MD Physics

[green]

Gotta admit, I don't like the idea of soldering a loop of high-conductivity copper up with a random blob of low-conductivity lead alloy. Maybe it was not leaded solder, but lead-free? Was it a neat butt joint, with flat faces, or did you just chop the wire with cutters and leave that distinctive chopped taper etc? Did you file the solder down afterwards, or leave a spherical ball on the join ? Etc Etc?

Anyway: Case C , with the 10a + 9b bit in the formula is no use for single turns, ... it's for solenoids, I think?
You've obviously got L calculation problems, you're miles out, but your R value looks good.
Here's my maths for the 40mm coil:
Loop resistance = 3.25 mOhms ( using 1.68 x 10^-8 for resistivity )
Loop inductance = 97 nH = 0.097 microhenry
So Time-constant = L/R = 29.9 microsecs. Which fits nicely with the measured value of 33 usec.

The effect of the solder join could be investigated by making rings with 1, 2, 3, 4, 5 joints, and seeing how TC varies ( increases) ... the trouble is making a perfect cut without removing much metal.

I'm beginning to regret not starting the 'Ring target thread' now, we're going to have to copy stuff over.

Did you use https://technick.net/tools/inductanc...circular-loop/ to calculate inductance?

Used kester 60/40 solder. neat butt joint, with flat faces, spherical ball on the joint. AWG19 magnet wire.

[green]

The effect of the solder join could be investigated by making rings with 1, 2, 3, 4, 5 joints, and seeing how TC varies ( increases) ... the trouble is making a perfect cut without removing much metal. Why increase? Adding resistance should decrease TC.

Made another 80mm diameter coil(coil B). Recorded coil A and coil B, 1 join. Cut coil B and joined with 2, 3 and 4 joins. Recorded after each join. My cutter cuts a flat on one side of cut, then cut other side flat. Loose about 1mm cutting other side flat for each of the 2, 3 and 4 joins. Don't see much change in TC.

[Skippy]

Your coils look OK, I guess the resistive loss caused by the solder that's between the ends of the copper is offset by the extra solder around the outer of the wire? I guess for large loops like 40mm+ the contribution to R from the solder is small.
Example: if solder has conductivity of 10% IACS, then 0.1mm of solder in the joint is like 1mm of copper, which out of, say 120mm ( Pi x 40) is only a small error.
Edit: that increase/decrease thing is bad proof-reading ... I originally typed R increased, then I changed it to TC, but forgot to change the 'increased' to 'decreased'.

And no, I didn't use the online calculator, I did it properly on a pocket calculator. Now you've forced me to turn on javascript and try the TechNick thing out, I can confirm it does in fact work.
R = 0.02
a = 0.0004505
N = 1
u_r = 1

gives L = 9.7328..... x 10e-8 = 97 nH, as previously.

I should ask: is 40mm the mean diameter ? If it's the inner diameter, you'll need to adjust the calculations.

[green]

Your coils look OK, I guess the resistive loss caused by the solder that's between the ends of the copper is offset by the extra solder around the outer of the wire? I guess for large loops like 40mm+ the contribution to R from the solder is small.
Example: if solder has conductivity of 10% IACS, then 0.1mm of solder in the joint is like 1mm of copper, which out of, say 120mm ( Pi x 40) is only a small error.
Edit: that increase/decrease thing is bad proof-reading ... I originally typed R increased, then I changed it to TC, but forgot to change the 'increased' to 'decreased'.

And no, I didn't use the online calculator, I did it properly on a pocket calculator. Now you've forced me to turn on javascript and try the TechNick thing out, I can confirm it does in fact work.
R = 0.02
a = 0.0004505
N = 1
u_r = 1

gives L = 9.7328..... x 10e-8 = 97 nH, as previously.

I should ask: is 40mm the mean diameter ? If it's the inner diameter, you'll need to adjust the calculations.

Thanks. Stupid pill, kept making the same error, needed another .0 in wire radius. Calculates about 10% lower than my measurements(L=TC*R) for the three coils. Don't know where the difference is but close enough.

I should ask: is 40mm the mean diameter? If it's the inner diameter, you'll need to adjust the calculations. Wire was cut to length, then butted and soldered.

[Skippy]

You can enter the radius values in scientific notation, just copy the format that the output L value uses.

