Cable capacitance

I always been lazy with math i am a screwdriver person

So before i start digging in my rusty math i seek quick advise from Solid state people

In a class AB amplifier which has most of the times feedback lets suppose that one user uses cables that for some reason have some or excesive capacitance

--is the capacitance itself that introduces a filter in the circuit probably effecting high frequency response ? lets suppose that the cable has 10-20-50-or 100pf capacitance does this form a filter that is in the audible frecumncy and that will effect the total result ?

--is the relation of feedback involved and this will after all effect the total result ?

kind regards
Sakis
 

academia50

Member
2013-01-16 2:52 pm
In my experience, in a solid state amplifier, the cables to pay attention to are the low level signal input. They must be shielded cables, of course, and of good quality (Teflon insulation was my favorite) The magnetic phone input is the most delicate, short paths must be taken. For years, they were guilty of many headaches (Hum) Mechanical keys on the front of the board and inputs on the back panel of the amplifier made a very long round trip, increasing the probability of picking up buzzing coming from the magnetic fields of the PSU. Modern amplifiers use DC-powered relays, and are directly on the rear panel by passing the signal to the circuit boards (PCBs) in minimum paths and also deactivating all other inactive inputs (grounded)
So I would worry more than the capacitance of the cables that carry signal (if they are of good quality you can not have problems) by the distribution of feedback loops that are formed by a poor distribution of components.
I textually transcribe the book "Sound systems" by Phillips:

"In the power supply and the output stage, currents of various amperes circulate. The conductors that carry these currents should not be part of the input circuit, otherwise humming or instability may occur due to low but nevertheless significant resistance thereof "
 
I always been lazy with math i am a screwdriver person

So before i start digging in my rusty math i seek quick advise from Solid state people

In a class AB amplifier which has most of the times feedback lets suppose that one user uses cables that for some reason have some or excesive capacitance

--is the capacitance itself that introduces a filter in the circuit probably effecting high frequency response ? lets suppose that the cable has 10-20-50-or 100pf capacitance does this form a filter that is in the audible frecumncy and that will effect the total result ?

--is the relation of feedback involved and this will after all effect the total result ?

kind regards
Sakis

No sensible answer is possible unless you tell us where the cable is. Look at it as a capacitor. You can't say what a capacitor does unless you know where it is on or in the circuit.

Jan
 
Capacitance in the order of a mere 100 pF is not going to affect a power amp output at all. Try something like a few nF. Even then, small-signal frequency response in the audible range is entirely unaffected. You can thank Messrs. Thiele and Zobel for keeping stability in check though.
 

Attachments

  • amplifier-Cload-1-nozobel.png
    amplifier-Cload-1-nozobel.png
    56 KB · Views: 357
  • amplifier-Cload-1.asc
    1.4 KB · Views: 3
  • amplifier-Cload-1-zobel.png
    amplifier-Cload-1-zobel.png
    58.3 KB · Views: 360
With respect to the preamp/mixer cables to power amp.
I had 12' of RCA jack coax cables connecting my RA-88a disco mixer and my power amp.I realize 6' is traditional, but my room layout doesn't allow that. My Parts-express capacitance meter said 12' RCA coax cable is ~820 pf of capacitance. 1/2*pi*f*C says this would be 1/.094 or 10.2 ohms impedance. A 33078 op amp in that RA-88a mixer can drive 600 ohms to typical 10.7 v swing. Typical short circuit current is 30 ma. Fortunately I can't hear 20 khz, only 14 khz.
To try to lose some hum at maybe 55 db down, I changed to a 25' twisted pair 1/4 phone plug cable which the meter says is about 1050 pf. 1/2*pi*f*C is 1/.126 or 7.94 ohms. However the PV8 mixer with 1/4 phone sockets driving the TP cable has 4565 op amps which have a maximum voltage swing of +-10v @ 25 ma current. 4565 have short circuit current of 130 ma.
I'm hearing a little less hum on the TP. I'm hearing less AM radio pickup when the mixer is turned off. I'm not so sure of high freq response. The amp with RCA jack input and the amp with 1/4 phone jack input are different. The RCA jack amp (ST120) had 150 pf across the inputs. The 1/4 phone jack input CS800s doesn't have any input capacitor, just a 4560 + & -.
 
Last edited:
To try to lose some hum at maybe 55 db down,
Sounds like maybe a ground loop problem.

