On Sun, Nov 20, 2022 at 10:32 AM, John KN5L wrote:
Yesterday's post using SimSmith to solve for 11Mohm resistor capacitance
and Q by match Steve's Q measurements. Results were 0.4pF with Q=45.
How are you calculating the Q of the capacitor? ??????????????????????????????????????????????? Mikek ? -- Thanks, Mikek
|
Yesterday's post using SimSmith to solve for 11Mohm resistor capacitance and Q by match Steve's Q measurements. Results were 0.4pF with Q=45. Xc = 1/(2*Pi*1.7e6*0.4e-12) = 234kohm Rp = 234k x 45 = 10.53Mohm 11M || 10.53M = 5.38M On 11/16/22 12:07 PM, Steve Ratzlaff wrote: So plugging into Mike's formula: 980 x 635/ 2 pi 1.7e6 32.5e-12
345 = 5.196 megohms Safe to conclude Steve's 11Mohm resistor parallel resistance is between 5.4Mohm and 5.2Mohm at 1.7MHz? John KN5L On 11/19/22 4:43 PM, John KN5L wrote: On 11/19/22 1:47 PM, Jacques Audet wrote:
This resistor Rp comes with a capacitor in parallel Cp, the inter
electrode capacitance of the physical resistor.
This capacitor has a "poor" dielectric:? made by the actual resistance
material. On 11/16/22 12:07 PM, Steve Ratzlaff wrote:
1700 kHz, 32.5 pF, coil-alone Q 980. With resistor across coil, Q 635.
(32.1 pF) Attached is a SimSmith model of Steve's 11/16/22 measurement.
Set DUT Q to match measurement without 11M resistor.
Added 11M parallel resistor and adjusted CpC and CpQ for resonance and
Q, as shown in trace, Steve measured.
Modeling suggests 11M resistor has 400fF parallel capacitance with Q of 45.
John KN5L
|
On Sun, Nov 20, 2022 at 05:30 AM, Alan wrote:
The real issue of Q with additional resistors is the self capacitance of the inductor. If you look at the HP manual there is a nice graph showing the effect of self capacitance on indicated Q.
That is all true, but once you coil is built, the self capacitance will always be there, all you can do is build to minimize it. Don't you pretty much have calculate with indicated values when designing. ?About half way down this page is a program that calculates self capacitance from measurements you enter. You can do it with two data points, but this will allow at least 15 and give you inductance also. Oh, looks like coping the program name makes a link, the page has other useful programs, worth a look.
http://electronbunker.ca/eb/Downloads.html
It is a very real issue, the self capacitance can easily be around 30pF for a multi-turn close wound coil. So if you have 30pF of external tuning capacitance your resonance is at 0.707 of the true resonance frequency AND the inductance that you get from the resonant frequency? and the 30pF external tuning is a factor of 2 out !? Even with a fully meshed tuning capacitor the self capacitance is very significant with many inductors that have many turns of wire. It gets worse when the wire is close wound into many layers and becomes more random when scramble winding is used. Wave winding helps reduce internal C as does multiple piles.
? I think you meant to say pies, instead of piles. ???????????????????????? Mikek  ? -- Thanks, Mikek
|
The real issue of Q with additional resistors is the self
capacitance of the inductor. If you look at the HP manual
there is a nice graph showing the effect of self capacitance on
indicated Q. It is a very real issue, the self capacitance can
easily be around 30pF for a multi-turn close wound coil. So if you
have 30pF of external tuning capacitance your resonance is at
0.707 of the true resonance frequency AND the inductance that you
get from the resonant frequency? and the 30pF external tuning is a
factor of 2 out !? Even with a fully meshed tuning capacitor the
self capacitance is very significant with many inductors that have
many turns of wire. It gets worse when the wire is close wound
into many layers and becomes more random when scramble winding is
used. Wave winding helps reduce internal C as does multiple
piles.?
