And don’t you ever again call my wife’s precious (*) carpet a ‘rug’
Persian carpets are well known for their vibration damping and antitriboelectric properties, the perfect Hi-End cable lifter.
Then the 3 inch spacers are made from a space-age material, direct out from secret labs. I painted it yellow for not to attract attention from the intelligence community.
Hans, the other bundle is 4 zip cords unsplit
George
(*) it smells wool when it comes in contact with the soldering iron
Attachments
I used the LCR meter DER5000. The measurements at 100kHz are the most consistent.
L measured with far end of the cable shorted.
C measured with far end of the cable open
RFZ= SQRT(Ls/Cs)
Single zip: 132 Ohm
4 zips paralleled: 35 Ohm
4 zips split and spaced 3 inches: 447 Ohm
At 100kHz even using any of the Ls, Lp, Cs, Cp readings, ends in an RFZ very close (+/- 5%) to the Ls/Cs value.
RFZ using the TDR method (less time consuming)
Single zip: 155 Ohm
4 zips paralleled: 38 Ohm
4 zips split and spaced 3 inches: 443 Ohm
Tomorrow I’ll review and post the REW impedance sweeps
Hans, how did you produce these Zo traces? They look a bit strange
George
The 100 ohm was with the split wires unconstrained as seen at post 84. Zip cord for speaker test
The impedance of the speaker is at post 82 Zip cord for speaker test
George
The impedance of the speaker is at post 82 Zip cord for speaker test
George
I agree that the traces are looking weird, but that's what my VNA setup returned.
Zipped at 100Khz: 4,3uH+275pF resulting in 31R for 4 zips, close to your result.
Unzipped at 100Kz and spaced far apart: 16uH+34pF resulting in 171R for 4 unzipped.
The 275pF and 34pF being already corrected for the 15pF probe.
But that is just for one single frequency, and taking Zo=sqrt(Zopen*Zshort) gives an exact result because it also takes into account the R and the G of Zo=Sqrt[(R+L)/(G+C)].
Hans
.
Attachments
George,
Looking at the above images:
At 12Khz for the zip version, L=4,3uH and C measures 0pF.
Same for unzip at 63kHz, L=16uH and C measures 0pF.
So taking Zo=sqrt(L/C) leads in both cases to infinity.
Only G is preventing this from happening, but still results in a high impedance, as can be seen here
Zip cord for speaker test
That your unzipped impedance is so much higher, means that you are closer to the point where C measures 0pF.
Hans
Looking at the above images:
At 12Khz for the zip version, L=4,3uH and C measures 0pF.
Same for unzip at 63kHz, L=16uH and C measures 0pF.
So taking Zo=sqrt(L/C) leads in both cases to infinity.
Only G is preventing this from happening, but still results in a high impedance, as can be seen here
Zip cord for speaker test
That your unzipped impedance is so much higher, means that you are closer to the point where C measures 0pF.
Hans
Last edited:
Hi Hans
So there are three up to now who have gone down the rabbit hole (John is out of the counting. He's the one who dug the hole)
Thanks
I thought the Z plots were a result of simulation.
The discontinuities are a result of calculation. I agree that the later formula is the one to use.
R and L from short end measurement
G and C from open end measurement
George
Hi George,
I'm rather surprised by the odd characteristics of this zip cord.
The cable that I showed here with the yellow trace
Zip cord for speaker test
was a 45*0.2mm or 1.5mm^2 power cord, internally twisted, of which I only used 2 of the 3 conductors, see image below.
This cable obviously has a "decent" char. imp. gradient versus frequency without the large peaks of the zip chord.
However, since this cable subjectively produced inferior sound to the MH-770, it would probably be a complete waste of money to repeat John's test with this cable.
Hans
.
I'm rather surprised by the odd characteristics of this zip cord.
The cable that I showed here with the yellow trace
Zip cord for speaker test
was a 45*0.2mm or 1.5mm^2 power cord, internally twisted, of which I only used 2 of the 3 conductors, see image below.
This cable obviously has a "decent" char. imp. gradient versus frequency without the large peaks of the zip chord.
However, since this cable subjectively produced inferior sound to the MH-770, it would probably be a complete waste of money to repeat John's test with this cable.
Hans
.
Attachments
Cat 5 and 6 cables are 100 ohm pairs, and by design they are twisted orthogonally, so that each pair does not magnetically communicate with any other. So paralleling those twisted pairs, means the final impedance will be 100/N, with N being the number of pairs connected.
I don't care for the cat cable overall conductor size however.
Four conductors arranged in quadrature (as in two zips tied together in antiparallel, will have a lower external magnetic field so will have a lower impedance than a single zip. I suspect it will halve the impedance, but there actually may be a bit more reduction as a quadrupole field drops to zero in the exact center. I've not tried to measure it, but it is interesting. If it drops impedance more that 3 times, it might just be a simpler way to tie wrap two zips together antiparallel...
I just designed a cable tray run using a star quad arrangement using 4 535 kcmil wires to keep the external magnetic field down.
Starhex? That's a new one on me..sextupole fields are even less storage than quadrupoles, but it's not a geometically stable configuration unless you put a same diameter core that the 6 go around. If you make the core slightly larger than the outer 6 conductors, then there will be an optimal twist pitch that keeps the cable nice and tight. A good application for PLA filament, it comes in 1.75mm and 3mm. Fishing lines are all over the range.
I use "6 around 1" superconductors on my machines, so have a lot of experience with that mechanical geometry.
jn
Last edited:
