Question on Q of toroid coil

[joseph2k]

Now, *this* is really interesting. But how would you accurately measure
the
two phase angles? Please share whatever information you have on using
this method of determining Q.

Thanks,

Dave

Oh come on, do you really need the underlying physics and electronics of
standard Vector Network Analysers (VNA) explained to you?

 

[Dave]

Hi Dave, You can use an oscilloscope and a signal generator to get a
pretty
good idea of Q. Do you have an oscilloscope and signal generator?
If so, I'll post the method I have used.
Mike

Hey Mike,

I have the Oscope, and am planning on getting a signal generator.
Icurrrently have a source of RF, but only at one frequency (for whatever
crystal I plug into it.) Please, speak! Knowledge is never wasted, and I
would like to hear what you have to say...

Thanks,

Dave

 

[Dave]

[...]
Huh. A Q meter. This interests me. Any ideas on where I could find
info
on building one of those? (I have a lot more time than money, and it
does
sound interesting...)

Thanks for the info.

Dave

This one is a nice design but needs an external signal generator.
http://users.tpg.com.au/users/ldbutler/QMeter.htm

I printed that article, and am puzzled by one thing. Some of the resistors
in the circuit diagram are labled things like "4k7" and "4m7." Is that 4.7K
or 47K? And 4.7Meg or 47Meg?

It does look like a good project, in spite of my earlier comments about the
"Poor Man's Q-Meter."

Thanks,

Dave

 

[Mike Monett]

I printed that article, and am puzzled by one thing. Some of the
resistors in the circuit diagram are labled things like "4k7" and
"4m7." Is that 4.7K or 47K? And 4.7Meg or 47Meg?

You can often tell by looking at the function of the resistor. For example,
R1 and R3 set the bias voltage of V1 at approximately VCC/2. They are
labeled 4k7, so that means 4.7k. That makes sense since the network has to
supply the base current for V1.

BTW, you might need a 50 ohm termination at the input when working at
higher frequencies, say >10MHz.

Resistor R15 grounds the gate of the jfet, V4. Here, the leakage current is
very low. It is labeled 4m7, so a 4.7meg resistor would be appropriate.

So he uses 4k7 to mean 4.7k, and 4m7 to mean 4.7megohm. That should help
identify any other components in question.

It does look like a good project, in spite of my earlier comments
about the "Poor Man's Q-Meter."

Yes, it does look very nice. I'd do the low impedance driver made up of V1,
V2, and V3 a bit differently, but it should work the way it is shown. You
might want to put that part in LTspice and see how it can be improved. You
want the lowest possible driver source impedance.

I particularly like the calibration method. It should help to minimize
the effect of unavoidable gain variations from one end of the frequency
range to the other.

Thanks,

Dave

Regards,

Mike Monett

Antiviral, Antibacterial Silver Solution:
http://silversol.freewebpage.org/index.htm
SPICE Analysis of Crystal Oscillators:
http://silversol.freewebpage.org/spice/xtal/clapp.htm
Noise-Rejecting Wideband Sampler:
http://www3.sympatico.ca/add.automation/sampler/intro.htm

 

[amdx]

Hey Mike,

I have the Oscope, and am planning on getting a signal generator.
Icurrrently have a source of RF, but only at one frequency (for whatever
crystal I plug into it.) Please, speak! Knowledge is never wasted, and I
would like to hear what you have to say...

Thanks,

Dave

I thought I had this written up in a word file but couldn't find it,
probably lost
in a harddrive crash or upgrade.
First you need to resonate the inductor at the frequency you want to know
the Q.
You should use a good low loss capacitor, I usually use polystyrene.
This method uses the difference between the upper and lower 3db points.
The less loading of you resonate circuit the more accurate the result.
Let's use a 10uh inductor and a 1000pf capacitor as a parallel resonate
circuit.
This should resonate at about 1.59Mhz, however when you put the (X10)
scope probe across the circuit it will slightly change the resonance and
loading.
{More on that later.}
You need to lightly couple some energy from your signal generator into the
resonate circuit. This can be done by placing the generator wires near the
resonate circuit close enough to get the scope signal level you need but as
far
away as possible so you don't load the circuit. You can couple it with high
value resistors if desired, but this increases loading on the inductor.