It is easy to add external capacitance and work out the self
capacitance of the inductor or you take a second resonance at a
different external tuning capacitance, this is why the manuals for
Q meters detail the self capacitance measuring methods.
The Q meter is an indicator of Q, when used without caution you
get misleading results. Every Q meter should have a Mental Health
Warning label " Inductor measurements without self capacitance
measurements may drive you nuts".
The high value resistor issue is very part type dependant, you
can't say that all resistors above x ohms and y MHz will show
different reactance at HF, you HAVE to say what kind of?
resistor.? A spiralled metal or carbon film resistor? has
distributed capacitance between the turns of the spiral , that
capacitance is the difference between a uniform uncut film and a
spiral cut film at high frequencies, a simple model for a 1 Meg
resistor could be ten 100K resistors in series each with a very
small capacitor in parallel.
A surface mount or a wire ended non-spiralled resistor generally
lacks the complexity of the spiralled part and are good to GHz
frequencies in sensible values.
Quick Reactance Calculations.
Hand waving descriptions are frequently very misleading, always
best to " do the numbers".? You only need remember:-
1uF = 159 ohms @ I KHz (? Xc= 1/ 2*pi*C*F)
Then? 1nF =159 ohms at 1MHz? and? 1pF =159 ohms at 1 GHz
So from 1nF at 1MHz ,? 250pf? at? 500KHz? =? 159? *4 ( cap is
1/4) *2 ( FX is 1/2) =159 *8= 1272 ohms
Inductance is even easier, 1uH=6.28 ( 2 pi) ohms at 1 MHz??
( Xl=2*pi*L*F)?? so 100uH @ 600KHz = 6.28 * 100? * 0.6=376.8 ohms.
Resonance?? LC= 25330/F*F???? or? F = SQR? 25330/LC? or ?
L= 25330/F*F*C ? or L=25330/F*F*C? where F=MHz, L=uH, C=pF
For a sanity check or to be pure lazy use the Kits and Parts
Toroid calculator?
Pop in two from L,C or F then calc the other and the basic Xc=Xl?
comes out as well. ( the link is for turns on a T68-6 core but
ignore the turns, the F,C,L and Z numbers will be good.)
Regards, Alan G8LCO
Specs for T68-6 RF Toroids | Physical Dimensions
OD = .690 in / 17.5 mm +/- 0.02 in
ID = .370 in / 9.40 mm +/- 0.02 in
Ht = .190 in / 4.83 mm +/- 0.02 in | | ?AL=4.7 +/- 5 % ???? uH=(AL*Turns2)/1000? | | Temperature Stability (ppm /∼C) = 35 | | Color Code = Yellow / Clear | Optimum Resonant Circuit Range
for highest Q and lowest core loss
3 MHz - 40 MHz | Orders and Pricing
|
|
Freq/L/C/Z/Turns Calculator for T68-6
Includes 1 inch / 2.5 cm pig-tails
| MHz
| uH
| pF
| ohms
| turns
| inches - cm
| ?
| ?
|
enter uH to Calc number of turns, or
enter number of turns to Calc uH, or
enter two (2) items: MHz, uH, pF, ohms or turns to Calc all values. |
Impedance Matching Network Calculator using a T68-6 where Z2 > Z1
Enter F(MHz), Z1(?) and Z2(?) below, then click the Calc button
 | | XL = sqrt ( Z1*Z2 - Z1^2 ) ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?XC = Z1*Z2 / XL | F(MHz)
| Z1(?)
| Z2(?)
| L(uH)
| L(turns)
| L(inches - cm)
| C(pF)
| ?
| ?
|
Impedance Matching Network Calculator using a T68-6 where Z2 > Z1
Enter F(MHz), Z1(?) and Z2(?) below, then click the Calc button
 | | XL = Z2 * sqrt ( Z1 / (Z2 - Z1) ) ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?XC = Z1*Z2 / XL | F(MHz)
| Z1(?)
| Z2(?)
| L(uH)
| (Lturns)
| L(inches - cm)
| C(pF)
| ?
| ?
|
|
|
|
|
Re: How to make a transistor, off topic and I'll delete this in a few days
Here is the image I tried to insert between the dashes in the previous message. The attach image freaked out, returned a bunch of errors and posted the message without the image.
