Re: More final comments on the Heathkit QM-1.
Science advances not by repetition of experiments and analysis
of large amounts of data, but by carrying out that one crucial test / experiment
that refutes present knowledge. With proper research design, you only need one
data-gathering event. Then you move on to the next conjecture, or if you like,
further hypothesis generation.
A precise example of this is when Charles Darwin visited Western
Australia in 1836 aboard the HMS Beagle. His prior experience in London, say,
Hyde Park, was that all swans were white. On arrival in Western Australia, he
found birds of the same build, physiology and skeletal structure, but they were
black. So, in that ONE OBSERVATION, he refuted the prior hypothesis that ‘All
swans are white’.
Cheers, Brian, PhD
?
toggle quoted message
Show quoted text
From: [email protected]
[mailto:[email protected]] On Behalf Of Jeff
Green
Sent: Thursday, 17 November 2022 1:52 AM
To: [email protected]
Subject: Re: [Test Equipment Design & Construction] More final
comments on the Heathkit QM-1.
?
As a chemist I am all
too familiar with uncertainty and the limits of knowledge.
That is a very nice
article.
?
In introduction to Chemical Engineering we had to watch a
TV series from 1966, “Design of Experiments” where a professor explained how to
get the most information possible with the fewest tests possible and the series
stressed that perfect knowledge of any chemical process would require an
infinite number of perfect experiments, and no chemical experiment is perfect,
there are always unknowns.
Even something as
simple as hydrogen and oxygen can form hydrogen peroxide. H2O2
?
These two videos cover
it in a slightly different way then what we watched.
?
What Is Design of
Experiments? Part 1
What Is Design of
Experiments? Part 2
?
The goal of a competent
chemist is to conduct as few experiments as possible to obtain useful data.
Chemical experiments
are expensive;, consume time, which is also a cost.
?
At times it felt like
more of an art then pure engineering.
?
We probably spent 7
times as much on quality control as we did for all other expenses. When making
biologicals, you do not cut corners or people die.
?
?
One of my professors
wore his father’s ring. I’ll bet a large softdrink that no one involved with
the Florida pedestrian bridge that collapsed graduated from a Canadian school.
?
This one line is the
biggest “take away:”
“罢丑别??was
miscalculated, and thus was not enough to prevent the??causing
cracks in the truss concrete.”
?
Engineering is all about getting your sums correct.
My EE friend and I will do a complete check of the QM-1 tomorrow. Too many
other topics came up.
The weather here turned nasty a month early and people who were planning new
antenna installations are trying to come up with work arounds.
My EE friend insists "the performance of an antenna is inversely related
to the weather when it was erected." If he's right, a lot of new antenna
installations are going to be wonderful.
?
|
Re: More final comments on the Heathkit QM-1.
Mine's on the shelf in front of me.
Vince - K8ZW.
toggle quoted message
Show quoted text
On 11/16/2022 08:22 PM, Paul via groups.io wrote: Some of us remember how to use a slide rule, accurate to 3 places - maybe.
Paul, W8AEF
--
Michigan VHF Corporation -- nobucks dot net
K8ZW -
Etsy Shop:
|
Re: More final comments on the Heathkit QM-1.
Some of us remember how to use a slide
rule, accurate to 3 places - maybe.
?
Paul, W8AEF
?
toggle quoted message
Show quoted text
From: [email protected]
[mailto:[email protected]]
On Behalf Of G8DQX list
Sent: Wednesday, November 16, 2022
5:54 PM
To: [email protected]
Subject: Re: [Test Equipment
Design & Construction] More final comments on the Heathkit QM-1.
?
Jeff,
trust me,
engineering is an art, the art of
compromise between what one wishes to achieve, what technology (&c.)
permits, how much it's all going to cost (and how long it might take &c.),
and how exactly does this erudite maths relate to the problem at hand…
Robin,
G8DQX (Chemistry to A-level)
On 16/11/2022 14:52, Jeff Green wrote:
?
At
times it felt like more of an art then pure engineering.
?
|
I had a good measurement 9,985,727Ω for a 10MΩ resistor.
The manual suggests using low capacitance for large resistors and high cap for smaller resistors.
I also use a lower Q inductor at about 232. The delta Q was 14. and the capacitance was 40pf.