Then I adjust coupling and signal generator output to get 7 units
on the scope. Why 7 units? I'm glad you ask! You want to move the signal
generator frequency up until the voltage on the scope drops to 5 units.

Some explanation; We want to measure the upper and lower frequency points
where the voltage drop is 3db or .707 times the resonate voltage.
So, back to the 7 units, .707 times 7 units equals 4.949 units or 5 units
when
I'm looking at my scope.

So we adjust the signal generator frequency to peak the waveform on the
scope.
Let's say the waveform peaks at 1,596,200hz.
To get the 7 units sometimes I adjust the generator drive and sometimes I
change
the scope variable attenuator.

I move the frequency up until the scope reads 5 units, Record this
frequency.
Lets say it's 1,600,200hz
Now move the frequency down until the scope reads 5 units, Record this
frequency.
Lets say this is 1,592,219hz
Do the math 1,600,200 - 1,592,219 = 7981
then using the resonate frequency of 1,596,200 / 7981 = 200
The Q of your inductor is 200

{More on that later.} Ok this is later, I have been known to isolate the
scope probe with a
100k or 1meg ohm resistor. This helps reduce the loading on the resonate
circuit, But this
can also induce some 60hz into the scope wave form.
I think I got this all correct, it has been a few years since I did this. I
used it a lot when I was
trying methods to improve the Q of some potcore inductors.
Let me know if you have questions.
Mike

 

[Dave]

I thought I had this written up in a word file but couldn't find it,
probably lost
in a harddrive crash or upgrade.
First you need to resonate the inductor at the frequency you want to know
the Q.
You should use a good low loss capacitor, I usually use polystyrene.
This method uses the difference between the upper and lower 3db points.
The less loading of you resonate circuit the more accurate the result.
Let's use a 10uh inductor and a 1000pf capacitor as a parallel resonate
circuit.
This should resonate at about 1.59Mhz, however when you put the (X10)
scope probe across the circuit it will slightly change the resonance and
loading.
{More on that later.}
You need to lightly couple some energy from your signal generator into the
resonate circuit. This can be done by placing the generator wires near the
resonate circuit close enough to get the scope signal level you need but
as far
away as possible so you don't load the circuit. You can couple it with
high
value resistors if desired, but this increases loading on the inductor.

Then I adjust coupling and signal generator output to get 7 units
on the scope. Why 7 units? I'm glad you ask! You want to move the signal
generator frequency up until the voltage on the scope drops to 5 units.

Some explanation; We want to measure the upper and lower frequency points
where the voltage drop is 3db or .707 times the resonate voltage.
So, back to the 7 units, .707 times 7 units equals 4.949 units or 5 units
when
I'm looking at my scope.

So we adjust the signal generator frequency to peak the waveform on the
scope.
Let's say the waveform peaks at 1,596,200hz.
To get the 7 units sometimes I adjust the generator drive and sometimes I
change
the scope variable attenuator.

I move the frequency up until the scope reads 5 units, Record this
frequency.
Lets say it's 1,600,200hz
Now move the frequency down until the scope reads 5 units, Record this
frequency.
Lets say this is 1,592,219hz
Do the math 1,600,200 - 1,592,219 = 7981
then using the resonate frequency of 1,596,200 / 7981 = 200
The Q of your inductor is 200

{More on that later.} Ok this is later, I have been known to isolate the
scope probe with a
100k or 1meg ohm resistor. This helps reduce the loading on the resonate
circuit, But this
can also induce some 60hz into the scope wave form.
I think I got this all correct, it has been a few years since I did this.
I used it a lot when I was
trying methods to improve the Q of some potcore inductors.
Let me know if you have questions.
Mike

Wow. Totally cool! Thank you very much, Mike. This is going to be a most
engrossing experiment. Thank you, thank you, very much.

