I got a prefab (supposedly) TDA7266 board from AE to use as an amplifier for two home cinema satellite speakers on my desk. I put a bluetooth module to play music from my phone at its input and the sound is surprisingly good considering that I got it for something like EUR2.00.
Now the TDA7266 is supposed to have a mute circuit for the speakers at power on but, of course, whoever designed the SMD board was too cheap to put in the two extra resistors and capacitors, so I got this massive “pop” sound on my speakers at power on because of the onboard supply capacitor, and again at power off as the capacitors of the cheap “wall wart” SMPS discharged.
Coincidentally, I happened to salvage a few relays from junk that was donated by a friend, including a very nice subminiature NAIS (Matsushita/Panasonic) DPDT with gold cladded contacts. So I took this as a learning opportunity to design a relay circuit to eliminate the pop sounds. I was planning to use a 555 as a timer for the delay, but as I was browsing online for odd CMOS ICs I came across CD4541 which looks perfect for the job. Just three parts to set the timer and by tying some of the pins accordingly you can get various divisions and/or an inverted output. So there’s no need to use big capacitors for long delays.
As luck would have it, before I got to put more things in the basket and send the order, I noticed a PCB protruding from the recycle bin at work. It was some sort of burned out LED lights timer with, believe it or not, a CD4541 on board. Of course I got the PCB home and, after verifying that the IC wasn’t damaged, I came up with the following circuit which works great to eliminate the nasty pop sounds at both power on and power off.
At power on the CD4541 produces a pulse after about 3 seconds, which is amplified by the transistor which, in turn, energizes the relay coil and completes the circuit of the + side of the speakers. At power off, the DPDT releases the relay by disconnecting the base of the transistor. The Zener helps the relay release faster, before the “wall wart” PSU capacitors are discharged, by redirecting the remaining energy of the coil towards the transistor (see this discussion).
Interestingly, I couldn’t get this to work with a 78L05 regulator to power the relay. It seems that the relay draws more than the nominal 28mA and the regulator could not provide that much current fast enough for the timer to work properly. A simple “ballpark” voltage drop resistor (actually 100+47Ω) did the trick (5.45V at relay pin 1).
EDIT: Corrected time and R6 placement; clarity.