RCA interconnect cable capacitance, ok with 800 pF ?

[cbueche]

Hello everyone, I hope this is the right forum for this.

I have as source a CD player, phono preamp and Raspberry Pi. They are connected to a Teac UD-503 DAC with their respective connections (coax, RCA in, USB). The DAC serves as pre-amp/source selection. It drives the power amplifier, a Millett Engineer Amp, using unbalanced RCA interconnects.
Due to my limited space and speaker placement, I had to use long RCA interconnect cables between DAC and amplifier. I made my own interconnects using Sommer Cable club MK22, cable length about 4m (say 14 ft). I measured the interconnect capacitance at about 750-800 pF.
The DAC output impedance (from doc) is 150 ohms, the power amplifier has a Broskie "Front panel" signal selector and attenuator. From doc : "The FrontPanel stereo attenuator presents a nominal load resistance of 28k to the signal source".

The whole setup works well and has very low hum. Can someone here tell me if my setup makes sense, or if my cabling kills the high range of the spectrum too much ? Any formula to apply to calculate the atenuation considering values above ?
Or should I better move the amp, use shorter interconnects and instead long HP cables ? They would be 4 and 7 meters.
thx!
Charles

 

[Baldin]

http://sim.okawa-denshi.jp/en/CRtool.php

Low pass filter 150 ohm / 800pF => Fc = 1.3 MHz ..... Think you should be ok 😉

Long interconnects is a problem if you get hum, otherwise not. Lowest signal voltage should be kept shortest in general ...

 

[merlin el mago]

@Baldin.
Wrong calculated: It's a high pass filter not a low pass filter, R is 28K Broskie & C is 800pF cable capacitance, the cutt-off frequency is 7.105Hz

@Charles if you have to choose between short or long cables always better short the RCA & long the loudspeakers cables. If you measured well you are listen a tweeter....or better said a super-tweeter.

 

[MarcelvdG]

@merlin el mago Unless you connect the cable in a very unorthodox manner, the cable capacitance is in parallel with the DAC output. Baldin's calculation is spot on. One could nitpick and state that the calculation should really be done with the parallel value of 150 ohm and 28 kohm, but the difference will be negligible anyway.

The only thing that's missing are possible non-linear effects if the DAC output should run out of current, but those shouldn't be an issue either.

 

[Markw4]

Perhaps it should also be noted that output load capacitance to some extent becomes part of an amplifier feedback loop transfer function. The particular extent depends on various factors such as, for example, build-out resistors, output inductors, zobel network, etc. In some cases SQ can be audibly affected even if instability does not occur. Easy to check by listening though, comparing short cables with long ones.

In addition if one believes in cables having 'a sound,' other than simply due to capacitance and or Z0, then length of cables may also be involved in perceived difference effects for less well understood reasons.

 

[Mark Tillotson]

For low impedance audio signals cable capacitance isn't usually noticable. Its high impedance sources like guitars and phono cartridges where this is a concern.

 

[cbueche]

I should have drafted the circuit as I understand it. Here it is.

So yes it is a low-pass filter and the cut-off frequency is 1.3 MHz. Meaning I have no issue with my bandwith and the only potential issue is hum, which is ok. Thanks all for the help !

 

[merlin el mago]

@Baldin.
Wrong calculated: It's a high pass filter not a low pass filter, R is 28K Broskie & C is 800pF cable capacitance, the cutt-off frequency is 7.105Hz

@Charles if you have to choose between short or long cables always better short the RCA & long the loudspeakers cables. If you measured well you are listen a tweeter....or better said a super-tweeter.

I'm sorry guys my bad. The capacitance it's in parallel to the load not in series.

 

[Hierfi]

Perhaps it should also be noted that output load capacitance to some extent becomes part of an amplifier feedback loop transfer function. The particular extent depends on various factors such as, for example, build-out resistors, output inductors, zobel network, etc. In some cases SQ can be audibly affected even if instability does not occur. Easy to check by listening though, comparing short cables with long ones.

In addition if one believes in cables having 'a sound,' other than simply due to capacitance and or Z0, then length of cables may also be involved in perceived difference effects for less well understood reasons.

For low impedance audio signals cable capacitance isn't usually noticable. Its high impedance sources like guitars and phono cartridges where this is a concern.

Agreed, though I wouldn't confess to cables having a sound unless willing to accept the wrath of the gods or double blind testers demanding whatever truth be proven before the faintest utterance of an opinion. Okay, I do (I couldn't make the font any smaller).

It is commonly observed that build-out resistors have values as high as 100 to 200 Ohm's, values thought much too high for minimizing dielectric influences in subsequent cabling if such exists. Some high-end gear (read pricier) have output resistances in the order 5 to 10 Ohms. It can be imagined that if the output impedance is zero Ohms all electrons going into and coming out of the dielectrics in a cable are either replaced or removed as not to affect the voltage output of the amplifying device delivered signal to subsequent devices. Notwithstanding that cables have a "sound" it seems good practice to maintain low values of any build-out resistors used.

It was often found that networks operating close to instability often sounded best. To maintain stability can require an included output network to support a low impedance drive and isolation from unpredictable loading.

 

[MarcelvdG]

You could use some LRC network instead of a resistor, as is usually done in main amplifiers.

 

[Hierfi]

True. I have often used a coil across a build-out resistor.