Dave

 

[Dave]

You can often tell by looking at the function of the resistor. For
example,
R1 and R3 set the bias voltage of V1 at approximately VCC/2. They are
labeled 4k7, so that means 4.7k. That makes sense since the network has to
supply the base current for V1.

BTW, you might need a 50 ohm termination at the input when working at
higher frequencies, say >10MHz.

Resistor R15 grounds the gate of the jfet, V4. Here, the leakage current
is
very low. It is labeled 4m7, so a 4.7meg resistor would be appropriate.

So he uses 4k7 to mean 4.7k, and 4m7 to mean 4.7megohm. That should help
identify any other components in question.


Yes, it does look very nice. I'd do the low impedance driver made up of
V1,
V2, and V3 a bit differently, but it should work the way it is shown. You
might want to put that part in LTspice and see how it can be improved. You
want the lowest possible driver source impedance.

I particularly like the calibration method. It should help to minimize
the effect of unavoidable gain variations from one end of the frequency
range to the other.



Regards,

Mike Monett

Antiviral, Antibacterial Silver Solution:
http://silversol.freewebpage.org/index.htm
SPICE Analysis of Crystal Oscillators:
http://silversol.freewebpage.org/spice/xtal/clapp.htm
Noise-Rejecting Wideband Sampler:
http://www3.sympatico.ca/add.automation/sampler/intro.htm

I came to the same conclusion, RE: 4k7 and 4m7 labeled resistors, but the
rest I hadn't gotten to yet. Thank you.

This is definetly going to help...

Dave

 

[amdx]

Wow. Totally cool! Thank you very much, Mike. This is going to be a
most engrossing experiment. Thank you, thank you, very much.

Dave

Hi Dave,
Something I always wanted to do but never go to, is build a board to fix
things
in position. Always seemed to get the info I wanted just laying things on
the bench.
A board should be made of low loss material. It should have pins to solder
the capacitor
and pins to solder the inductor to. I might add a permanent air variable
capacitor
to fine tune the resonate frequency. It would need an input for the signal
generator.
Maybe two parallel lines to couple signal into the circuit. One of these
lines would
run from the cap to the inductor, the other would be from the signal
generator.
Hmm, Just thinking out loud. Let me know when you try this.
Mike
PS. Proud owner of a Boonton 260A Q Meter

 

[[email protected]]

amdx ha scritto:

I thought I had this written up in a word file but couldn't find it,
probably lost
in a harddrive crash or upgrade.
First you need to resonate the inductor at the frequency you want to know
the Q.
You should use a good low loss capacitor, I usually use polystyrene.
This method uses the difference between the upper and lower 3db points.
The less loading of you resonate circuit the more accurate the result.
Let's use a 10uh inductor and a 1000pf capacitor as a parallel resonate
circuit.
This should resonate at about 1.59Mhz, however when you put the (X10)
scope probe across the circuit it will slightly change the resonance and
loading.
{More on that later.}
You need to lightly couple some energy from your signal generator into the
resonate circuit. This can be done by placing the generator wires near the
resonate circuit close enough to get the scope signal level you need but as
far
away as possible so you don't load the circuit. You can couple it with high
value resistors if desired, but this increases loading on the inductor.

Then I adjust coupling and signal generator output to get 7 units
on the scope. Why 7 units? I'm glad you ask! You want to move the signal
generator frequency up until the voltage on the scope drops to 5 units.

Some explanation; We want to measure the upper and lower frequency points
where the voltage drop is 3db or .707 times the resonate voltage.
So, back to the 7 units, .707 times 7 units equals 4.949 units or 5 units
when
I'm looking at my scope.

