SPICE VS real life..

See my posting earlier with lots of measurements.

• Since it allows EVERYTHING except what i eXplicitly mark, localnet.com does not know beans from duck eggs.

• Got it all.

Comment relative to the degaussing coil..full AC voltage is never applied,, and what is applied decreases rapidly. If i remember correctly, a positive TC power thermistor is in series with the coil; starts at low value when cold, and increases with temperature. Power to degaussing coil is only for a few seconds at most.

• Circuit? Where?

Value 8.000 printed on xtal case (no units); most xtals show more digits..

They use a different PIC. I don't know if that makes any difference, but the real value of the site is they published the hex code for the PIC. I'll see if I can find a PIC disassembler and find out how the code works. Maybe it will work on any PIC.

A shorted turn would mean the inductance is less than the calculated value. We do not have a calculated value for the coil.

However, the problem is the discrepancy between the oscillation frequencies for the Hartley and Colpitts, and the inductance value indicated by the AADE.

The AADE claims the coil is 21.29mH. The Hartley and Colpitts give the following inductance values:

Colpitts: 16.28 mH Hartley : 16.53 mH

All three readings are on the same inductor.

If there is a shorted turn, it should affect all three readings. Presumably the Q would drop, but this should show up in the LTspice results and require a lower emitter resistance to reach the same p-p voltage across the tank.

capacitive

Many professsional instruments ground the case. Measuring the frequency is always problematic due to scope loading. But we don't need accuracy to

if

Are they all the same type? Maybe they are ferrite core and that is offsetting the readings. The ferrite has higher permeability which means fewer turns so the distributed capacitance woud be lower.

I'm wondering if distributed capacitance is throwing the AADE measurement off.

We really need the AADE oscillation frequency when connected to the degauss coil. The calculation will show if the frequency matches the value expected with 680pf or 1680pf in parallel.

those

That's a lot of work, but I don't know what the goal was. I'll wait for your analysis.

Most recent measurement: L1 degaussing coil, unknown turns R = 17.83

Actually Robert Baer posted diameter and number of turns and Awg

174T of 24Awg wire yielded 16 ohms and 25.6mH, fairly close on the resistance, but way off for the inductance.

working backwards from diameter yielded 158 turns to get the 'measured' inductance, which meant the possibility that 16 turns were shorted out there. [perhaps, high impedance short?] However, putting that primitive model into LTspice NEVER got the circuit to oscillate, so don't know the shorted turns effect.

Using the very odd value of 25Awg, not in library, use 0.45481mm as strand diameter also perhaps the actual conductor is NOT pure copper with a conductivity of 58MS/m, but more like an 'alloy' that has 48 to 52MS/m ?? for now assume 58.

Interestingly, shorted turns makes little difference near DC.

The Q is at least 80 - 100 from 'measurements'.

Interestingly, 118T yields a coil that made the LTspice osc near 22kHz, but the Rdc was waaaay off. have to be as small as 26Awg wire, perhaps the triple dip coating used on AC mains coils just made the wire's OD 'appear' to be larger. Don't know.

I'll be looking at RB's latest set of tables to wee if I spot anything.

don't know if this will post all, but at a first pass: =-=-= NO CX: [approx] calculate 17.1mH at 10kHz, Rdrive=10k calculate 179uH at 10kHz, Rdrive=1k calculate 199uH at 10kHz, Rdrive=100 calculate 2.79mH at 1kHz, Rdrive=100 that table is really wonky!

trust this a bit more... Cx [just first two entries] calculate L=10.3mH; calculate L=10.1mH; self resonance implies 1nF ??!!.

When I put 11mH into the LTspice circuit I get 22kHz oscillation which is also around 118T.

Something is STILL wonky here!.

PS: When I use a function generator as a source, it has a little spikey at turn around point. So at parallel resonance [which is normally very difficult to find] I simply make certain that spikey appears on top the peak signal. By lining up the phase is a bit more accurate, then read out on frequency counter. f is easily 0.01% accurate that way.

interestingly, that second value is 17.9mH *IF* Rdrive=100k, not the indicated 1k

• Degaussing coils do have a LOT of tape. Say 200T for grins; that translates to 5pF per turn which is not unreasonable. 150T gives 6.7pF/T which is high but not unreasonable.

Correction/clarification: Equipment used:Protek 6500, AADE L/C Meter IIB L1 L2 o--+--uuuuuu-+-uu--+---o--vvvv--o TOP GND | | Rdrive +----||---------+ Cx

L1 degaussing coil, unknown turns R = 17.83 L2 added secondary, 11 turns R = 0.70 Resistance reported is measured value minus 0.02 ohms, the meter lead resistance. Est error 0.01 ohms all values above.

no Cx; Rdrive=100 1%| L2+L1 Frequency | Vtop | Vtap ----------+-------+-------- 10KC |5.878V |5.782V ----------+-------+-------- 1KC |4.476V |2.219V ----------+-------+-------- 100~ |4.223V |0.701V ----------+-------+-------- These numbers should pin down the inductance of the coils.

