biasing electrolytics in speaker crossovers?

I know there has been plenty of discussion about biasing electrolytics and film in DC blocking apps.

However, a search for crossover AND electrolytic AND bias drew a blank.

Wouldn't it make sense to replace the large bipolar electrolytics in speaker crossovers with two polar ones of twice the capacity and bias them, e.g. with a 1 k resitors to the positive or negative rail of the power amp?
 
biasing electrolytics etc

Hi capslock,

You're right, electrolytics behaver differently depending on the signal polarity. The applications of 'lytics in xovers I have seen/used always had either bi-polar 'lytics or back-to-back "normal" 'lytics. Any particular reason for you to use a polarised set-up (or just wondering)?

Cheers, Jan Didden
 
Jan,

go to the "electrolytic decoupling ..." thread in this forum and look at the Jung/Marsh article which is the first or second link provided by H.H.

Fig. 3 gives two 6u8 caps back to back with 680 R load in a DC blocking (high pass) configuration with no bias (C), and 5, 10, 15 V bias (D, E, F).

Occured to me that that might be a neat way to fix the large crossover capacitors that aren't readily replaced by film devices.

In my speakers, I have already replaced all the series capacitors, but the shunts in the woofer branch were simply too big. Conventional wisdom has it that it does not really matter what happens in the shunted branch (see e.g. Harry's shunted volume control). I am not really convinced because the shunt impedance is after all part of a voltage divider and will have some influence on what passes through...

Eric
 
capslock said:
Jan,

[snip]
Conventional wisdom has it that it does not really matter what happens in the shunted branch (see e.g. Harry's shunted volume control). I am not really convinced because the shunt impedance is after all part of a voltage divider and will have some influence on what passes through...

Eric

Eric, OK I'll check the Jung/Marsch article. Should have it in my lab library, but its from waaay back when.

On the above quote, I am strongly convinced that whatever you do in (a) shunt DOES influence the resuls! As you say, the shunt is as much part of the circuit as the series arm. Consider a node that has three leads connected: the input from a series circuit (let's call it the source), a shunt to ground and a wire to the input of an amplifier. Thevenin has it that the sum of currents into a node is zero, so whatever goes to the input of the amp is the sum of the current from the source + the shunt. QED.

Cheers, Jan Didden