Tom, wb6b
toggle quoted message
Show quoted text
On Sun, Nov 20, 2022 at 01:23 AM, Tom, wb6b wrote:
But in one way we have really gone backwards in technology. In 1948 you could buy a radio to install in your automobile, that not only could charge your car's battery, but could power your electric shaver. Where are radios like that today. And to think that could do that way Bach in 1948. What is holding this technology back?
|
Re: How to make a transistor, off topic and I'll delete this in a few days
On Sat, Nov 19, 2022 at 01:18 PM, Jeff Green wrote:
You can use those as described in:
?
Fascinating. Spent about an hour going through the articles and especially the ads. Wow, TVs ranged from $99 kit ($1,200 in todays dollars) to some kind of small rear projection TV that would cost around $12,000 in today's money. No wonder very few people had TVs back then. But in one way we have really gone backwards in technology. In 1948 you could buy a radio to install in your automobile, that not only could charge your car's battery, but could power your electric shaver. Where are radios like that today. And to think that could do that way Bach in 1948. What is holding this technology back? ------ ------ By the way I recently read an article, and have tried it out, in reproducing the first LED discovered around 1912.? Take a chunk of silicon carbide, wrap a wire around it and hook it to a power supply positive terminal. Put a small wire or nail on the Negative lead, probe around and you will see spots where the crystal will glow. It is an actual semiconductor junction converting electrons to photons; not heat.? My experiment required around 24 VDC (Three 9v batteries work great) at around 50 MA to light up. I used a small 24 volt incandescent lamp as a safety current limiter. But, really cool. Some the the crystal formations glow on the edge, giving a asteroid space ship window effect.? Fortunately the efficiency of purpose made LEDs has become a lot higher than the accident of nature LED junctions in a silicon carbide crystal. Tom, wb6b
|
On Sat, Nov 19, 2022 at 05:40 PM, Mikek wrote:
I started another drawing, but decided I'm to tired to think now, But the new Idea is to go via crazy to making with a ring that is at ground potential on the opposite side of the PCB to ground the 7 smd resistors to. See picture.
? -- Thanks, Mikek 
|
On Sat, Nov 19, 2022 at 03:26 PM, John Kolb wrote:
Nothing magic about 7 resistors. I used it because it was easy to draw in Visio with resistors on 45 deg slants to make 8 in a circle.
Yep, I looked up the link and saw they hav0.140 resistor so, no need to do anything different, there is enough confusion! :-)
?
?
There is a type of SMT component,? 1020(reversed), 0915(reversed), 0612(reversed), etc, with the solder pads along the long edge rather than the usual short edge.? These might be much lower inductance, so fewer used.? Longer solder joint would mean lower added resistance.
?
The 260 has transformer drive (coupling coil on the RF oscillator tank) driving about 3 ohms (the thermocouple) in series with the 0.02 ohm drive source resistor, so transformer coupling into a resistive diver is old news? :)?
?
We only need 1 amp to get? 20mV across 20 milliohms, but reducing the current out of the power amp with a 2:1, 4:1, etc transformer would be good.?
Yes, but, wouldn't you design for a little extra to cover losses? I would, even if it's only 20%.
?
?
Your PCB is an interesting concept, but my hand shakes when I reach forward with a soldering iron, so I'm not going to be soldering parts on the edge of the PCB.?? The GND banana jack could be coupled to the top etch by replacing the plastic insulator with a metal washer or by wrapping a copper strap around the edge of the PCB to the back side to reach the plug mounting nut or both.?? Not sure how you would use this.? It supplies the GND and LO terminals of a Q meter but you still need to connect the cap and voltmeter.? The circle doesn't lend itself to mounting to the top of a Q meter next to the capacitor/voltmeter terminals.