• BTW, i cannot understand what FEMM shows for the provided files, and see no obvious(?) way to even lay down a one-turn loop much less a wire.

Well, the pk-pk amplitude (18V) is limited by clipping to the voltage source (9V).

• I have NEVER seen any hokey connection to the case on the outside of a crystal, and never any indication by any manufacturer of an internal added wire to the case. If there was one, then there would be a polarity indication of some sort. Foil caps do have such an indication for outer foil, which could be used as a shield.

• Does it matter? 2 were same type and maker; one prolly had a ferrite core as it was a lot smaller (dia, length) than the other two. They clearly measured well within their 5% spec.

• If the AADE uses a different cap for different inductance ranges,then the cap used would be suspect.

• was interested in the actual coil inductance, went to SPICE and got a burp.

• Wire size is a crude guess.

• So,except for inductance, something is right...

• #26 wire is possible, triple dip not reasonable due to cost. But construction with some turns,tape wrap, more turns, wrap, etc is not unreasonable.

I dont know what you mean by hokey, but crystal case connection is common enough

NT

They help but surprisingly not a lot, because 100 ohms is too small to read the impedance at 10kHz, but try it anyway.

I get, as a simple indutor model near DC 10.8mH, 18.53 ohms, and 1nF [from your resonance circuit] That value of inductor causes your circuit to osc near 22kHz. See below for simple Test Circuit.asc Problem with this coil is that you need to add a Laplace equation to represent the skin effect taking you up to near 30 ohms at 10kHz. but you get the drift of how to use LTspice to 'verify' measurements over frequency. Didn't I first estimate around 11mH?

you have to RUN the provided files, then VIEW the .ans file that gets created. Some step by step instructions are in the $Notes.txt I sent [also I don't quite understand what you're asking here?

There are two ways to represent wires in femm. Easiest is to simply define a large enough cross section block to contain a number of turns of ?? Awg wire. The other way is to actually draw in EACH conductor wire, which is much more accurate, but VERY tedious and slow on the analysis.

Version 4 SHEET 1 880 680 WIRE -384 32 -480 32 WIRE -240 32 -304 32 WIRE -208 32 -240 32 WIRE 0 32 -128 32 WIRE 32 32 0 32 WIRE 32 64 32 32 WIRE 192 64 32 64 WIRE -480 80 -480 32 WIRE 32 112 32 64 WIRE 192 160 192 64 WIRE 32 224 32 192 WIRE 32 336 32 304 WIRE 192 336 192 224 WIRE 192 336 32 336 WIRE -480 384 -480 160 WIRE 32 384 32 336 FLAG 32 384 0 FLAG -480 384 0 FLAG -240 32 top FLAG 0 32 tap SYMBOL res -400 48 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 0 56 VBottom 2 SYMATTR InstName Rs SYMATTR Value 50 SYMBOL voltage -480 64 R0 WINDOW 123 24 124 Left 2 WINDOW 39 0 0 Left 2 SYMATTR InstName Vs SYMATTR Value "" SYMATTR Value2 AC 6 SYMBOL res -224 48 R270 WINDOW 0 32 56 VTop 2 WINDOW 3 0 56 VBottom 2 SYMATTR InstName R2 SYMATTR Value 100 SYMBOL res 16 208 R0 SYMATTR InstName RL SYMATTR Value 18.53 SYMBOL ind 16 96 R0 SYMATTR InstName L SYMATTR Value 10.81mH SYMBOL cap 176 160 R0 SYMATTR InstName CL SYMATTR Value 1nF TEXT -320 160 Left 2 !.ac dec 200 100 10k TEXT -256 240 Left 2 ;Rdc = 18.53 ohms\nRac1kHz = 22 ohms*\nRac10kHz = 32 ohms*\n* estimated

From what i have seen over a 40 year period of time, a crystal case connection is so un-common as to be called "never". Can you provide any real examples (even if i have to rip them apart)?

Back in the day, HC-18 crystals were often mounted on their sides because they were so tall. To prevent them from falling off due to vibrationally-induced fatigue, they often had a short lead soldered to the top (i.e. the end opposite the leads). I've seen that often in radios.

Cheers

Phil Hobbs

I've seen 32.768 KHz crystals soldered to the board, at the base to stabilize them. I have some unused crystals with a lead welded to the case, to hold them in place.

I have seen several amateur radio articles, in which the crystal casing has been soldered all over to the ground plain, in which a single point connection to ground would have been enough.