Useful trick to mute multiple LMxxx6 chipamps

[sidiy]

Hi all,
I did my first Gainclone back in '96-'97... It was a pretty standard implementation (1 chip per channel, dual mono construction) so the AN schematic was enough to make it work without any glitches. It served me well for a couple of years but then I moved to more powerful stuff.
Now I decided to fiddle again with chipamps, this time something more complex, a bridge/parallel configuration for a 'cost sensitive' secondary system. It occurred to me that cross-conduction at power on/off might unnecessarily strain the chips so I adopted the solution shown bellow. I feel more comfortable with a CCS turning the whole gang on and off and with the ability to choose the threshold where muting occurs or how much total current to sink.
It's pretty obvious how it works: Q2 and Q3 implement a wired AND, only when both are 'on' the chips are un-muted. But at power-on Q2 turns on slowly, after about 3sec. while at power-off Q3 turns off as soon as Vcc-Vee<30V (+/-15V) avoiding thumps.
The values shown are for a +/-25V supply and about 1mA sinked from each chip. For any other (reasonable) number of chips R1 must be recalculated (eg. 620ohm for 2 chips,...) D1, D2 are 15V zeners that might help protecting the mosfets.

I hope you'll find it useful (and sorry if something similar was posted before, I'm not watching the Chip Amp section on a regular basis but I saw lately quite a few fellow-members asking for help on this issue)

[tomchr]

Neat! Thanks for sharing.

~Tom

[ratza]

Nice approach. So the LMs should unmute gradualy, isn't it?

[pra3718]

Good, Have you any idea for LM1875 (Chip has no mute pin)

[sidiy]

Originally Posted by ratza

Nice approach. So the LMs should unmute gradualy, isn't it?

Thanks!

Hopefully not sharp, but how slow the un-muting happens depends on how fast the rails are raising ( how much filter capacitance...), divider ratio (R2/R3) and how Rds is varying around the threshold. (2N7002 is a logic level type, so...)

@pra3718: Unfortunately I can only think at the classic solution (relay in series with the output...there must be plenty schematics around)

[thetube0a3]

I'm not sure of why not to, but When I built my BPA200's I tied all 4 mute pins together and used a single very large (800+uf) cap and I forget what resistor value. All 4 chips mute pretty fast but theres a significant delay. The mute circuit is driven off of a 15v regulated rail.

This post leads me to believe there may be a problem tying the 4 pins together? The 2 amplifiers have been in use since mid 2008 with 0 problems so far.