Frequency was 1.44Mhz.
???????????????????????????????????????????????????????? Mikek
--
Thanks, Mikek
|
Re: More final comments on the Heathkit QM-1.
Jeff,
trust me, engineering is an art, the art of compromise
between what one wishes to achieve, what technology (&c.)
permits, how much it's all going to cost (and how long it might
take &c.), and how exactly does this erudite maths relate to
the problem at hand…
Robin, G8DQX (Chemistry to A-level)
On 16/11/2022 14:52, Jeff Green wrote:
toggle quoted message
Show quoted text
?
At times it
felt like more of an art then pure engineering.
?
|
Using an LF coil at lower freq, with the nominal 98 megohm,
results are not good.? 459 x 448 /? 2? pi? 159,269? 450e-12? 11?
=? 41.512 megohms
On 11/16/2022 3:53 PM, Steve Ratzlaff
via groups.io wrote:
toggle quoted message
Show quoted text
On the idea perhaps the resonating capacitance was too small
compared with any (tiny) inductance of the resistor, I resonated
at the other end, using 450 pF, using the nominal 98 Mohm.
1447 x 1423 / 2? pi? 491,437 Hz? 450 e-12? 24? =? 61.75
megohms. Much better than before but still far from nominal 98
megohms.
On 11/16/2022 12:41 PM, Steve
Ratzlaff via groups.io wrote:
I tried again, this time with four 22 megohm in series with
10 megohm for a nominal 98 megohms total.
1700 kHz, 32.3 pF, coil-alone Q 990; with resistors Q 850.?
990 x 850 / 2? pi? 1.7e6? 32.3e-12? 140? = 17.42 megohms!???
Much much worse than before. Obviously there is something else
going on here, assuming Mike's formula as stated is correct.
Steve
On 11/16/2022 11:07 AM, Steve
Ratzlaff via groups.io wrote:
I tried it with my HP4342A Q meter, and a standard 5% 10
megohm resistor which measures about 10.97 megohms on my
Fluke 87V. (The reading keeps creeping up, never seems to
finally stop....)
1700 kHz, 32.5 pF, coil-alone Q 980. With resistor across
coil, Q 635. (32.1 pF)
So plugging into Mike's formula: 980 x 635/ 2?? pi ? 1.7e6
? 32.5e-12? 345? =? 5.196 megohms----far from what it should
be....
Steve AA7U in AZ
On 11/16/2022 8:35 AM, Mikek
wrote:
On Sat, Nov 12, 2022 at 07:42 AM, Mikek wrote:
Impact of 100M ohm
A simple work around is to measure Q of a coil then to add
100M ohms in parallel then retune and take a second Q
reading then calculate out the 100M ohm first reading.
However most? measurements don't need such a process
because the 100M ohm loading is <1% Q reduction.? If
you have an excellent 100uH inductor? Xl=628 ohms at 1MHz,
so for a Q=2000? the parallel loss will be 1.26M ohm .
Adding 100M ohm in parallel to 1.26M produces 1.2443, a
1.26% drop from Q=2,000 to 1975.? In reality a large coil
near to the metalwork of a Q meter will read low Q because
of eddy currents, in my experience the Q drop due to eddy
currents is often far greater than 1%!
Would some perform this experiment on their Q meter?
?I continuously get a about a 19%? low error. I have used
two different Q meters. I'm attaching thin wires to an 0805
resistor.
But please use what you have.
How to:
?Put an inductor on your meter and measure the Q, and find
the capacitance that was used to resonate.
Record both.
Now put a 10MegΩ across the inductor (retune if needed) and
measure the new lower Q.
Record.
Use this formula to find the value of the measured value of
the 10MΩ.
Rp = Q1 x Q2 / 2 x pi x f x capacitance x delta Q.
Capacitance is in Farads, i.e, 150pf = 0.000000000150.
The delta Q is first Q measured minus the second Q measured.
?
I want to know why I don't get an accurate answer.
This method of measuring high value resistors
is in both the Boonton 260A manual and the HP4342A manual.
?
?? Thanks for any incite you can add to this.? Mikek
|
On the idea perhaps the resonating capacitance was too small
compared with any (tiny) inductance of the resistor, I resonated
at the other end, using 450 pF, using the nominal 98 Mohm.
