+ supply volts on my speakers,help!


2008-02-10 11:40 am
I have just made an amp using LM4766T.Made it exactly to the datasheet except I added 1u caps on inputs.Using a sig gen with only one output(1khz 200mv sine) I fed the signal to each input in turn.What happens is this.
1 Input on R,No output from R speaker,no dc,nothing.
2 The L speaker(with no input signal)goes up to full plus supply voltage with heavy ripple,same heavy ripple on plus supply,negative supply is ok.
3 If input signal is applied to L input,same problem occurs on R output.
4 One test sig applied to both inputs,no dc on speaker outputs and no tone output even when test sig. is increased.

The amp is only on for a couple of seconds at a time while I troubleshoot as I know this will fry speakers.LM4766 stays cool all the while.Supply is +/- 32VDC,6800u smoothing caps and transformer are from the dvd player I got the ic's from.


  • 100_1338.jpg
    85.7 KB · Views: 162
#1 You can't leave the mute pins unconnected.

#2 How did you actually build it? Did you make all connections to both A & B channels per the datasheet, or did you follow your hastily drawn sketch? For example, in your drawing, the only connection to the chip's B channel is the Vcc connection on pin 15. If you only connected Vcc and not Vee for the B channel, that would explain your problem.


2008-02-10 11:40 am
I built both channels but drew only one for clarity.Pin 2 and 15 are + supplys for both channels but this ic only has one negative supply pin(pin 4) for both channels.
I will try the mute pin thing but I am sure when I checked the pcb the ic's came from the mute pins (6,11)were just shorted together.
Thanks for your advice guys.


2008-02-10 11:40 am
I just checked the datasheet again and this time READ the application notes,not just the diagram.
The diagram shows the mute switch as normally open and I assumed this was un-muted.I was wrong as switch is closed for un-muted operation.Silly me.
Does the input to ground resistor go directly from + input pin to ground or from between input cap and input 1k resistor?
Is the absense of this resistor the reason for unused input making the corresponding output "go high"?