Indianajo, try making some Rane Note 110 #17 or #13 cables. (Measure where ground potential on the mixer is with respect to protective earth with a multimeter in AC mode. If ~identical, use as-is. Otherwise connect cable shield to output ground.)

I am not sure why you are even having these issues still - the Peavey's input as described as "quasi-balanced" with a "relatively low impedance" ground loop eliminator circuit (maybe 100 ohms || 100 nF to chassis/PE or so?), and the Stereo 120 appears to be floating altogether, so I don't see any major ground loop currents flowing there.

For testing purposes, could you short the cable at the mixer end and connect just the ground? If so, hum or no hum?
I would also suggest investigating the issue with a minimal system if at all possible - one source, one mixer, one power amp.
 
820pF at 20kHz has an impedance of magnitude about 10k
1 / (2π * 2e4 * 8.2e-10) = 1 / 1.03e-4 = 9.7e3 ~= 10k
Whoops, 820 pf is 820 e-12, not e-9. Great, no problem. 33078 op amp will drive 10k impedance.
sgskloss, both sets cables TP & RCA coax dead silent shorted input on 101 db 1w1m speakers with gain all the way up on CS800s. CS800s gain specified @ 29 db stereo 4 ohms. My speakers are 8 ohms. My hearing is normal to 14 khz a 2008 test @ work reported. RA-88a disco mixer is wall transformer DC power, case ground connected to turntable shell green wire. Turntable is 2 pin power plug, floating. PV8 mixer is floating case wall transformer power. Cs800s amp has earth pin case not tied to either input pin.
Quietest hum setup is RA-88a ( MM LP, CD, mono FM radio) to PV8 (RA-88a+stereo FM radio) to CS800s via 25' TP cable. Still has a little very faint hum. Stereo FM radio and mono FM radio are battery powered. Mono FM radio is disconnected, down for new e-caps since December.
Don't have a 12' TP phone plug cable. Have 100' of 16 ga TP bulk with braid shield. That picked up AM radio 30 db down as 30' speaker cable off Allen S100 amp. Shield was unused. Allen S100 lacks output coil or RC network on output. Conclusion, ungrounded shields are bad news, make good AM antennas.
ST120 under workbench for hum analysis; a black wire popped loose on AX6 board and when I put it "back" started humming that channel maybe 40 db down. Must be in the wrong place now, hum was about -55 db for 3 years.
On 2 vac scale of tenma 72-7725 DVM, RA-88a mixer ground terminal for turntable to a IEC power cord receptacle 3rd pin 6 mv. PV8 ground ring of in input, 11 mv.
 
Last edited:
In a class AB amplifier which has most of the times feedback lets suppose that one user uses cables that for some reason have some or excesive capacitance

--is the capacitance itself that introduces a filter in the circuit probably effecting high frequency response ? lets suppose that the cable has 10-20-50-or 100pf capacitance does this form a filter that is in the audible frecumncy and that will effect the total result ?


Filtering is purely due to the RC circuit formed - the output impedance of the source driving the cable forms a divider with the capacitance of the cable.


One important case is for an MM phono cartridge, which are usually carefully designed to require something like 300pF of load capacitance. The intention is that most of this is due to the cable, and the rest is provided as a selectable capacitance in the phono preamp input circuit. This is a more complex setup, the cartridge has resistance and inductance, and the load resistance and capacitance are specified to give the best, flattest, frequency response across the audio range (ignoring the RIAA correction which is then applied). So the cable and the input circuit of the preamp form a filter designed to match the requirements of the cartridge.


If phono cartridges were made with a flat response that assumed no capacitance, the phono preamp would have to be built into the tonearm.


Signal interconnects between preamps and power amps typically have a low driving impedance (a few hundred ohms), and its takes quite a lot of capacitance to be an issue, several nF at least.


For a power amp to 8 ohm speakers even 1uF would have minimal effect as the amps output impedance is < 1 ohm. However too much capacitive load on an amp can make it unstable, which is another matter.


No realistic cable has 1uF of capacitance!
 

Elvee

Member
2006-09-08 2:04 pm
Capacitance just by itself is normally not going to affect audio frequency parameters.
A long enough figure-of-8 cable will attenuate high frequencies, but that's because of its inductance, not capacitance.

A cable however is not just the combination of a lumped capacitance and a lumped inductance: both are distributed, and this does not matter at all for audio frequencies, but it does matter for audio amplifiers: a mismatched cable can display any impedance depending on the frequency, and this may stress the compensation, typically in the MHz range.