1447 x 1423 / 2? pi? 491,437 Hz? 450 e-12? 24? =? 61.75 megohms.
Much better than before but still far from nominal 98 megohms.
On 11/16/2022 12:41 PM, Steve Ratzlaff
via groups.io wrote:
toggle quoted message
Show quoted text
I tried again, this time with four 22 megohm in series with 10
megohm for a nominal 98 megohms total.
1700 kHz, 32.3 pF, coil-alone Q 990; with resistors Q 850.? 990
x 850 / 2? pi? 1.7e6? 32.3e-12? 140? = 17.42 megohms!??? Much
much worse than before. Obviously there is something else going
on here, assuming Mike's formula as stated is correct.
Steve
On 11/16/2022 11:07 AM, Steve
Ratzlaff via groups.io wrote:
I tried it with my HP4342A Q meter, and a standard 5% 10
megohm resistor which measures about 10.97 megohms on my Fluke
87V. (The reading keeps creeping up, never seems to finally
stop....)
1700 kHz, 32.5 pF, coil-alone Q 980. With resistor across
coil, Q 635. (32.1 pF)
So plugging into Mike's formula: 980 x 635/ 2?? pi ? 1.7e6 ?
32.5e-12? 345? =? 5.196 megohms----far from what it should
be....
Steve AA7U in AZ
On 11/16/2022 8:35 AM, Mikek wrote:
On Sat, Nov 12, 2022 at 07:42 AM, Mikek wrote:
Impact of 100M ohm
A simple work around is to measure Q of a coil then to add
100M ohms in parallel then retune and take a second Q
reading then calculate out the 100M ohm first reading.
However most? measurements don't need such a process because
the 100M ohm loading is <1% Q reduction.? If you have an
excellent 100uH inductor? Xl=628 ohms at 1MHz, so for a
Q=2000? the parallel loss will be 1.26M ohm . Adding 100M
ohm in parallel to 1.26M produces 1.2443, a 1.26% drop from
Q=2,000 to 1975.? In reality a large coil near to the
metalwork of a Q meter will read low Q because of eddy
currents, in my experience the Q drop due to eddy currents
is often far greater than 1%!
Would some perform this experiment on their Q meter?
?I continuously get a about a 19%? low error. I have used two
different Q meters. I'm attaching thin wires to an 0805
resistor.
But please use what you have.
How to:
?Put an inductor on your meter and measure the Q, and find the
capacitance that was used to resonate.
Record both.
Now put a 10MegΩ across the inductor (retune if needed) and
measure the new lower Q.
Record.
Use this formula to find the value of the measured value of
the 10MΩ.
Rp = Q1 x Q2 / 2 x pi x f x capacitance x delta Q.
Capacitance is in Farads, i.e, 150pf = 0.000000000150.
The delta Q is first Q measured minus the second Q measured.
?
I want to know why I don't get an accurate answer.
This method of measuring high value resistors
is in both the Boonton 260A manual and the HP4342A manual.
?
?? Thanks for any incite you can add to this.? Mikek
|
By standard, is that 10MΩ Is that thru hole? That 5MΩ is about what I got when I was using a thru hole.
I think--- maybe, just maybe I found the answer. If you look at section about measuring "Large Resistors" it
has the formula I posted, but then goes on to say,
"If the resistor is also Reactive,
?Xp = 1 / 2 x pi x F x (C1-C2) " ?
?So, I think we have the? Rp of our original LC, in parallel with 10MΩ, in parallel with Xc.
I don't know how to do imaginary numbers. So, still having a problem.
I'm hoping that math me work out. I seems the manual would have warned about such a thing though!
? I'm still a confused though, because I thought the capacitance resonated away, so why does it affect the resistor.
Heck, it has a much bigger capacitor resonating the inductor!
?????????????????????????????????? Mikek
|
I tried again, this time with four 22 megohm in series with 10
megohm for a nominal 98 megohms total.
1700 kHz, 32.3 pF, coil-alone Q 990; with resistors Q 850.? 990 x
850 / 2? pi? 1.7e6? 32.3e-12? 140? = 17.42 megohms!??? Much much
worse than before. Obviously there is something else going on
here, assuming Mike's formula as stated is correct.
Steve
On 11/16/2022 11:07 AM, Steve Ratzlaff
via groups.io wrote:
toggle quoted message
Show quoted text
I tried it with my HP4342A Q meter, and a standard 5% 10 megohm
resistor which measures about 10.97 megohms on my Fluke 87V.
(The reading keeps creeping up, never seems to finally stop....)
1700 kHz, 32.5 pF, coil-alone Q 980. With resistor across coil,
Q 635. (32.1 pF)
So plugging into Mike's formula: 980 x 635/ 2?? pi ? 1.7e6 ?
32.5e-12? 345? =? 5.196 megohms----far from what it should
be....
Steve AA7U in AZ
On 11/16/2022 8:35 AM, Mikek wrote:
On Sat, Nov 12, 2022 at 07:42 AM, Mikek wrote:
Impact of 100M ohm
A simple work around is to measure Q of a coil then to add
100M ohms in parallel then retune and take a second Q reading
then calculate out the 100M ohm first reading. However most?
measurements don't need such a process because the 100M ohm
loading is <1% Q reduction.? If you have an excellent 100uH
inductor? Xl=628 ohms at 1MHz, so for a Q=2000? the parallel
loss will be 1.26M ohm . Adding 100M ohm in parallel to 1.26M
produces 1.2443, a 1.26% drop from Q=2,000 to 1975.? In
reality a large coil near to the metalwork of a Q meter will
read low Q because of eddy currents, in my experience the Q
drop due to eddy currents is often far greater than 1%!
Would some perform this experiment on their Q meter?
?I continuously get a about a 19%? low error. I have used two
different Q meters. I'm attaching thin wires to an 0805
resistor.
But please use what you have.
How to:
?Put an inductor on your meter and measure the Q, and find the
capacitance that was used to resonate.
Record both.
Now put a 10MegΩ across the inductor (retune if needed) and
measure the new lower Q.
Record.
Use this formula to find the value of the measured value of the
10MΩ.
Rp = Q1 x Q2 / 2 x pi x f x capacitance x delta Q.
Capacitance is in Farads, i.e, 150pf = 0.000000000150.
The delta Q is first Q measured minus the second Q measured.
?
I want to know why I don't get an accurate answer.
This method of measuring high value resistors
is in both the Boonton 260A manual and the HP4342A manual.
?
?? Thanks for any incite you can add to this.? Mikek
|
I tried it with my HP4342A Q meter, and a standard 5% 10 megohm
resistor which measures about 10.97 megohms on my Fluke 87V. (The
reading keeps creeping up, never seems to finally stop....)
1700 kHz, 32.5 pF, coil-alone Q 980. With resistor across coil, Q
635. (32.1 pF)
So plugging into Mike's formula: 980 x 635/ 2?? pi ? 1.7e6 ?
32.5e-12? 345? =? 5.196 megohms----far from what it should be....
Steve AA7U in AZ
On 11/16/2022 8:35 AM, Mikek wrote:
toggle quoted message
Show quoted text
On Sat, Nov 12, 2022 at 07:42 AM, Mikek wrote:
Impact of 100M ohm
A simple work around is to measure Q of a coil then to add 100M
ohms in parallel then retune and take a second Q reading then
calculate out the 100M ohm first reading. However most?
measurements don't need such a process because the 100M ohm
loading is <1% Q reduction.? If you have an excellent 100uH
inductor? Xl=628 ohms at 1MHz, so for a Q=2000? the parallel
loss will be 1.26M ohm . Adding 100M ohm in parallel to 1.26M
produces 1.2443, a 1.26% drop from Q=2,000 to 1975.? In reality
a large coil near to the metalwork of a Q meter will read low Q
because of eddy currents, in my experience the Q drop due to
eddy currents is often far greater than 1%!
Would some perform this experiment on their Q meter?
?I continuously get a about a 19%? low error. I have used two
different Q meters. I'm attaching thin wires to an 0805 resistor.
But please use what you have.
How to:
?Put an inductor on your meter and measure the Q, and find the
capacitance that was used to resonate.
Record both.
Now put a 10MegΩ across the inductor (retune if needed) and
measure the new lower Q.
Record.
Use this formula to find the value of the measured value of the
10MΩ.
Rp = Q1 x Q2 / 2 x pi x f x capacitance x delta Q.
Capacitance is in Farads, i.e, 150pf = 0.000000000150.
The delta Q is first Q measured minus the second Q measured.
?
I want to know why I don't get an accurate answer.
This method of measuring high value resistors
is in both the Boonton 260A manual and the HP4342A manual.
?
?? Thanks for any incite you can add to this.? Mikek
|
On Sat, Nov 12, 2022 at 07:42 AM, Mikek wrote:
Impact of 100M ohm
A simple work around is to measure Q of a coil then to add 100M ohms in parallel then retune and take a second Q reading then calculate out the 100M ohm first reading. However most? measurements don't need such a process because the 100M ohm loading is <1% Q reduction.? If you have an excellent 100uH inductor? Xl=628 ohms at 1MHz, so for a Q=2000? the parallel loss will be 1.26M ohm . Adding 100M ohm in parallel to 1.26M produces 1.2443, a 1.26% drop from Q=2,000 to 1975.? In reality a large coil near to the metalwork of a Q meter will read low Q because of eddy currents, in my experience the Q drop due to eddy currents is often far greater than 1%!
Would some perform this experiment on their Q meter? ?I continuously get a about a 19%? low error. I have used two different Q meters. I'm attaching thin wires to an 0805 resistor. But please use what you have. How to: ?Put an inductor on your meter and measure the Q, and find the capacitance that was used to resonate. Record both. Now put a 10MegΩ across the inductor (retune if needed) and measure the new lower Q. Record. Use this formula to find the value of the measured value of the 10MΩ. Rp = Q1 x Q2 / 2 x pi x f x capacitance x delta Q. Capacitance is in Farads, i.e, 150pf = 0.000000000150. The delta Q is first Q measured minus the second Q measured. ? I want to know why I don't get an accurate answer. This method of measuring high value resistors is in both the Boonton 260A manual and the HP4342A manual. ? ?? Thanks for any incite you can add to this.? Mikek
|
Re: Ringdown Q Meter Built an d working
On Tue, Nov 15, 2022 at 05:46 AM, Leon wrote:
I have seen that before, I thought it was in the Files section, but I don't see it. I put the scope on my Ringdown Meter to look at the waveform. ?It gets a shake and ringsdown, at some point It gets another shake, I can't tell if that is at a specific time period after the cycle count or at a level. Btw, I'd have to count 220 cycles on an inductor with a Q of 1000, I'm glad the electronics do that for me. I don't really know if it measures at the 50% level though. ?????????????????????????????? Mikek ?Ignore the frequency, it is confused by the pulses, should read about 800kHz. Cursors are at 50% 
|
Re: Ringdown Q Meter Built an d working
|
Ringdown Q Meter Built an d working
I received the Ringdown Q meter Kit and have got it built into a case and working. The frequency range is 400kHz to 9.2MHz. Minimum inductance is 10uh, but I have 9.2uh and it worked. Here are some pictures of the build and tests of my 6" styrene pipe couple wound with 660/46 Litz wire. ?Fun Project. ???????????????????????????? Mikek   -- Thanks, Mikek
|
Re: More final comments on the Heathkit QM-1.
On Tue, 15 Nov 2022 at 12:49, Jeff Green < Jeff.L.Green1970@...> wrote: I bought the QM-1 on a whim. I don’t have any need for a Q-meter at my knowledge level. I was following the threads on Q-meters and when I saw the Heathkit for such a good price, I had to buy it. Worst case I was out a few?
That sounds great.?
I am a bit amazed at the dedication to measuring that last digit displayed herein. And that is not criticism. I appreciate the energy, determination, and, skill sets required.
A lot of people, myself included, enjoy trying to make something better, even if we don't really need it.
?
And here I thought I was doing pretty good when I bought a 4 and half digit DVM for accuracy!
That's good enough for most things. I have 6.5 digits, and that's good enough for everything. But I know a few people with 8.5 digits, but none need it. But if I could afford 8.5 digits, I would have one. Note each extra digit is a 10x increase in resolution, and unless that's accompanied by a 10x decrease in uncertainty, it is usually not that useful.
One suggestion - take a look at the first 5 sections of
to learn the difference between terms like error (which we can never know), uncertainty, accuracy and other related terms.
I would avoid buying semiconductors from places like eBay, because there are a lot of fakes around. It will be frustrating if you build something which does not work because the semiconductor you? are using is fake. A lot of old articles use obsolete devices which have not been made for years. But you will find Chinese sellers on eBay having thousands of them. You can be 95% sure they are fake.
Dave
_._
|
Re: Concluding comments on Heathkit QM-1 Q-Meter
I would agree that all the LCR meters I have seen may be good at measuring LCR but really bad with Q accuracy. My point to Jeff was if he needed a wider range of frequencies or range of Q, get something other than a QM-1.? VNAs are in vogue today.? My QM-1, for the frequency range it supports, seems to be as good as a Boonton 160 for accuracy. All these dedicated Q meters built last century are probably great for the last century but we need a new design, and new thinking, to get the accuracy we need today.
I understand the uncertainty issue.? We have measuring devices with large uncertainty ranges used to measure coils with large uncertainty ranges based on how they are attached to the measuring devices. Using a VNA is the modern method but it still has uncertainties. Even though it may be a very accurate device, it is still hampered by the interconnect with the device being measured especially at any high frequency.??
I would also maintain that even if you could build a coil and then measure the Q accurately, you still have to put it into a circuit, which is not electrically or environmentally the same as the measuring device, so the Q in circuit will be different than measured anyway.? The only way to get a real accurate measure of Q is by figuring out how to measure it accurately in the final circuit or with the circuits characteristics.? I would say we should be able to get a reasonable indication of Q through the measurements and calculations to help predict how it will work in a circuit, this coil is better than that or this method of winding, wire or trace should perform the best, but until it is tried in circuit, it will be hard to accurately measure what the Q will actually be.?
This is just my opinion but I don't see any commercial devices that can accurately predict what the Q of a coil will be once installed in a circuit. Though some might disagree, I think a major success metric might not be getting the Q measurement to a couple of decimal points initially but rather reducing the uncertainty down toward zero. Until the uncertainty is removed, the accuracy cannot be trusted.? I would like to see us get to three digits of accuracy but it does no good if the uncertainty of the measurement eclipses the value measured. If I measured a Q of 99 with an old Q meter and the same coil at 99.456 with a precision Q meter, based on even +/- 1% uncertainty, both will be about the same. The lower accuracy 99 measurement is somewhere between 100 (99.99) and 98 (98.1) while the precision accuracy values end up somewhere between 100.451 and 98.461. The decimal digits don't really help as the value is still around 99 in either test. Imagine this at +/- 10% uncertainty.
I have to wonder if creating a device that has some variable resistance/impedance on the input and output would more accurately characterize the Q and make the device able to better predict what the Q will be in circuit. Perhaps the Qmeter measures Q with two or three input/output resistances/impedances and uses this to filter out some of the uncertainty. This is a little different than a simple component Q tester but might be more useful for actually designing something.? Though it may seem a little off, I am trying to stay on topic to reach a way of getting to decimal digits that can be trusted and useful.? I would like us to get to a device we can use to measure Q which performs as well as the average DMM measuring resistors today.?
Hopefully this all makes sense.
|
Re: Concluding comments on Heathkit QM-1 Q-Meter
Yes Jeff, metal film resistors can be spiral cut during trimming.
But the inductance increase is usually very? small and in series
with a much higher resistance. The inductance of a wire ended
quarter Watt resistor is around 0.02uH? max so XL is around 0.1
ohms at 1MHz. For many metal film resistors the inductance is
smaller than 0.02uH - depends on the manufacturer and how they
trim. Some metal film wire ended resistors are not spiral cut.
Surface mount resistors can be trimmed with a short line, they
have very low inductance hardly altered by trimming.
The spiral trimming process has been around since the 1930's,
originally with carbon film resistors and abrasive loaded air
jets. Makers could use a standard rough coating process then
spiral cut the coating to make resistors of 10-20 times higher
value and charge more for higher precision. Generally don't worry
about the inductance of metal film wire ended resistors unless you
are using 100 ohms or less at frequencies above 10MHz.