I've often wondered if copper wire actually has the resistivity figures it 'should' have. After all, repeated drawing processes completely changes the crystal structure, eliminating faults, void, cracks, and elongating each individual crystal. It wouldn't surprise me if those IACS figures of 103% you see on some data tables for 'annealed' copper are for wire, processed by heat.
Some info on the topic here:
https://www.atlascables.com/design-conductors.html

( drawing stainless steel billet into 'wire' for bicycle spoke purposes makes a significant change to the strength: Bulk S.S has ultimate tensile strength about 400 MPa, a regular 2.0mm spoke is at 1200 MPa, and fancy double-butted ones have a 1.5mm section that's 1500 MPa.
So it's surprising how little the electrical characteristics of wire vs bulk differ.)

[green]

Thanks. Stupid pill, kept making the same error, needed another .0 in wire radius. Calculates about 10% lower than my measurements(L=TC*R) for the three coils. Don't know where the difference is but close enough.

I should ask: is 40mm the mean diameter? If it's the inner diameter, you'll need to adjust the calculations. Wire was cut to length, then butted and soldered.

Woke more than once last night wondering where the 10% difference is. Not close enough. Maybe doing something else stupid, not the first time I got the decimal point in wrong place with metric. Target: cut wire to length(flat ends), wrap in circle, solder joint. Thinking of some other things to try. Wire length 2mm shorter or longer. AWG16 or AWG28 magnet wire. Wondering if there is another way to measure inductance. I measure coil inductance by connecting measured capacitor across coil and excite resonance with my bench circuit. Wondering if I solder a ceramic chip capacitor in series with the single loop, should the wire be cut the same size as above or cut shorter by the length of the capacitor before soldering? Not as sharp in my old age as I was, new more about what I was trying to measure and would catch the stupid mistakes. Some possible reasons for the difference, log amplifier not correct, test procedure, measurement errors, others? I've been using https://www.powerstream.com/Wire_Size.htm for wire resistance. Any thoughts what to try or where I might be making an error or another stupid mistake? 80mm coil has less of a difference than the 40mm coil, 40mm or 80mm coil for further testing? 10% probably close enough since it has nothing to do with metal detecting. Not close enough if I'm trying to learn something. If my TRT tester isn't correct I would like to fix it.

[Skippy]

I have some ideas, will post later.
Do you have an accurate L meter that will measure 200 - 2000 uH correctly ?

[green]

I have some ideas, will post later.
Do you have an accurate L meter that will measure 200 - 2000 uH correctly ?

Don't have a L meter. Have 2 multimeters that measure capacitance, think I get a good capacitance measurement. Is there some reason calculating inductance with resonance decay and correct capacitance would give a wrong inductance? The 40mm single loop coil has an inductance of about .1uH, is there a good way to measure it's inductance?

[Skippy]

I can't believe you haven't got an L meter, with all those coils you've made etc, I assumed you must have one. ?
The 'usual' way to measure tiny L is put it into an oscillator with precision caps ( mica, polystyrene ) and measure the freq with a good DFM, not easy, and potentially lots of errors.
If you're thinking your test rig is suspect, perhaps restart the 'test rig' thread. But without some kind of 'standard' targets, you're not going to get very far, and having no L meter is a major handicap.

[green]

I can't believe you haven't got an L meter, with all those coils you've made etc, I assumed you must have one. ?
The 'usual' way to measure tiny L is put it into an oscillator with precision caps ( mica, polystyrene ) and measure the freq with a good DFM, not easy, and potentially lots of errors.
If you're thinking your test rig is suspect, perhaps restart the 'test rig' thread. But without some kind of 'standard' targets, you're not going to get very far, and having no L meter is a major handicap.

I've been watching some of the replies about L meters. Some like what they purchased, some not. L meter doesn't measure SRF. Would need another tester to measure SRF. My test circuit measures both. How accurate I don't know. Carl suggested in another thread my method might not be best because capacitor tolerance might be 10 to 20%. Don't know why my multimeter measuring capacitance wouldn't be as accurate as a L meter measuring inductance. I have two different multimeters that read the same value for capacitors I use(within 1%). I do have the error reading frequency which I wouldn't have with a L meter. L=2.46E6/f^2 for 9.9nf cap and 2.53E5/f^2 for 100nf cap. Searched resonance decay, appears resonance lowers with increased R(LCR circuit), didn't find a formula for how much. Tried with spice and didn't see much drop with low damping. Anyone know the formula? Including scope pictures measuring SRF and L for a coil.