I started another drawing, but decided I'm to tired to think now, But the new Idea is to go via crazy to making with a ring that is at ground potential on the opposite side of the PCB to ground the 7 smd resistors to. See picture.
?
?
Here's a RF voltmeter circuit which would work well to monitor a 20 mV level.?
I don't see any frequency range, it may only be useful in the broadcast band. TL071 has a gain bandwidth product of 3 MHz.
?
Ignore the part where it's used as a demodulator driving an audio amp. ? If used ad a voltmeter with a meter, one would have to look at a correction chart for voltages below 100 mv, but the correction factor doesn't matter if used to detect a fixed voltage level for an AGC loop.? The H11F1 should work if used between the signal generator and the? low output impedance amp.
? -- Thanks, Mikek
|
Hi Mike,
Nothing magic about 7 resistors. I used
it because it was easy to draw in Visio with resistors on 45 deg
slants to make 8 in a circle.
There is a type of SMT component,?
1020(reversed), 0915(reversed), 0612(reversed), etc, with the
solder pads along the long edge rather than the usual short edge.?
These might be much lower inductance, so fewer used.? Longer
solder joint would mean lower added resistance.
The 260 has transformer drive (coupling
coil on the RF oscillator tank) driving about 3 ohms (the
thermocouple) in series with the 0.02 ohm drive source resistor,
so transformer coupling into a resistive diver is old news? :)?
We only need 1 amp to get? 20mV across
20 milliohms, but reducing the current out of the power amp with a
2:1, 4:1, etc transformer would be good.?
Your PCB is an interesting concept, but
my hand shakes when I reach forward with a soldering iron, so I'm
not going to be soldering parts on the edge of the PCB.?? The GND
banana jack could be coupled to the top etch by replacing the
plastic insulator with a metal washer or by wrapping a copper
strap around the edge of the PCB to the back side to reach the
plug mounting nut or both.?? Not sure how you would use this.? It
supplies the GND and LO terminals of a Q meter but you still need
to connect the cap and voltmeter.? The circle doesn't lend itself
to mounting to the top of a Q meter next to the
capacitor/voltmeter terminals.
Here's a RF voltmeter circuit which
would work well to monitor a 20 mV level.?
Ignore the part where it's used as a
demodulator driving an audio amp. ? If used ad a voltmeter with a
meter, one would have to look at a correction chart for voltages
below 100 mV, but the correction factor doesn't matter if used to
detect a fixed voltage level for an AGC loop.? The H11F1 should
work if used between the signal generator and the? low output
impedance amp.
John ? KK6IL
On 11/19/2022 6:47 AM, Mikek wrote:
toggle quoted message
Show quoted text
Hi John, ?
I like the 7 smd resistor idea, I went to bed thinking about
how to make it work.
?I didn't think of much last night, I was already out of
mental energy when I wrote back,
I had to check the logic of what I wrote 3 times and still, I
was hesitant to send it.
? It will help to skip the 50次 resistor and just have a low
impedance current source,
that has a little more than 0.1 volt output while driving
0.02次.
? As I recall the HP4342A injection drive circuit was found to
have about 0.2次 output impedance, (rising with freq)
I = E/R = .1V / .02次 = 5amps (Wow!).
? What happens if we insert a 10 to 1 transformer? 1V input at
500Ma is transformed to 0.1V at 5 amps.
The 0.02次 load reflected back through the transformer is 2次,
seems easier to drive 2次 at 1V.
?Just throwing out ideas. Since we have smd parts vs the large
platinum plated disk* that Boonton honed to have
the proper resistance without inductance, it maybe a better
way.
? But now we have inserted components that will have
resistive, coupling, inductive and other characteristic to
cause the injection voltage to vary with load and frequency.
So I'm back to monitoring the injection voltage and
?and controlling the level of the transformer input voltage.