Regards, Alan G8LCO
|
Re: Concluding comments on Heathkit QM-1 Q-Meter
Jeff,
general purpose round-bodied metal film (MF) resistors will often
have been trimmed with a spiral, but there are absolutely no
guarantees with this. Specialist resistors are usually trimmed
longitudinally, and this can be checked with the manufacturer's
data sheet. SMD resistors will usually be trimmed longitudinally,
but, again, this is not guaranteed. All of this is at the
manufacturer's discretion, and for a wide enough tolerance band
the manufacturer may simply not bother to trim. The manufacturer
may also mark resistors according to a measured value, which is
also an option for an end user who selects the required value from
a batch.
A resistor will possess, besides the quality of resistance an
amount of self-inductance and an amount of self-capacitance.
Depending on the frequency of operation, and the accuracy that is
required, then one is in to doing sums.
So yes, generic MF or metal oxide (metox) resistors will usually
suffer from an excess of self-inductance, as will carbon film
resistors. Carbon composition resistors do not, but have very
serious problems in the contexts of precision and stability.
When one is concerned with the self-inductance of a ?Watt
precision resistor, then the inductance and capacitance of wiring
and other components become serious issues.
And all of this is why test gear, and knowing exactly what it is
that one is measuring, and to what accuracy, is such fun/jolly
hard work/something of a physics experiment (delete as
applicable)…
HTH, 73, Stay Safe,
Robin, G8DQX
On 14/11/2022 10:10, Jeff Green wrote:
toggle quoted message
Show quoted text
My EE friend
sent me an email last night.
He
remembered modern metal film resistors are laser trimmed to
the desired value and believes the trimming might be in a
spiral, which would form an inductor. The coating on the metal
film resistors appears to be epoxy or fired ceramic.
He’s
concerned the added inductance might affect accuracy.
Anyone have
any ideas on that possibility?
?
Thank you
for your observations.
We’ll check
for wiring errors tomorrow.
|
Re: Concluding comments on Heathkit QM-1 Q-Meter
On Mon, 14 Nov 2022 at 08:50, tgerbic < tgerbic@...> wrote: Jeff,
If you need an accurate measurement of high Q values and much higher frequencies, get a Boonton or some more modern LCRQ meter/VNA.
I don’t think any LCR meter is particularly good at Q above 20. I have an HP 4285A (75 kHz to 30 MHz. Basic accuracy 0.1%). There’s an example in the manual of the uncertainty of a Q measurement at the following approximate values, where the Q displayed is 30
25 MHz 220 nH
The uncertainty is something like +107/-13, so the Q can be anywhere between 17 and 137.?
I am a bit of a restoration nut and have a bunch of HP, Tek, Heath, LeCroy, Fluke and other equipment. Not sure why we chase precision Q measurements other than it is one of the harder ones to nail down. Voltage, current, resistance and frequency measurements are pretty easy to build accurate equipment for but building an accurate wide range Q meter is like wrestling a greased pig.? I don't think the intent of this group is to get super accurate readings, such as Q to three decimal points, but rather to get a method/device that has a higher frequency and Q value range than the Boonton and HP gear we currently have, and reduce the sources of inaccuracy/uncertainty that plagues Q meters and Q measurements.
I set the group up. It’s description at includes the statement?
“?There are no limits to the complexity or cost of the design, so both a cesium fountain clock and a basic voltmeter are acceptable. items such as?"ghost detectors"?are?not?considered measurement devices. Measuring devices for hobby / amateur radio use are encouraged, as are those that exceed the capabilities of standard laboratories such as NIST, NPL, METAS etc.”
meters So if there are ways of getting to 3 decimal places, that’s on topic.?
It is great you have saved another QM-1. Have fun.
Dave? -- Dr. David Kirkby, Kirkby Microwave Ltd, drkirkby@...Telephone 01621-680100./ +44 1621 680100 Registered in England & Wales, company number 08914892. Registered office: Stokes Hall Lodge, Burnham Rd, Althorne, Chelmsford, Essex, CM3 6DT, United Kingdom
|
Saw a couple of HP 8640Bs in there.............
|
Brian
Leon
Dr. David Kirkby, Kirkby Microwave Ltd