So we adjust the signal generator frequency to peak the waveform on the
scope.
Let's say the waveform peaks at 1,596,200hz.
To get the 7 units sometimes I adjust the generator drive and sometimes I
change
the scope variable attenuator.

I move the frequency up until the scope reads 5 units, Record this
frequency.
Lets say it's 1,600,200hz
Now move the frequency down until the scope reads 5 units, Record this
frequency.
Lets say this is 1,592,219hz
Do the math 1,600,200 - 1,592,219 = 7981
then using the resonate frequency of 1,596,200 / 7981 = 200
The Q of your inductor is 200

{More on that later.} Ok this is later, I have been known to isolate the
scope probe with a
100k or 1meg ohm resistor. This helps reduce the loading on the resonate
circuit, But this
can also induce some 60hz into the scope wave form.
I think I got this all correct, it has been a few years since I did this. I
used it a lot when I was
trying methods to improve the Q of some potcore inductors.
Let me know if you have questions.
Mike

Due to frequency drift a good idea would be to use a variable high Q
air capacitor in series with the inductance to fine tune the LC tank to
resonance.

Cheers

Charles

 

[Dave]

Hi Dave,
Something I always wanted to do but never go to, is build a board to fix
things
in position. Always seemed to get the info I wanted just laying things on
the bench.
A board should be made of low loss material. It should have pins to
solder the capacitor
and pins to solder the inductor to. I might add a permanent air variable
capacitor
to fine tune the resonate frequency. It would need an input for the signal
generator.
Maybe two parallel lines to couple signal into the circuit. One of these
lines would
run from the cap to the inductor, the other would be from the signal
generator.
Hmm, Just thinking out loud. Let me know when you try this.
Mike
PS. Proud owner of a Boonton 260A Q Meter

Hey Mike,

I have a couple of questions on the proceedure you outline below for
calculating the (approximate) Q of a coil. In your example you used a fixed
value capacitor along with the coil under test, and resonate these at the
appropriate frequency (1.59 MHz in your example.) Once that is done, you
move the frequency from the signal generator first up from resonance to get
the lower reading on the Oscope (.707 x the original voltage level displayed
on the scope, or five units rather than seven), and then down from resonance
for the same purpose, jotting down the starting and stopping frequencies
each time. Then, subtracting the higher frequency from the lower provides a
value which is noted. The Q of the coil is found by dividing the resonant
frequency by the difference between the two formar values (the previously
mentioned "noted value.") In my case, I ran the signal from the generator
into the actual circuit I am using, which consists of a variable cap and a
fixed value inductor. Bottom line is, 4.0 MHz to 4.5 MHz minus 4.0 to 3.5
MHz is 1 MHz. Then dividing 4.0 by 1.0 gives me .25. This indicates that
the coil under test has a Q of 25. Is that right? Am I doing this
correctly?

The coil involved is an Amidon T-50-2 toroidal core, wound with #36 wire,
and I expected something more along the lines of a Q around 200, as
indicated was supposedly typical for this frequency according to their
catalogue. Now, granted, this is a lot better than I would have gotten with
my original coil, which consisted of #24 wire wound around a toilet paper
core, but it is still a little disappointing. If I am doing this right, do
you have any ideas on how to increase the Q of this coil?

Many thanks for your help in the original posting. Without that, I would be
a lot more lost than I already am.

Cheerfully yours,

Dave

 

[Dave]

Hey Mike,

I have a couple of questions on the proceedure you outline below for
calculating the (approximate) Q of a coil. In your example you used a
fixed value capacitor along with the coil under test, and resonate these
at the appropriate frequency (1.59 MHz in your example.) Once that is
done, you move the frequency from the signal generator first up from
resonance to get the lower reading on the Oscope (.707 x the original
voltage level displayed on the scope, or five units rather than seven),
and then down from resonance for the same purpose, jotting down the
starting and stopping frequencies each time. Then, subtracting the higher
frequency from the lower provides a value which is noted. The Q of the
coil is found by dividing the resonant frequency by the difference between
the two formar values (the previously mentioned "noted value.") In my
case, I ran the signal from the generator into the actual circuit I am
using, which consists of a variable cap and a fixed value inductor.
Bottom line is, 4.0 MHz to 4.5 MHz minus 4.0 to 3.5 MHz is 1 MHz. Then
dividing 4.0 by 1.0 gives me .25. This indicates that the coil under test
has a Q of 25. Is that right? Am I doing this correctly?