A well damped Zobel should take care of that though
 
Capacitance just by itself is normally not going to affect audio frequency parameters.

Many years ago we built two identical amplifiers with a friend.
One of them suffered from very poor performance at high frequencies.
The capacitor of the Baxandhall network in the treble section in one of them was 0.01 mf. , and in the other of 0.10 mf.

Guess what the problem was?
 

Elvee

Member
2006-09-08 2:04 pm
Many years ago we built two identical amplifiers with a friend.
One of them suffered from very poor performance at high frequencies.
The capacitor of the Baxandhall network in the treble section in one of them was 0.01 mf. , and in the other of 0.10 mf.

Guess what the problem was?
The discussion is about speaker cables, not capacitors in general.
0.2pF misplaced in a UHF tuner can make a huge difference
 
The discussion is about speaker cables, not capacitors in general.
0.2pF misplaced in a UHF tuner can make a huge difference




The discussion is about the importance of capacitance in the cables in the amplifiers (specifically SS, as I mentioned when the OP opened the wire), the UHF is what has nothing to do with it here.
The answer is that a small difference in capacity can attenuate frequencies in an audio amplifier, because it acts as a filter in a tone control.
And if you follow the reasoning, a difference in the capacitance of the cables can also influence, and much if it is in the wires of a magnetic phono input, either those of the TT or those of the input circuit of, RIAA curve, gain , etc, etc, etc.

Regards, friend ;)
 
The Baxandall tone control networks are high Q, adjustable active filters. You could well expect them to be very sensitive to their C, R values. Audio cable capacitances are generally small and low Q, meaning they should be scarcely any different when used with their intended or typical source and load impedances.

yes, I agree with that, it is correct, in my example, when we located the wrong value, the treble returned to the defective amplifier in its FR
Regards friend
 
The answer is that a small difference in capacity can attenuate frequencies in an audio amplifier, because it acts as a filter in a tone control.
And if you follow the reasoning, a difference in the capacitance of the cables can also influence, and much if it is in the wires of a magnetic phono input, either those of the TT or those of the input circuit of, RIAA curve, gain , etc, etc, etc.

Regards, friend ;)


We've covered this already really, most cables have ~100pF per metre or so and it takes >100nF to make a non-trivial difference to a speaker given the source impedances involved, ie cable capacitance has no effect on a speaker with any conceivable cable, other than to affect the stability of the amplifier driving it, (which is why we have output inductors).


For small signal cables the impedances are higher so capacitance can be much more important - source impedances of 100 ohms or more are common, so several nF of capacitance (10's of m of cable) will start to be an issue sometimes.


For MM cartridges capacitance of the cable is all important, the source impedance is of the order of 1kohms + 0.5H, forming an LCR circuit with the cable and preamp with a pole near 15--20kHz to provide modest peaking at the top end.
 
Subjective differences between speaker cables are huge! Generally, best sounding speaker cables are very low inductance, pleated or flat foil. Unfortunately, such cables are nightmare for amplifiers that does not use output coil. That's because of high capacitance of such cables. You would have to add coil to your favorite ESP P3A amplifier if you want to try some of these cables. The result is very much dependent on loudspeakers you use. If you use mini monitors (like Proac Tablette) you already have all detail and resolution you need so no low inductance speaker cable is necessary. If you have big 15" Tannoy dual concentric in a small room, you will try to retrieve all the fine details from your system because bottom end sound signature of your speakers will dominate. In that case it's worth trying low inductance loudspeaker cables. Your taste also matters. If you like darker, laid back sound even miniature loudspeakers like Harbeth P3ESR, can leave the impression of big loudspeaker. You must experiment to get the sound you like. Sometimes the only way to get satisfying reproduction is to use preamp with tone controls.

Some people say that for audio you don't need pleated speaker cable. That kind of cable is for RF antenna amplifiers. But in practice you will notice huge difference with cable like Kimber 4TC . When we speak about resolution and fine detail it's simply much better than ordinary figure of 8 cable.

To completely cancel the effect of loudspeaker cables you need to build amplifier with remote sense feedback using two pairs of loudspeaker cables, one to deliver signal to loudspeaker and the other to take feedback from loudspeaker terminals. With such topology you can use the cheapest loudspeaker cables on earth and still get great sound.
 
Last edited: