Re: Making a Q-meter / References etc
On Thu, 25 Aug 2022 at 03:48, MAX < max@...> wrote: It seems that I joined this list at just the right time.? For several years I have been contemplating building a Q meter with no restriction on the type of technology used.? I have a Boonton 260 which I have restored to near like new performance which I am quite proud of.? I also use it frequently.? ?But I would like to have a Q meter that doesn't depend on a tube that is absolutely and totally unobtainable.? To that end I have given some thought to how to go about building one. <snip> Two methods I have contemplated is to use a low resistance resistor and monitor the current using a tungsten lamp instead of the thermocouple.? ? I think that would be problematic because of the inductance of the lamp. I can see that working at low frequency, but not at RF.
I wondered about heating a wire, or probably a thin wide strip of nichrome, but instead of using a thermocouple, use a non-contact device like an IR thermometer. I know there are IR imaging techniques that can detect a 1 mK temperature change. I know someone who worked on military thermal imaging systems. I recall asking him once about the temperature change that could be detected, but he would not say as it is covered by the Official Secrets Act in the UK. But I know 1 mK can be detected for sure. So I don't think there's a need to heat something to 100's of degrees C. As a rough guess, I would have thought heating something 20 deg C would give a decent enough change to measure, but without the limited lifetime of bulbs.
> Heathkit's entry into the Q meter realm was interesting in that instead of a small resistor they used a large capacitor.? It was a 5000 pf capacitor which was constructed rather like a mica compression capacitor except that everything was screwed down tight.? Modern surface mount capacitors might have less inductance and higher Q than their leaded counterparts.? I don't know that I am suggesting it as a possibility.? Meanwhile some of the alternate circuits that have been discussed bear investigating.
Thank you David for inviting me to join.? Although the Q meter traffic was overwhelming at first things seem to be settling down.? ?
One? can always mute a topic of no interest, although I realise in this case you did have an interet.?
73 (Regards).
Max K 4 O D S.
|
Re: Optimal full scale deflection on analog meters
On Thu, 25 Aug 2022 at 02:14, MAX < max@...> wrote: Since micro controllers are on topic here I can suggest that an analog meter face and moving pointer could be generated as a graphic on an L C D ?display.? This would make it easy to implement the earlier suggested automatic feature of an expanded scale with the current reading at the center of the scale. ? 73 (Regards). ? Max K 4 O D S.
I have mentioned this before. I reckon if there are 400 pixels, that's probably enough to equal any analog meter.
I have thought of driving an analog meter with a DAC, because the frequency response will not be totally flat. As long as something is controlling the signal generator, it should be possible to do a sweep of the frequency, measure the response on a ADC, then use to make the meter be scale independent of frequency.
Dave
|
Re: Optimal full scale deflection on analog meters
No surprise, as the movement would normally be damped by the very low resistance of the thermocouple to which it should be connected.
With the data you have, make a 0-1 Amp ammeter and test that with a 1/2 FSD current.
You will find it exhibits 'normal' damping. That is unless you have a thermocouple to try it out on!
Pete G4GJL
On Wed, Aug 24, 2022 at 9:34 PM Mikek < amdx@...> wrote: Here's a video I posted of the Multiply by Q meter from the 260A.
?I just found it interesting that it was so Undamped.
????????????????????????????????? Mikek
|
Re: Making a Q-meter / References etc
Keeping the freq the same looks like an interesting interesting approach. Should reduce possible error sources.
Need to get someone to move my 160A from the garage to the workbench. Hasn't been powered in over 20 years, so probably needs some new caps.
John KK6IL
toggle quoted message
Show quoted text
On 8/23/2022 5:57 AM, John KN5L wrote: Hi John,
On 8/23/22 3:13 AM, John Kolb wrote:
"Approximations can be made by resonating a coil on the Q meter, and
then noting the reduction in Q obtained when the capacitor is added."
If the Caps are equal, Qc = Q1*Q2/2*(Q1-Q2) I believe measurement should be performed at same frequency. Attached is
a SimSmith model example. Using Voltage Ratio method as shown in
Suppose want to measure Q of a 120pF capacitor at nominal 4MHz.