The coil involved is an Amidon T-50-2 toroidal core, wound with #36 wire,
and I expected something more along the lines of a Q around 200, as
indicated was supposedly typical for this frequency according to their
catalogue. Now, granted, this is a lot better than I would have gotten
with my original coil, which consisted of #24 wire wound around a toilet
paper core, but it is still a little disappointing. If I am doing this
right, do you have any ideas on how to increase the Q of this coil?

Many thanks for your help in the original posting. Without that, I would
be a lot more lost than I already am.

Cheerfully yours,

Dave

Nevermind. After a minute considering the numers, I see that I had a Q of
4. Also figured out that my freq counter was loading down the circuit
something terrible. The addition of a 100K resister helped that, and I now
come up with: 6.4 MHz - 6.1 MHz and 6.4 MHz - 6.3 MHz equals a Q of 16, and
the numbers are still somewhat suspect. All of this is spit and bailing
wire, but the latest results seem more reliable (at least to me.) Suspect I
need to rewind the coils with more attention to spacing, and resolder all of
the connections. I *know* the Q has been much higher because at one point I
had to add a fine tuning knob to the device after realizing that those
little clicks and pops weren't static, but were shortwave stations going by
at the speed of light. The fine tuning knob allowed the addition or
subtraction of capacitance one pF at a time and allowed me to pull in very,
very weak stations. Not doing that now though, so something has changed.
Much has been done since that time, so I have a lot of solder joints to
re-work. Still, it's looking better.

Thanks again for your original post. If you don't see this and reply I'll
email you and report, in view of your request to let you know when I tried
to put all of this into action.

Regards,

Dave

 

[john jardine]

[...]

Nevermind. After a minute considering the numers, I see that I had a Q of
4. Also figured out that my freq counter was loading down the circuit
something terrible. The addition of a 100K resister helped that, and I now
come up with: 6.4 MHz - 6.1 MHz and 6.4 MHz - 6.3 MHz equals a Q of 16, and
the numbers are still somewhat suspect. All of this is spit and bailing
wire, but the latest results seem more reliable (at least to me.) Suspect I
need to rewind the coils with more attention to spacing, and resolder all of
the connections. I *know* the Q has been much higher because at one point I
had to add a fine tuning knob to the device after realizing that those
little clicks and pops weren't static, but were shortwave stations going by
at the speed of light. The fine tuning knob allowed the addition or
subtraction of capacitance one pF at a time and allowed me to pull in very,
very weak stations. Not doing that now though, so something has changed.
Much has been done since that time, so I have a lot of solder joints to
re-work. Still, it's looking better.

Thanks again for your original post. If you don't see this and reply I'll
email you and report, in view of your request to let you know when I tried
to put all of this into action.

Regards,

Dave

Just as a side comment, Q about 200 seems not unreasonable for this coil. A
rewind/altered spacing should make little difference to that value . It
seems your tuned circuit is somehow being loaded with about 9kohm across
it, giving the measured Q=16 value. (either that, or the coil resistance is
reading about 30ohms!). If your only loading is a 100k resistor feeding the
circuit and a 1Mohm scope across it, then you should be able to see upto
Q=150. Sure that 100k isn't a 10k?.

 

[Dave]

[...]