?The H11F1 won't do the job at these low impedances, so...
another transformer to step up the impedance?
? Hrmph! If it was easy, it wouldn't be any fun to think
about.
?I have attached another implementation of the 7 SMD low
inductance PCB, for your critic.
?An after drawing thought, need to mount bananas on sturdy
insulative plate, then solder pcb to banana connectors.
??????????????????????????????????????????????? Later, Mikek
?*I read towards the end, with employee turnover at Boonton,
they had lost the technical details of how to
properly make the disc and they lost some accuracy.We only
need
On 11/19/2022 1:08 AM, John Kolb
wrote:
Oops, SMT resistors
in a circle directly around the binding post post can make a
very low inductance 20 milliohm resistor, but a 2500:1
divider is probably not very useful.
John?? KK6IL
On 11/18/2022 7:33 PM, Mikek wrote:
I went with you your
numbers, seven 0.140次 resistors in parallel is 0.02次.
Oh, I see, a 50次 to 0.02 resistive divider is 2500 to 1 ,
not the 50 to 1 I said.
So 1V into a 50次 to 0.02次 voltage divider has an output of
0.5mV
?A transformer with a 50 to 1 ratio, converts 50次 to
0.02次, (2500 to 1) but has 20 mV output with one volt in.
??????????????????????????????????????? Later,? Mike
On 11/18/2022 9:09 PM, John Kolb
wrote:
Generating 20
mV drive level with a 50:1 divider only requires 1V RMS,
2.828 P-P, which I would expect most signal generators
would do, or not hard for a DIY generator.? Higher drive
levels for very low Q coils would be a problem.? The
virtue of a low inductance resistor array like I drew is
that the resistor values could be much lower than 0.02
ohms.? 1 milliohm anyone?? That rapidly creates
challenges of either drive power or voltmeter
sensitivity or both.
John?? KK6IL
?
On 11/18/2022 3:32 PM, Mikek
wrote:
On Fri, Nov 18, 2022 at 02:16 PM, John Kolb wrote:
The author's approach is not to cancel the
inductance of the 0.2 Ohm (not 0.02)
Oh, he does say 0.2次--- Now I don't like it :-)? I've
going through the Boonton 260A manual and have the 0.02次
etched in my mind.
?I see one problem with the resistive 50 to 1 voltage
divider, it is the low voltage, it is 1/50th of what a 50
to 1 transformer would be.
It's always something!
??????????????????? Mikek
?
--
Thanks, Mikek
--
Thanks, Mikek
--
Thanks, Mikek
|
On 11/19/22 1:47 PM, Jacques Audet wrote: This resistor Rp comes with a capacitor in parallel Cp, the inter
electrode capacitance of the physical resistor.
This capacitor has a "poor" dielectric:? made by the actual resistance
material. On 11/16/22 12:07 PM, Steve Ratzlaff wrote: 1700 kHz, 32.5 pF, coil-alone Q 980. With resistor across coil, Q 635.
(32.1 pF) Attached is a SimSmith model of Steve's 11/16/22 measurement. Set DUT Q to match measurement without 11M resistor. Added 11M parallel resistor and adjusted CpC and CpQ for resonance and Q, as shown in trace, Steve measured. Modeling suggests 11M resistor has 400fF parallel capacitance with Q of 45. John KN5L
|
On Sat, Nov 19, 2022 at 11:47 AM, Jacques Audet wrote:
What caused Rp to fall ?
Good question !
One would expect that the resistance would increase with frequency, as predicted by the skin effect.
Yep, my thinking was backwards, skin effect would cause the value to rise.
But this shows up on conductors or "low" value resistors.
Resistors above the 100K range do not follow these rules.
This resistor Rp comes with a capacitor in parallel Cp, the inter electrode capacitance of the physical resistor.
This capacitor has a "poor" dielectric:? made by the actual resistance material.
As with many dielectrics, the losses increase with frequency.?
