Cable measurements: Brushing up my knowledge

Hari Iyer

Well-Known Member
Joined
Feb 8, 2010
Messages
3,822
Points
113
Location
Mumbai
I have been using my telephone cable interconnect for now 1+ month and was quite contented with them. This week i hooked up my stock Marantz CD player IC swhich was supplied and noticed that this cable had more high-end than my telephone cable IC. i swapped multiple time to check if my findings were true and to my sesurprise it was. So i headed for my LCR meter yesterday evening to measure primarily the cable capacitance and this is what i got,

Marantz CD player stock IC - cable capacitance was - 144pF which is around 1.5 meter long cable.
Telephone cabe ( 4 core + 4 core) - cable capacitance was - 310pF which is around 2 meter long cable.

Needless to say, i had to remove the extra 3 + 3 cores to reduce the cable capacitance and i did so. Also reduced the cable length to 1.5 meter to keep consistency.
I measured again and this time for the Telephone cable i got a capacitance of around - 74pF.

Now this set me to do some math as below,
Input impedance of my tube amplifier - 100Kohms
cable capacitance - 74pF

If i calculated the -3dB roll-off point from the forumula F=1/(2*pi*z*c) the roll-of frequency is 21507Hz.

Now if the cable capacitance increases to 150pF, for the same input impedance the roll-off changes to 10610Hz and for a cable capacitance of 300pF the roll-off is a scary 5305Hz.

When i replaced the Marantz stock IC with the new modified telephone cable IC, i got crisp and detailed highs, very good midrange and astonishing bass. I did not know the cable capacitance also has +ve influence on the low end too. Also i noticed that the Marantz stock interconnects were a bit distorted and vocals were not very smooth. I attribute this to the PVC dielectric which messup the sound.

Hope my assertions are right and if not will like to hear from FMs.
 
Hari sir, I have a spare Blue Jeans LC1 cable which claims a capacitance of 12 pf per feet.
I can send it to you for running a comparison.

cheers
kannan
 
In your F calculation, you have considered C of the cable under test yet considered the Z of the destination amp (input Z of 100K). Why not Z of the cable under test? Replace the 100K figure with a more realistic figure like the characteristic impedance of the cable, say, 75R and recalculate it.

Also, I'm not sure if the formula for impedance that you used calculates the -3 dB roll-off point. IIRC, it simply calculates the impedance of a component at F frequency.
 
With ic, capacitance has to be as low as possible. With speaker cable, inductance and resistance has to be as low as possible

That’s the general thumb rule
 
In your F calculation, you have considered C of the cable under test yet considered the Z of the destination amp (input Z of 100K). Why not Z of the cable under test? Replace the 100K figure with a more realistic figure like the characteristic impedance of the cable, say, 75R and recalculate it.

The input impedance of the amplifier will be the load to the cable and hence that was considered. Its similar to how/why me match the input impedance of a phono preamp using different values of R & C across the phono cartridge terminal. Moreover the capacitive reactance and inductive reactance of the cable is a function of Frequency.

Xc = 1(2*pi*F*C) and XL = 2*pi*F*L

so Z of the cable = R + 1/Xc + XL

Hari sir, I have a spare Blue Jeans LC1 cable which claims a capacitance of 12 pf per feet.
I can send it to you for running a comparison.

cheers
kannan
Kannan, thanks for the offer. By any means can you measure them if it agree with the specs?
 
The input impedance of the amplifier will be the load to the cable and hence that was considered. Its similar to how/why me match the input impedance of a phono preamp using different values of R & C across the phono cartridge terminal. Moreover the capacitive reactance and inductive reactance of the cable is a function of Frequency.

Xc = 1(2*pi*F*C) and XL = 2*pi*F*L

so Z of the cable = R + 1/Xc + XL


There's a fundamental flaw in there somewhere.

The impedance of a cable of 2m or less can't be as high as 100K. Physically impossible. At the most its characteristic impedance will be sub-100R or 600R for balanced cabling. The load is kept x10 or more of the output impedance of the source. Hence the typical 47K input impedance, and sometimes 100K. The input impedance of the destination device cannot be assigned to the cable which is a mere conduit from source to destination.
 
There's a fundamental flaw in there somewhere.

The impedance of a cable of 2m or less can't be as high as 100K. Physically impossible. At the most its characteristic impedance will be sub-100R or 600R for balanced cabling. The load is kept x10 or more of the output impedance of the source. Hence the typical 47K input impedance, and sometimes 100K. The input impedance of the destination device cannot be assigned to the cable which is a mere conduit from source to destination.
The impedance of the cable will be less than 1 ohm if you apply the above forumula as R is as low as 0.2 ohms for a 1.5 meter cable. So if you measure the capacitive reactance say at 20KHz for the cable which has 74pF capacitance and around 5uH as inductance (as measured by me) then you will get an cable impedance of 0.66 ohms.
Z = sqrt(R^2+X^2)
 

I don't know if I am getting it wrong. As per the online formula in the page, capacitance seems to have no impact on audible frequencies for lengths under 1000 meters if the capacitance is under 100 per meter.

Values I took:
Output impedance: 100 ohms
capacitance per meter: 100
Cable length for roll off to reach 15000hz: 1000 meters.

Output impedance: 100 ohms
capacitance per meter: 300
Cable length for roll off to reach 15000hz: 350 meters.
 
Hari, quick question: how tightly twisted are your telephone cable ICs? Higher twist rate equals higher capacitance. If you cut open balanced cables the twist rate is rather low to keep capacitance in check.
 
Hari, quick question: how tightly twisted are your telephone cable ICs? Higher twist rate equals higher capacitance. If you cut open balanced cables the twist rate is rather low to keep capacitance in check.

Arn't balanced cables dual core coax?
 
No.

They're just two wires twisted together and surrounded by one or two shielding layers.

I was under impression that balanced cables are 2 straight cores shielded by braided layers. Didn't know that they are twisted inside.
 
Hari, quick question: how tightly twisted are your telephone cable ICs? Higher twist rate equals higher capacitance. If you cut open balanced cables the twist rate is rather low to keep capacitance in check.
I have not done additional twisting from my end. It's a loose twist that comes standard with purchase.
Yesterday I replaced my tape deck IC too. 1 meter measured 45pF which is quite cool.
 
Guys, today I borrowed his DIY interconnect cables mentioned above from Hari and connected in my system. Let me also tell you I replaced Van den Hul and Nordost Blue Heaven with these ones.

And I was pleasantly surprised by the improvement in detail, especially the upper harmonics. I don’t understand his calculations, but I surely can vouch for the sound enhancement. Even my wife arrested same.

Hari, thanks for opening my eyes - costlier brands don’t necessarily work better.
 
Check out our special offers on Stereo Package & Bundles for all budget types.
Back
Top