Set up Q meter such that Q meter capacitor is 240pF, resonate a coil
near 4MHz. In the SimSmith example, C1=240pF L1 is an inductor selected
for nominal 4MHz resonance, 5.5uH and Q meter measured Q=99.987 at
4.3806MHz, as shown in Blue trace.
Add 120pF capacitor to Q meter and re-resonate to 4.3806MHz. In SimSmith
added capacitor is C2, 120pF with Q=1k. Q meter capacitor is now 120pF,
C3. Violet trace indicates Q=95.231
Calculated 120pF Qc = (99.987 x 95.231) / (2 * (99.987 - 95.231)) = 1001
John KN5L
|
Re: Making a Q-meter / References etc
It seems that I joined this list at just the right time. For several years I have been contemplating building a Q meter with no restriction on the type of technology used. I have a Boonton 260 which I have restored to near like new performance which I am quite proud of. I also use it frequently. But I would like to have a Q meter that doesn't depend on a tube that is absolutely and totally unobtainable. To that end I have given some thought to how to go about building one. I have created a folder and moved all messages pertaining to making a Q meter into it. The count is 97 messages but I haven't been through them to see which contain useful information and which do not. The earliest one is dated August 4th. I have spot checked a few and I see there are other methods than the one used by Boonton. Two methods I have contemplated is to use a low resistance resistor and monitor the current using a tungsten lamp instead of the thermocouple. A brightness matching method used in optics could be use to set the current by matching the current monitor lamp to another one operating from DC. Heathkit's entry into the Q meter realm was interesting in that instead of a small resistor they used a large capacitor. It was a 5000 pf capacitor which was constructed rather like a mica compression capacitor except that everything was screwed down tight. Modern surface mount capacitors might have less inductance and higher Q than their leaded counterparts. I don't know that I am suggesting it as a possibility. Meanwhile some of the alternate circuits that have been discussed bear investigating.
Thank you David for inviting me to join. Although the Q meter traffic was overwhelming at first things seem to be settling down.
73 (Regards).
Max K 4 O D S.
I've Never Lost the Wonder.
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of John KN5L Sent: Tuesday, August 23, 2022 7:58 AM To: [email protected]Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / References etc Hi John, On 8/23/22 3:13 AM, John Kolb wrote: "Approximations can be made by resonating a coil on the Q meter, and
then noting the reduction in Q obtained when the capacitor is added."
If the Caps are equal, Qc = Q1*Q2/2*(Q1-Q2) I believe measurement should be performed at same frequency. Attached is a SimSmith model example. Using Voltage Ratio method as shown in Suppose want to measure Q of a 120pF capacitor at nominal 4MHz. Set up Q meter such that Q meter capacitor is 240pF, resonate a coil near 4MHz. In the SimSmith example, C1=240pF L1 is an inductor selected for nominal 4MHz resonance, 5.5uH and Q meter measured Q=99.987 at 4.3806MHz, as shown in Blue trace. Add 120pF capacitor to Q meter and re-resonate to 4.3806MHz. In SimSmith added capacitor is C2, 120pF with Q=1k. Q meter capacitor is now 120pF, C3. Violet trace indicates Q=95.231 Calculated 120pF Qc = (99.987 x 95.231) / (2 * (99.987 - 95.231)) = 1001 John KN5L
|
Re: Optimal full scale deflection on analog meters
Since micro controllers are on topic here I can suggest that an analog meter face and moving pointer could be generated as a graphic on an L C D ?display.? This would make it easy to implement the earlier suggested automatic feature of an expanded scale with the current reading at the center of the scale. ? 73 (Regards). ? Max K 4 O D S. ? I've Never Lost the Wonder. ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of tgerbic
Sent: Friday, August 19, 2022 7:13 PM
To: [email protected]
Subject: Re: [Test Equipment Design & Construction] Optimal full scale deflection on analog meters ? To be clear in the last paragraph, I mean something like 4V at the extreme left and 6V at the extreme right side.