?OK, I was blaming Cp before but with the wrong reasoning. Help me understand, I still believe we tune out the capacitance of the resistor, are you saying we do but, the dielectric losses of Cp acts as a large resistance in parallel with the 10M次?
The Q of the capacitor shows a decreasing slope
at higher frequencies.? See my attached doc.
(If the Rp was constant, then the Q would always increase with frequency)
I also checked a 10 Mohms resistor.
It shows that the resistance Rp decreases at much lower frequencies and to quite low values.
Same for the Q which peaks at 3 MHz, instead of 20 MHz for the 1 Mohm resistor.
? Very Cool, thanks for the testing. Looking at the 3rd graph (capacitance and Rp of a 10M次) The 10M次 resistance drops to 6M次. ? Is it reasonable to assume a 10M次 0805 smd resistor which has lower Cp, would have less dielectric loss and and therefore would drop less. ? My measurements are fairly consistently around 8M次. ?? I think you have provide the info I needed to calm my mind! ??????????????????????? Thank you.
Testing high impedance amplifiers can be done with the Q meter, connected across a tuned circuit.
This allows for measuring high Rp values on a device with one side connected to ground.
Take care not to overdrive/saturate the amplifier, since that may influence/lower the measured Rp.
?I think I may have run into that on one of my tests, I added a 0.3pf cap on the input which made a very close to a 10 to 1 divider with the FET gate capacitance. And the measurement acted better. ?
Or only check the input stage.
To improve resolution, use a high Q coil and connect a DVM across the Q meter's meter.
Find the relation between the Q reading and the DC volts on the DVM and use it to compute the Q
with better resolution.? Like if you read 200 mVDC for a Q of 600, then Q=VDC*3.
Beware of the published input resistance specs.? Assume it's at DC, unless otherwise.
?A word about that, I have a Boonton 260A, in order to measure high Q inductors, I need to decrease the 1X injection voltage of 20mV down to 4mV, so I have 5X. I then need to multiply the Q meter reading by 5. ?The Q meter is demarked to 250 with 5 unit divisions, not terribly hard to be within 1 unit when reading the meter. It may be worth bring some leads out for my DVM. ? The problem with the very high input impedance is it changes Q Very little even with an inductor that has a Q over 1000. OK, yes, good idea to connect the DVM :-) ? I would like to know the input impedance of an amp from 500kHz to 30MHz, I think I would need 3 or 4 high Q inductors and some high Q capacitors to get that range. That will be interesting. ? Thanks again for all of the info. ??????????????????????????????? Mikek
Jacques,
VE2AZX
? -- Thanks, Mikek
|
Hi Mikek,
What caused Rp to fall ?
Good question !
One would expect that the resistance would increase with frequency, as predicted by the skin effect.
But this shows up on conductors or "low" value resistors.
Resistors above the 100K range do not follow these rules.
This resistor Rp comes with a capacitor in parallel Cp, the inter electrode capacitance of the physical resistor.
This capacitor has a "poor" dielectric:? made by the actual resistance material.
As with many dielectrics, the losses increase with frequency.? The Q of the capacitor shows a decreasing slope
at higher frequencies.? See my attached doc.
(If the Rp was constant, then the Q would always increase with frequency)
I also checked a 10 Mohms resistor.
It shows that the resistance Rp decreases at much lower frequencies and to quite low values.
Same for the Q which peaks at 3 MHz, instead of 20 MHz for the 1 Mohm resistor.
Testing high impedance amplifiers can be done with the Q meter, connected across a tuned circuit.
This allows for measuring high Rp values on a device with one side connected to ground.
Take care not to overdrive/saturate the amplifier, since that may influence/lower the measured Rp.
Or only check the input stage.
To improve resolution, use a high Q coil and connect a DVM across the Q meter's meter.
Find the relation between the Q reading and the DC volts on the DVM and use it to compute the Q
with better resolution.? Like if you read 200 mVDC for a Q of 600, then Q=VDC*3.
