew a RF voltmeter with
frequency range much above 10MHz. Can not use a HP400E as a back end
detector and meter display because of it's 10MHz BW limit. Therefore,
whatever is designed requires front end to RF detector.
I am testing high Q inductors and will post to the list later today.
Today is a peak Q=600 device. I also have a nominal peak Q=1200 device.
John KN5L
On 9/27/22 8:15 AM, Mikek wrote:
John KN5L,
?So far it looks like you are the only one to have actually run with the
ball.
?Do have any thoughts about actually building this in an instrument?
As always once you get to 1MHz and up strays become
important to minimize, but, apparently a Q meter can be built that does
this.
? I ask because, I would want a variable capacitor, I would also have a
difficult time
measuring a 6 inch diameter coil on the small pcb you have.
?I made the suggestion that unless we can find a better way to make the
50 to 1 transformer so it is wideband,
?we may have to have 3 switchable transformers for low, medium, and high
frequencies. The OPs design goal was
1.5MHz to 150MHz. I received no response on the idea. I have enjoyed
this thread and your posts.
?????????????????????????????????????????????????????? Mikek
|
?Yes, it tracks pretty close to the HP measurements, The numbers show an 11% low reading at 1200kHz.
It is just such a shame and yet a grace* that Q is such an elusive number, 10 people test it and we get 11 answers.
?I suspect the difference in Q between the COG caps and the air cap is, the COGs have more loss than the air cap.
??????????????????????????????????????? Mikek
* Because it is fun to work at!
|
Hi Mikek,
After making VNWA BW measurements this morning, my VNWA has the dreaded,
"USB device not licensed or wrong activation code" error. I'll need to
wait for a fix from Tom to continue with measurements.
My measurements from this morning:
Using an air variable capacitor and VNWA bandwidth Q measurements are
rather close to HP Q Meter results. Adding a 120pF C0G de-Qs a little.
330pF C0G has excessive Q.
John KN5L
toggle quoted message
Show quoted text
On 9/27/22 9:48 AM, John KN5L wrote: Hi Mikek,
Making a stand alone voltage ratio Q Meter with range up to 150MHz does
not seem realistic. Limitation is volt meter frequency range. HP-400E
and home-brew voltmeter is limited to about 10MHz.
I'm not sure I have the ability to home-brew a RF voltmeter with
frequency range much above 10MHz. Can not use a HP400E as a back end
detector and meter display because of it's 10MHz BW limit. Therefore,
whatever is designed requires front end to RF detector.
I am testing high Q inductors and will post to the list later today.
Today is a peak Q=600 device. I also have a nominal peak Q=1200 device.
John KN5L
On 9/27/22 8:15 AM, Mikek wrote:
John KN5L,
?So far it looks like you are the only one to have actually run with the
ball.
?Do have any thoughts about actually building this in an instrument?
As always once you get to 1MHz and up strays become
important to minimize, but, apparently a Q meter can be built that does
this.
? I ask because, I would want a variable capacitor, I would also have a
difficult time
measuring a 6 inch diameter coil on the small pcb you have.
?I made the suggestion that unless we can find a better way to make the
50 to 1 transformer so it is wideband,
?we may have to have 3 switchable transformers for low, medium, and high
frequencies. The OPs design goal was
1.5MHz to 150MHz. I received no response on the idea. I have enjoyed
this thread and your posts.
?????????????????????????????????????????????????????? Mikek
|
On Tue, Sep 27, 2022 at 06:34 AM, Mikek wrote:
I'm looking at this page in the files section,?
/g/Test-Equipment-Design-Construction/files/Q-factor%20%28Q%20is%20the%20inverse%20of%20dissipation%20factor%20DF%29./Method%20simplifies%20testing%20high-Q%20devices;%20Alan%20Victor,%20EDN%20Feb%202002.pdf
?It gives a chart of Notch depth vs series resistance (Rs). That doesn't work well with a 240uh coil with a Q of 1000 at 1MHz.
?I received a PM, showing pages from Experimental Methods in RF Design book, showing a calculation to get Q from Notch depth.
We have one good example in the PDF I posted to work out.
?"You set the signal generator at 65MHz and use a variable, air-di-electric capacitor to fine-tune the notch at 65 MHz. The in-
ductance?is 460 nH at 65 MHz. The measured notch depth is 36 dB. RS is 0.4V, and the, unloaded Q is 469"
Sooo... the formula form EMRAD is, Qs = 4pifLu/Z x (10^A/20 -1). L is in uH, frequency in MHz, Z is 50¦¸, the system impedance.4 x 3.14159 x 65 x 0.46 / 50 X (10^(36/20) -1) =
12.56 x 65 x 0.46 / 50 x (10^1.8 -1) =
375.5 (63.0957 -1)
375.544/50 x (63.1 -1) =
751.088 x 62.09 = 466.35 The answer they presented was Q=469. I put in all the precise numbers because with truncated numbers I didn't get 469.
That didn't help. Looks like they probably extrapolated from the chart to get notch depth and it was not precisely a 36db notch.
????????????????????????????? Mikek
|
Hi Mikek,
Making a stand alone voltage ratio Q Meter with range up to 150MHz does
not seem realistic. Limitation is volt meter frequency range. HP-400E
and home-brew voltmeter is limited to about 10MHz.
I'm not sure I have the ability to home-brew a RF voltmeter with
frequency range much above 10MHz. Can not use a HP400E as a back end
detector and meter display because of it's 10MHz BW limit. Therefore,
whatever is designed requires front end to RF detector.
I am testing high Q inductors and will post to the list later today.
Today is a peak Q=600 device. I also have a nominal peak Q=1200 device.
John KN5L
toggle quoted message
Show quoted text
On 9/27/22 8:15 AM, Mikek wrote: John KN5L,
?So far it looks like you are the only one to have actually run with the
ball.
?Do have any thoughts about actually building this in an instrument?
As always once you get to 1MHz and up strays become
important to minimize, but, apparently a Q meter can be built that does
this.
? I ask because, I would want a variable capacitor, I would also have a
difficult time
measuring a 6 inch diameter coil on the small pcb you have.
?I made the suggestion that unless we can find a better way to make the
50 to 1 transformer so it is wideband,
?we may have to have 3 switchable transformers for low, medium, and high
frequencies. The OPs design goal was
1.5MHz to 150MHz. I received no response on the idea. I have enjoyed
this thread and your posts.
?????????????????????????????????????????????????????? Mikek
|
ebrucehunter
John Kolb