DC feedback at pin 7 of the LM 3886?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Use pin 7 as DC feedback by means of a servo unit

Hello,

Maybe my question is not clear. Normally pin 7 of a LM3886 is connected to ground. If you see the equivalent schematic, pin 7 is the pin which begins inside with a resistor af 10k.

In the case of a single supply voltage circuit, pin 7 is used to bring the output voltage on half the supply voltage (see application note).

So my idea is to use pin 7 for adjusting the output offset to zero. This can be done by means of a servo unit.

I hope somebody will reflect on this idea.

Marc Vi.
 
As you see in LM4780 / 3886 / 3875 / 1875 Comparison (LM3886 Annotated Equivalent Schematic) pin 7 sets the ground voltage. If you use an input capacitor (which I always advise) to block an input DC voltage but not a feedback capacitor, because that will be an electrolytic one, then a output offset voltage will occurre. By putting a low voltage with a low impedance on pin 7, you can cancell out the output offset voltage. So this pin sets the output on the midpoint of the rails. Pin 7 is apart from the signal ground, so to my opinion it doesn't affect the quality, on the contrary.
 
Is pin 7 part of the protection systems?
Does pin 7 allow turn off in event of fault/abuse detection?
Does pin 7 allow delay if the supply rails are not up to minimum specified operating voltage?

Can the simplified schematic be relied upon to use features that are not in the datasheet?

There's only one way to find out!
 
You are right if you use an capacitor in the feedback way. This capacitor always has a value of 10 microFarad or more, so it will be an electrolytic one and that is not so good for the sound quality.

If you don't use a feedback capacitor then the output offset voltage will be multiplied by the amplification factor, mostly 10 or more. The output offset voltage will then be 120 mV or more and that is too high.

Marc.
 
I think it is a good way to correct the offsetvoltage. Than you do'nt have to use capacitators in the signal-inputs other than an inputcapacitator to block the dc input and to get a roll-of at the lowest frequency. The inputimpedance is much higher than the impedance in the feedbackpath. Therefore the inputcap is rather small.
 
No, this will NOT work. The ground is only "active" as an "input" while the mute is on. Otherwise the output is determined solely by the + and - inputs.

I recall Nelson Pass posting here (in one of the many chipamp threads) suggesting that distortion could perhaps be reduced by increasing the current into the mute pin past the 0.5mA suggested in the data sheet. This prompted me to trace out the circuit (or maybe just follow along with his post).

The LM3886 equivalent schematic (at the earlier-posted link and on page 6 of the LM3886 data sheet) shows the input and mute circuitry. The mute input goes to the emitter of an NPN whose base goes to the ground pin and the collector to the base of a longtail pair. With no current into the mute pin, the right-hand transistor conducts, switching in the two transistors with 4k7 resistors. One's bass input connects to ground through the 10k resistor. The other connects to the output through a 10k resistor. The collectors drive the transistors along the bottom that drive the VAS and output.

With sufficient current through the mute input, the left-hand transistor turns on and the right-hand one turns off. This powers the two transostors with 1k1 resistors in series with their emitters. The bases connect to the emitters of the input buffer transistors, and these then pass the + and - input signals to the output.

It appears you CAN control the output offset with the ground pin while in mute mode, and it appears to be a gain-of-one amplifier with the ground pin as the input. It might be desirable to make the DC output in mute mode the exact same as the DC output in normal mode, but it would be most desirable to make them both as nearly zero as is practical.

Does the output have significant offset in mute mode? I'd guess it's pretty close to zero.
 
okay Benb, I've read your explanation and I agree. There are two differential pairs, one is used to handle the inputsignals and the other is to handle the mute-status of the chip.
When mute is on then the output-offset wil be the offset from the internal diff-amp, only some mV's.
When mute is off the other diff-amp is functionning. Than you can use a servo-amp on the -input to reduce output-offset.
Jac
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.