Beware of the published input resistance specs.? Assume it's at DC, unless otherwise.
Jacques,
VE2AZX
|
Re: How to make a transistor, off topic and I'll delete this in a few days
Lester B Veenstra K1YCM M?YCM W8YCM 6Y6Y W8YCM/6Y 6Y8LV (Reformed USNSG CTM1)
lester@...
452 Stable Ln
Keyser WV 26726 USA
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
-----Original Message----- From: [email protected] [mailto: [email protected]] On Behalf Of cheater cheater Sent: Saturday, November 19, 2022 2:30 PM To: [email protected]Subject: Re: [Test Equipment Design & Construction] How to make a transistor, off topic and I'll delete this in a few days Where did you obtain the germanium crystal? On Thu, Nov 17, 2022 at 2:58 PM Jeff Green <Jeff.L.Green1970@...> wrote:
I received an email asking how our daughter made a point contact transistor.
She started with the description in:
Look on page 39.
|
Re: How to make a transistor, off topic and I'll delete this in a few days
Where did you obtain the germanium crystal?
toggle quoted message
Show quoted text
On Thu, Nov 17, 2022 at 2:58 PM Jeff Green <Jeff.L.Green1970@...> wrote:
I received an email asking how our daughter made a point contact transistor.
She started with the description in:
Look on page 39.
|
On Sat, Nov 19, 2022 at 07:27 AM, Mikek wrote:
?I can understand why strays would cause a change of impedance with increasing frequency, but as I understand it,
?your graph shows Resistance changes with frequency. What is the cause of that change?
Oh, I just found a reference to skin affect in resistors, could that explain all if the change? ? -- Thanks, Mikek
|
On Fri, Nov 18, 2022 at 12:01 PM, Jacques Audet wrote:
Please take note that high value resistors have their resistive part decreasing with increasing frequency.
?I can understand why strays would cause a change of impedance with increasing frequency, but as I understand it, ?your graph shows Resistance changes with frequency. What is the cause of that change? ?Link to graph. ? -- Thanks, Mikek
|
I got confus  ed again, I thought we had a private conversation going, but I guess all can chip in now!
Hi John, ?
I like the 7 smd resistor idea, I went to bed thinking about how to make it work.
?I didn't think of much last night, I was already out of mental energy when I wrote back,
I had to check the logic of what I wrote 3 times and still, I was hesitant to send it.
? It will help to skip the 50次 resistor and just have a low impedance current source,
that has a little more than 0.1 volt output while driving 0.02次.
? As I recall the HP4342A injection drive circuit was found to have about 0.2次 output impedance, (rising with freq)
I = E/R = .1V / .02次 = 5amps (Wow!).
? What happens if we insert a 10 to 1 transformer? 1V input at 500Ma is transformed to 0.1V at 5 amps.
The 0.02次 load reflected back through the transformer is 2次, seems easier to drive 2次 at 1V.
?Just throwing out ideas. Since we have smd parts vs the large platinum plated disk* that Boonton honed to have
the proper resistance without inductance, it maybe a better way.
? But now we have inserted components that will have resistive, coupling, inductive and other characteristic to
cause the injection voltage to vary with load and frequency. So I'm back to monitoring the injection voltage and
?and controlling the level of the transformer input voltage.
?The H11F1 won't do the job at these low impedances, so... another transformer to step up the impedance?
? Hrmph! If it was easy, it wouldn't be any fun to think about.
?I have attached another implementation of the 7 SMD low inductance PCB, for your critic.
?An after drawing thought, need to mount bananas on sturdy insulative plate, then solder pcb to banana connectors.
??????????????????????????????????????????????? Later, Mikek
?*I read towards the end, with employee turnover at Boonton, they had lost the technical details of how to
properly make the disc and they lost some accuracy.
-- Thanks, Mikek
|
SimSmith Injection Transformer Q-Meter model attached.
Block L = Voltmeter with 100Mohm input impedance
Block B = inductor DUT with added Parallel resistor.