Tried soldering a 10nf capacitor(measured 9.9nf)in series with 125mm AWG19 magnet wire and measured resonance with my tester. Calculates .1044uH, Technick calculates .0974uH, L=TC*R=1.1uH. Cut the wire 125mm with the added capacitor instead of 125.7(approximate) when measuring TC, not sure what length I should have cut.

[Skippy]

Lots of questions.
Here's the maths for damped LC circuits, lots of it is hideous, but the 'change in resonant frequency' formula is simple. If damping is reasonably light, the change in freq ( to a lower freq) is very small. Your 40mm diam loop calculates out at Q = 1000, ( X_L = 3 Ohms, R = 3 mOhm ), so no change in freq there.
https://en.wikipedia.org/wiki/RLC_circuit

One potential problem is your C value. a 10n in an 0805 package is going to be pretty temperature-sensitive, a few % drift is easy, plus there's the unknown factor of the effect of heating the little blighter to solder it in. And what's the series inductance of the cap? Obviously pretty low, but if it's just 5 nanoHenry, then add that to 97nH, you see how errors of a few % creep in.

To accurately assess your target tester, you need to perform readings on many targets, of differing TC. Then analyse the results,eg by straight-line graph plotting, to show up rogue values, trends, offsets etc.
Maybe if we could come up with a selection of stuff that we both have, that would be a start. There's probably some benefit to mailing stuff across the pond to each other, I don't expect our Royal Mail is going to be cheap .... I'll get some up-to-date price guides.

[green]

Lots of questions.
Here's the maths for damped LC circuits, lots of it is hideous, but the 'change in resonant frequency' formula is simple. If damping is reasonably light, the change in freq ( to a lower freq) is very small. Your 40mm diam loop calculates out at Q = 1000, ( X_L = 3 Ohms, R = 3 mOhm ), so no change in freq there.
https://en.wikipedia.org/wiki/RLC_circuit

One potential problem is your C value. a 10n in an 0805 package is going to be pretty temperature-sensitive, a few % drift is easy, plus there's the unknown factor of the effect of heating the little blighter to solder it in. And what's the series inductance of the cap? Obviously pretty low, but if it's just 5 nanoHenry, then add that to 97nH, you see how errors of a few % creep in.

To accurately assess your target tester, you need to perform readings on many targets, of differing TC. Then analyse the results,eg by straight-line graph plotting, to show up rogue values, trends, offsets etc.
Maybe if we could come up with a selection of stuff that we both have, that would be a start. There's probably some benefit to mailing stuff across the pond to each other, I don't expect our Royal Mail is going to be cheap .... I'll get some up-to-date price guides.

Thanks for the formula. I looked at wiki/RLC but wasn't sure that was what I needed.

a 10n in an 0805 package is going to be pretty temperature-sensitive, a few % drift is easy, surprised me, I've used cog caps before and they were stable.

Another scope picture of the 40mm loop, Tx on and off. The slope with Tx on is steeper than when Tx off. Calculates L (L=TC*R)closer to L from resonance and Technick. Slope with Tx on is usually a little steeper(usually straight line decay from start) than when Tx is off(after straight line decay linear-log, could be steeper at start of decay). Any idea why that would be? Your TC measurements with a VLF are usually lower than my TC with Tx off so that would make them closer.

Like the idea of having the same targets(verify two targets read the same on my tester before shipping you one)to compare once we decide what they should be.

[Skippy]

If you're positive the caps are COG, then I guess you have to believe the datasheet.
I always felt that COG's main advantage was they were miles better than any other ceramic dielectric, most of which are terrible, especially the C variation with applied voltage. But I've not studied the datasheet of the latest high C value smd's, guess I expected they would be modestly flawed ...

I don't know about the turn on /off difference, that's a PI thing, I know little.

[Polymer]

I have seen "only for reflow soldering" on some datasheets for SMD MLCC capacitors.
Maybe the poor buggers are getting some wonkiness with hand soldering.

[waltr]

I have seen "only for reflow soldering" on some datasheets for SMD MLCC capacitors.
Maybe the poor buggers are getting some wonkiness with hand soldering.

Yes, the problem with hand soldering is the temperature difference across the cap's body and the much quicker temperature rise and fall times.
Both of these can cause major mechanical stress within the cap that can change the value significantly or even cause a total failure of the end metalization.