Nevermind. After a minute considering the numers, I see that I had a Q
of
4. Also figured out that my freq counter was loading down the circuit
something terrible. The addition of a 100K resister helped that, and I now
come up with: 6.4 MHz - 6.1 MHz and 6.4 MHz - 6.3 MHz equals a Q of 16, and
the numbers are still somewhat suspect. All of this is spit and bailing
wire, but the latest results seem more reliable (at least to me.)
Suspect I
need to rewind the coils with more attention to spacing, and resolder all of
the connections. I *know* the Q has been much higher because at one
point I
had to add a fine tuning knob to the device after realizing that those
little clicks and pops weren't static, but were shortwave stations going by
at the speed of light. The fine tuning knob allowed the addition or
subtraction of capacitance one pF at a time and allowed me to pull in very,
very weak stations. Not doing that now though, so something has changed.
Much has been done since that time, so I have a lot of solder joints to
re-work. Still, it's looking better.

Thanks again for your original post. If you don't see this and reply
I'll
email you and report, in view of your request to let you know when I
tried
to put all of this into action.

Regards,

Dave

Just as a side comment, Q about 200 seems not unreasonable for this coil.
A
rewind/altered spacing should make little difference to that value . It
seems your tuned circuit is somehow being loaded with about 9kohm across
it, giving the measured Q=16 value. (either that, or the coil resistance
is
reading about 30ohms!). If your only loading is a 100k resistor feeding
the
circuit and a 1Mohm scope across it, then you should be able to see upto
Q=150. Sure that 100k isn't a 10k?.

The highest Q calculattion I was able to obtain was 32, and the most
frequently repeated value was on the order of 1 or less. I don't believe
the coil resistance is anything like 30 ohms, but it now occurs to me that a
poor solder connection with the very fine wires I am using might look
something like this. And the 100 K resistor is actually 89K, but it is not
limiting the feed, it is limiting the load that the freq counter puts on the
circuit. Should Ihave another 100K resistor feeding the circuit? What
about a 1 Mohm,or a 1.8 Mohm? One more thing: I am using surplus (used)
switches to switch between the inductors, and I now wonder if they are the
problem.

Thanks for the feedback, and the ideas. Your input is much appreciated.

Dave

 

[john jardine]

[...]

Just as a side comment, Q about 200 seems not unreasonable for this coil.
A
rewind/altered spacing should make little difference to that value . It
seems your tuned circuit is somehow being loaded with about 9kohm across
it, giving the measured Q=16 value. (either that, or the coil resistance
is
reading about 30ohms!). If your only loading is a 100k resistor feeding
the
circuit and a 1Mohm scope across it, then you should be able to see upto
Q=150. Sure that 100k isn't a 10k?.

The highest Q calculattion I was able to obtain was 32, and the most
frequently repeated value was on the order of 1 or less. I don't believe
the coil resistance is anything like 30 ohms, but it now occurs to me that a
poor solder connection with the very fine wires I am using might look
something like this. And the 100 K resistor is actually 89K, but it is not
limiting the feed, it is limiting the load that the freq counter puts on the
circuit. Should Ihave another 100K resistor feeding the circuit? What
about a 1 Mohm,or a 1.8 Mohm? One more thing: I am using surplus (used)
switches to switch between the inductors, and I now wonder if they are the
problem.

Thanks for the feedback, and the ideas. Your input is much appreciated.

Dave

The setup should be similar to this ...

89k, 100k, 1Meg etc
___
o----------o------|___|------o-------o------,
| | | | |
| | | | |
.------o-. .-----o-. | | | .--------.
| o/p| | i/p| || | | | .-----.|
| | | | C| --- | | | ||
| | | | C| --- | | |-----||
|Sig-gen | |Counter| C| | | | '--- -'|
| | | | | | | | -Scope-|
| | | | | | o---o i/p |
| 0V| | 0V| | | o---o 0V |
'------o-' '-----o-' | | | '--------'
| | | | |
o----------o-----------------o-------o------'
Tuned circuit of interest

(created by AACircuit v1.28 beta 10/06/04 www.tech-chat.de)

I'm seeing a small part of another thread about a frequency counter, so will
wait to see how that pans out.
john