Block A = 50:1 Injection Transformer
Using Total Q = 1/(1/Qdut + 1/Qrp + 1/Qvm)
There's a little difference in computed Q using voltmeter shown in trace.
This model can be used to check other Q equations.
John KN5L
|
Oops, SMT resistors in
a circle directly around the binding post post can make a very
low inductance 20 milliohm resistor, but a 2500:1 divider is
probably not very useful.
John?? KK6IL
On 11/18/2022 7:33 PM, Mikek wrote:
toggle quoted message
Show quoted text
I went with you your
numbers, seven 0.140次 resistors in parallel is 0.02次.
Oh, I see, a 50次 to 0.02 resistive divider is 2500 to 1 , not
the 50 to 1 I said.
So 1V into a 50次 to 0.02次 voltage divider has an output of
0.5mV
?A transformer with a 50 to 1 ratio, converts 50次 to 0.02次,
(2500 to 1) but has 20 mV output with one volt in.
??????????????????????????????????????? Later,? Mike
On 11/18/2022 9:09 PM, John Kolb
wrote:
Generating 20 mV
drive level with a 50:1 divider only requires 1V RMS, 2.828
P-P, which I would expect most signal generators would do,
or not hard for a DIY generator.? Higher drive levels for
very low Q coils would be a problem.? The virtue of a low
inductance resistor array like I drew is that the resistor
values could be much lower than 0.02 ohms.? 1 milliohm
anyone?? That rapidly creates challenges of either drive
power or voltmeter sensitivity or both.
John?? KK6IL
?
On 11/18/2022 3:32 PM, Mikek wrote:
On Fri, Nov 18, 2022 at 02:16 PM, John Kolb wrote:
The author's approach is not to cancel the
inductance of the 0.2 Ohm (not 0.02)
Oh, he does say 0.2次--- Now I don't like it :-)? I've going
through the Boonton 260A manual and have the 0.02次
etched in my mind.
?I see one problem with the resistive 50 to 1 voltage divider,
it is the low voltage, it is 1/50th of what a 50 to 1
transformer would be.
It's always something!
??????????????????? Mikek
?
--
Thanks, Mikek
--
Thanks, Mikek
|
I went with you your numbers,
seven 0.140次 resistors in parallel is 0.02次.
Oh, I see, a 50次 to 0.02 resistive divider is 2500 to 1 , not
the 50 to 1 I said.
So 1V into a 50次 to 0.02次 voltage divider has an output of 0.5mV
?A transformer with a 50 to 1 ratio, converts 50次 to 0.02次,
(2500 to 1) but has 20 mV output with one volt in.
??????????????????????????????????????? Later,? Mike
On 11/18/2022 9:09 PM, John Kolb wrote:
toggle quoted message
Show quoted text
Generating 20 mV
drive level with a 50:1 divider only requires 1V RMS, 2.828
P-P, which I would expect most signal generators would do, or
not hard for a DIY generator.? Higher drive levels for very
low Q coils would be a problem.? The virtue of a low
inductance resistor array like I drew is that the resistor
values could be much lower than 0.02 ohms.? 1 milliohm
anyone?? That rapidly creates challenges of either drive power
or voltmeter sensitivity or both.
John?? KK6IL
?
On 11/18/2022 3:32 PM, Mikek wrote:
On Fri, Nov 18, 2022 at 02:16 PM, John Kolb wrote:
The author's approach is not to cancel the
inductance of the 0.2 Ohm (not 0.02)
Oh, he does say 0.2次--- Now I don't like it :-)? I've going
through the Boonton 260A manual and have the 0.02次
etched in my mind.
?I see one problem with the resistive 50 to 1 voltage divider,
it is the low voltage, it is 1/50th of what a 50 to 1
transformer would be.
It's always something!
??????????????????? Mikek
?
--
Thanks, Mikek
--
Thanks, Mikek
|
John Kolb
