DC coupling capacitor calculation

[luvdunhill]

I am looking to mod a preamp that uses coupling capacitors and wanted to know if there was a way I could measure the offset and compute the lowest value I could get away with, assuming the original manufacturer allowed for headroom in the design, or just threw in whatever was handy.

Is this possible, and how does one go about doing this?

[PigletsDad]

To have a reasonable safety margin, you want the voltage rating on coupling caps to exceed the power supply voltage in the pre-amp, so that the cap won't fail if big transient signal comes along. If the pre-amp has a dual rail supply (positive and negative), you want the difference between rails, as this is the largest voltage likely to come along under normal operation.

In practice, I would usually choose high rated parts, as they cost almost nothing more, and often have better performance. I would also try not use electrolytic caps as signal coupling devices if film types were practical (any value up to say 10uF or so). My experience is that pretty much any film cap is so much better than any form of electrolytic as to be a very cost effective upgrade. They are physically larger, and it can be hard to fit them into some designs.

[AndrewT]

Hi,
the size/value of the signal coupling capacitor does not affect the offset.

The coupling cap forms half of a high pass filter, the input impedance of the next stage forms the other half.

The F-3db of a power amp is usually set in the range 1Hz to 15Hz.
Preamps are usually set lower than this. eg Jensen suggest 220uF for the electrolytic coupling capacitors in each half of a balanced output.

I suggest an RC time constant >=150mS for your preamp.
This can be formed with 1uF+150k or 10uF+15k or 220uF+600r (well almost, what Jensen intended?)

[luvdunhill]

Originally posted by PigletsDad
To have a reasonable safety margin, you want the voltage rating on coupling caps to exceed the power supply voltage in the pre-amp, so that the cap won't fail if big transient signal comes along. If the pre-amp has a dual rail supply (positive and negative), you want the difference between rails, as this is the largest voltage likely to come along under normal operation.

sorry, I probably wasn't clear. I am referring to the value of the capacitor, not the voltage rating.

Thanks though!

[Mitch Alsup]

A long time ago (1973) I modded out a Citation 11 that had a low freq roll off around 4-5 Hz. I founnd the coupling caps on the board, read the numbers from the board parts, and then multiplied that number by 10 and added another cap in parrell. I left the 5-gang tone controls out of the mod. For what it is worth, and being ignorant of high quality caps, I used polyproplene caps, and got good service for many years via this mod.

A switch of the tone controls (with tone controls at 0db) on/off would allow one to hear the difference between a 4-5 Hz roll off and a 0.4-0.5 Hz roll off. It was audible, but not very (Bose 501 driven by Dynaco 150 power amp).

[Bobken]

Hi,

The reason why you could hear this difference is probably due to the fact that with a -3dB roll-off at say 5Hz, the roll-off actually *commences* at a decade higher than that. i.e. at 50Hz, and this is well within the normal hearing threshold, of course, although until at a point somewhere between these 2 frequencies when the reduction reaches maybe 0.5 dB or so, it is unlikely to be readily noticed.

Also, the phase change commences higher still at around 500Hz, which is also likely to be audible in a direct comparison, too, when still well within the normal hearing-range it reaches a significant level.

Regards,