This helps eliminate one of the weaknesses of most analog VOMs.? If you have a voltage like 250V but you need to make measurements with half volt accuracy (or could be trying to peak the value of a high Q circuit). You are stuck with setting the meter on something like the 300V or 500V scale but because of the closeness of markings it would be hard to read down to a couple of volts. If you could expand the scale so 250V was in the middle of the scale and read + or - 10V (from 240V to 260V,) it would be easy to make out very small increments of change. You could make the argument that in this case a DMM would be better, and it might, as long as the voltage change was slow enough that it would give the DMM time to settle to keep the digits from constantly spinning up and down, preventing getting any useful measurement. To some extent the Heathkit IM-25 has an advantage in that you can set the divider to the next lower scale and move the needle around to get a bit of an expanded scale. Not perfect but can be done.
Analog and digital meters have there strengths and weaknesses. One cannot completely replace the other so it is nice to have both and know when to apply each for the measurement you want.
|
Re: Optimal full scale deflection on analog meters
I¡¯ve been lurking here since August 4th but I can¡¯t resist weighting in on this thread.? There is a good reason why HP and others used the 1 3 10 sequence on their analog meters intended primarily for audio work.? The dB. There was and is a good reason why Tektronix used the 1 2 5 10 sequence on their oscilloscopes.? The difficulty of accurately reading the scale on an oscilloscope. ??Of course digital technology has made this discussion moot. As has already been pointed out analog meters are still needed to accurately adjust a control for a maximum or minimum reading. ?Digital meters are by their nature 1 10 100 animals.? ?If you aren¡¯t getting the accuracy you need just above 1/10 scale you add another digit.? Since most of us are building our own unique one-of meters we can do whatever we want and/or need with only a miniscule increase in cost.? There is one more alternative.? Suppose I have a very nice looking meter movement with preprinted scales.? ?Make the ranges to fit the scales.? I have in fact done this. ??Since you are the only one who will ever use it you are the only person you need to pleas.? So keep on having fun. ? 73 (Regards). ? Max K 4 O D S. ? I've Never Lost the Wonder. ?
toggle quoted message
Show quoted text
From: [email protected] [mailto:[email protected]] On Behalf Of Dr. David Kirkby, Kirkby Microwave Ltd
Sent: Friday, August 19, 2022 1:29 PM
To: [email protected]
Subject: Re: [Test Equipment Design & Construction] Optimal full scale deflection on analog meters ? On Fri, 19 Aug 2022 at 11:28, ray_g4lua via <raymond.gathergood=[email protected]> wrote: Having used all kinds of multi-meters over a 50 year career, I have concluded that no matter where the range limits are set, you always come across a measurement that? falls around the overlap!
AVO are perhaps the best known manufacturer of analogue meters, back in the day. I have personally owned an AVO Multi-minor, an AVO Model 7, an AVO model 8 and a much rarer "AVO Test Set Multi Range No1" - the Mil Spec version of the AVO 8?(NATO part number 6625-99-105-7050). The multi-minor, model 7 and Model 8 were all ranged as: 1.0, 2.5,10, 25... etc., up to?2500V FSD
The Test Set Multi Range No1 was different; it's ranges were 1.0, 3.0, 10, 30, 100... up to 3000V FSD However, my thoughts are 1.0, 2.5, 5.0, 10, 25, 50... would be better; based on the fact the the best accuracy for an analogue meter is above 50% FSD.
Obviously if one has specific uses, then different scales would be useful.? But otherwise, I think if one wants N different full--scales between each decade, then they should be differ by a factor of 10^(1/(N-1)). So optimal values would theoretically be To get above 50% of full scale, one would need to go insert another range However, I think that lot would be pretty dam confusing. Could a human brain cope with It might make the meter a bit messy, but would allow any reading to be over 50% of the scale. 73 de Ray G4LUA
|
Re: Optimal full scale deflection on analog meters
Here's a video I posted of the Multiply by Q meter from the 260A.
?I just found it interesting that it was so Undamped.
????????????????????????????????? Mikek
|
Re: Optimal full scale deflection on analog meters
Oops I forgot to attach the meter movement
|
Re: Optimal full scale deflection on analog meters
I found an auction on eBay today. It's for a meter movement out of a Boonton Q meter. There are some interesting things about this.
* Q increases from left to right, but there is also a delta Q, which goes from right to left.
* The Q scales are not perfectly linear.
I would buy the meter if the price was right, but $100 is too much. I might try to find an old AVO meter, to get me a decent size mirrored movement for low cost. In very good condition those meters fetch a fair bit, but average ones are not worth much.
|
Re: Making a Q-meter / References etc
Hi John, On 8/23/22 3:13 AM, John Kolb wrote: "Approximations can be made by resonating a coil on the Q meter, and
then noting the reduction in Q obtained when the capacitor is added."
If the Caps are equal, Qc = Q1*Q2/2*(Q1-Q2) I believe measurement should be performed at same frequency. Attached is a SimSmith model example. Using Voltage Ratio method as shown in Suppose want to measure Q of a 120pF capacitor at nominal 4MHz. Set up Q meter such that Q meter capacitor is 240pF, resonate a coil near 4MHz. In the SimSmith example, C1=240pF L1 is an inductor selected for nominal 4MHz resonance, 5.5uH and Q meter measured Q=99.987 at 4.3806MHz, as shown in Blue trace. Add 120pF capacitor to Q meter and re-resonate to 4.3806MHz. In SimSmith added capacitor is C2, 120pF with Q=1k. Q meter capacitor is now 120pF, C3. Violet trace indicates Q=95.231 Calculated 120pF Qc = (99.987 x 95.231) / (2 * (99.987 - 95.231)) = 1001 John KN5L
|
Re: Making a Q-meter / References etc
Hi David, On 8/23/22 2:27 AM, Dr. David Kirkby, Kirkby Microwave Ltd wrote: When you used the VNWA, did you connect the VNWA directory to the coil?
Doing that will load the circuit - you need to couple via only a small
amount. There are several different BW measurement configurations. Measuring parallel resonant circuit in shunt configuration requires small coupling capacitors. As described in, previously referenced, Jacques' document Figure 1. Another method is Series resonant in shunt as shown in Figure 4. Similar is parallel resonant in Series, as used in: Another set of measurement in: At end of web is section "SimSmith Series versus Shunt" which compares parallel resonant in series and parallel resonant in shunt with 1pF coupling capacitors. Note the coupling capacitors are loading the parallel resonant circuit. Parallel resonant in series has greater accuracy. John KN5L
|
Re: Making a Q-meter / References etc
Hello all,
I have a GR 1650 Universal Impedance Bridge.
It works up to 20 KHz but with an external generator even somewhat higher up in frequency.
In attachment you find' The general radio Experimenter' Vol.33 No3 March 1959 with some info.
There is some information about "the Orthonull " detection that allows the measurement of very low Q's.
You can find further info on Orthonull in patent 2,872,639.
You can find a manual at:
Jan
toggle quoted message
Show quoted text
-----Original Message----- From: [email protected] < [email protected]> On Behalf Of John Kolb Sent: Tuesday, August 23, 2022 10:13 AM To: [email protected]Subject: Re: [Test Equipment Design & Construction] Making a Q-meter / References etc The manual for the Advance Q Meter T2 gives a method to measure Q of a capacitor. "Approximations can be made by resonating a coil on the Q meter, and then noting the reduction in Q obtained when the capacitor is added." If the Caps are equal, Qc = Q1*Q2/2*(Q1-Q2) I haven't had any luck with this method. Q with 2 caps measures higher and calculated Q of C2 is a negative number. One C is a Jennings Vacuum Variable and the other cap from a BC221 freq meter, both set to 150 pF. q measured using the attenuation in a 50 ohm circuit method. Negative Q when either is measured first as C1. This method would require that the Q of the coil is the same at both freqs. also might be more accurate if the C of the coil was determined and added to C1 and C2 in the formula. John On 8/23/2022 12:27 AM, Dr. David Kirkby, Kirkby Microwave Ltd wrote: On Mon, 22 Aug 2022 at 15:16, John Oppenheimer <john@...
<mailto:john@...>> wrote:
I've investigated several Q measuring methods. Appended to:
<> is using Voltage
Ratio. DG1032 used as generator and Oscilloscope used as a RMS
voltmeter. An Oscilloscope is not normally a good voltmeter, but is
what I have. I did not expect an exact match to BW Q measurements,
close.
When you used the VNWA, did you connect the VNWA directory to the coil?
Doing that will load the circuit - you need to couple via only a small
amount. The use of 0.5 pF capacitors was mentioned in one of the
documents. The problem is the capacitors need to be equal each side,
and I'm not sure how to achieve that. There's a paper of over 100
pages in length in the files section about using a VNA to measure Q.
It was written by an employee of the National Physical Laboratory (NPL) in the UK.
Possibly inspire a micro-controller two channel Vrms meter with
greater accuracy.
A meter that does not load the circuit is a challenge.
As Q increases, voltage measured across series LC will decrease,
limiting measurement range depending on Vrms meter range.
For this measurement, measured Q is both L and C. As always, C Q is
necessary to isolate L Q.
I am not sure how to measure the Q of the C. The assumption that it is
irrelevant is not always true.
John KN5L
G8WRB
|
Re: Making a Q-meter / References etc
The manual for the Advance Q Meter T2 gives a method to measure Q of a capacitor.
"Approximations can be made by resonating a coil on the Q meter, and then noting the reduction in Q obtained when the capacitor is added."
If the Caps are equal, Qc = Q1*Q2/2*(Q1-Q2)
I haven't had any luck with this method. Q with 2 caps measures higher and calculated Q of C2 is a negative number. One C is a Jennings Vacuum Variable and the other cap from a BC221 freq meter, both set to 150 pF. q measured using the attenuation in a 50 ohm circuit method. Negative Q when either is measured first as C1.
This method would require that the Q of the coil is the same at both freqs. also might be more accurate if the C of the coil was determined and added to C1 and C2 in the formula.
John
toggle quoted message
Show quoted text
On 8/23/2022 12:27 AM, Dr. David Kirkby, Kirkby Microwave Ltd wrote: On Mon, 22 Aug 2022 at 15:16, John Oppenheimer <john@... <mailto:john@...>> wrote:
I've investigated several Q measuring methods. Appended to:
<> is using Voltage
Ratio. DG1032 used as generator and Oscilloscope used as a RMS
voltmeter. An Oscilloscope is not normally a good voltmeter, but is
what I have. I did not expect an exact match to BW Q measurements,
close.
When you used the VNWA, did you connect the VNWA directory to the coil? Doing that will load the circuit - you need to couple via only a small amount. The use of 0.5 pF capacitors was mentioned in one of the documents. The problem is the capacitors need to be equal each side, and I'm not sure how to achieve that. There's a paper of over 100 pages in length in the files section about using a VNA to measure Q. It was written by an employee of the National Physical Laboratory (NPL) in the UK.
Possibly inspire a micro-controller two channel Vrms meter with
greater accuracy.
A meter that does not load the circuit is a challenge.
As Q increases, voltage measured across series LC will decrease,
limiting measurement range depending on Vrms meter range.
For this measurement, measured Q is both L and C. As always, C Q is
necessary to isolate L Q.
I am not sure how to measure the Q of the C. The assumption that it is irrelevant is not always true.
John KN5L G8WRB
|
Re: Making a Q-meter / References etc
On Mon, 22 Aug 2022 at 15:16, John Oppenheimer < john@...> wrote: I've investigated several Q measuring methods. Appended to: is using Voltage Ratio. DG1032 used as generator and Oscilloscope used as a RMS voltmeter. An Oscilloscope is not normally a good voltmeter, but is what I have. I did not expect an exact match to BW Q measurements, close.
When you used the VNWA, did you connect the VNWA directory to the coil? Doing that will load the circuit - you need to couple via only a small amount. The use of 0.5 pF capacitors was mentioned in one of the documents. The problem is the capacitors need to be equal each side, and I'm not sure how to achieve that. There's a paper of over 100 pages in length in the files section about using a VNA to measure Q. It was written by an employee of the National Physical Laboratory (NPL) in the UK.?
Possibly inspire a micro-controller two channel Vrms meter with greater accuracy.
? A meter that does not load the circuit is a challenge.?
As Q increases, voltage measured across series LC will decrease, limiting measurement range depending on Vrms meter range.
For this measurement, measured Q is both L and C. As always, C Q is necessary to isolate L Q.
I am not sure how to measure the Q of the C. The assumption that it is irrelevant is not always true.?
John KN5L
G8WRB
|
Re: Making a Q-meter / References etc
On Tue, 23 Aug 2022 at 03:31, Steve Lindberg < steve_tech@...> wrote:
Maybe some audio test gear. Ultra low noise. Active cooling. I have an ESI that matches resistors to 8 1/2 digits. Differential rejection. I have Tek, AP audio tools. Also Heathkit, some others too.
I think you replied to the wrong thread. This one is about Q-meters, but another thread is about low noise amplifiers.
|
Re: Making a Q-meter / References etc
Maybe some audio test gear. Ultra low noise. Active cooling. I have an ESI that matches resistors to 8 1/2 digits. Differential rejection. I have Tek, AP audio tools. Also Heathkit, some others too.
toggle quoted message
Show quoted text
I think John Addis would enjoy this group.
|
Re: Optimal full scale deflection on analog meters: What would HP DO??
Hello Jerry,
I own a?R & S type UVN meter as shown earlier, and as you mentioned it increases its?sensitivity when rotated?in the?clockwise direction.? One switches?the instrument on by turning the?rotary knob clockwise?from?the 6 o¡¯clock or?¡¯Off¡¯ position; the first click checks the internal amplifier and the second click selects?the 300 Volts FSD range?(actually?320 V as mentioned earlier). My theory is that the instrument?was designed this way?for two?reasons: a) The input stage is protected against potential?over-voltage conditions to some extent when the input is already?connected to a signal source -?if we assume a user that pays attention to the readings. b) The rotary knobs of oscilloscopes and other measurement?instruments work?in the same way; i.e. their Volts/DIV?sensitivity *increases*?in the clockwise direction. Typically in the known?5¨C2-1 step sizes such as 50-20-10 Volts/DIV.? Side note:?Several?Multimeters work in the same way; at least for either?DC or AC?voltage measurements. Note there is no?uniform way.?I am including some examples below.
Other members may know more or are welcome to correct me here¡
Cheers,
Magnus?
AVO meters:?
Hello Magnus,
A usability, not technical question. The HP meter control begins at a low-range setting and increases as the knob is rotated in a clockwise direction. This seems rather 'natural' to me; clock faces, volume controls, most other analog instruments that I'm familiar with.
The R&S meter control begins as the high-range setting and decreases as the knob is rotated in a clockwise direction. I can see value to this design, but wonder if it causes any confusion when in use.
Jerry S
|
Re: Optimal full scale deflection on analog meters: What would HP DO??
Hello Magnus,
A usability, not technical question. The HP meter control begins at a low-range setting and increases as the knob is rotated in a clockwise direction. This seems rather 'natural' to me; clock faces, volume controls, most other analog instruments that I'm familiar with.
The R&S meter control begins as the high-range setting and decreases as the knob is rotated in a clockwise direction. I can see value to this design, but wonder if it causes any confusion when in use.
Jerry S
|
Re: Wanted - moderator in a different timezone
Thank you very much, I guess I'm just running a little slow.
????????????????????????????? Mikek
|
Dr. David Kirkby, Kirkby Microwave Ltd
Pete_G4GJL
John Kolb
