LME49720NA help!!

[klewis]

Originally posted by bassfrontier



What should I do if I really want to keep the circuit DC coupled?

Use a dc servo. I'll post a link to a thread when I find the thread.
ken

[klewis]

DC Servo question...

It's a bit of a read... but the final circuit posted by Tom Gootee is quite good.

Ken

[bassfrontier]

akis,

The buzz gets louder the longer the input wiring gets (i.e. cable running from signal source to the LME49720) and lower when I disconnect the signal input wiring and just leave the 10K to ground (off course I will not be hearing signal since the signal wire would be disconnected.. but I hear minimal buzz when there is no input witing connected- is what I mean)


klewis,

Thanks for the links. I really want to keep it DC coupled. Ill look into DC serco.
I like in Glendale.

[bassfrontier]

This is my current schem.

again..

- The longer the incoming signal input cable, the louder the "buzz" (minimal buzz with no incoming signal cable)

- The "buzz" is there so long as Im not touching a ground on the circuit (like the power supply chassis or the signal ground on the circuit).

- The "buzz" gets louder when I get my hand within 4cm or so of the chip.


Thanks, guys

[akis]

The buzz is generated by your body, and if your wife or son tried they'd get a different buzz. The longer input cable obviously picks up more noise, is it completely unterminated at the other end? With the input of 10K you have selected, I'd go for a 10 times rule, so connect a 1K resistor between the centre and the shield wires at the far end of the cable. Now, if there is any noise to be picked up it would have to be developed across a 1K resistor, so unless you have swallowed some radio-active juice it should stop making noises. Unless the input cable is sh*gged in which case replace it with a better quality one and see how it goes.

[klewis]

I'd remove the 18pf cap between the invert and non-invert pins, move the 10k resistor that is at the non-invert pin to the left side of the 1k resistor.

Hum is generally from bad grounding schemes. There is another thread on grounding, in this forum - have a look. Have you tied power ground to the signal ground? The thread will show you how.
Ken

[AndrewT]

What should I do if I really want to keep the circuit DC coupled?

Originally posted by klewis
Use a dc servo.

and fit a DC detect to the output that triggers a mute on the output whenever the DC servo cannot correct the output offset error.

There is a very comprehensive thread (post22) for DC servos in the Forum.

If you are struggling to cope with wiring up an AC coupled pre-amp then I'd advise that you do not consider adding the complexities required to go to DC coupled until you understand the simple stuff (AC coupled) and then understand the consequences of going to DC coupled.

Yes, remove the 18pF. Some opamps/amplifier become unstable if a capacitor is hung directly on the -IN terminal/trace

0.01uF is 10nF. That combined with {1k0 input resistor + Rs} kills all the treble response and some of the mid range as well.

I'll say it again.

FORGET DC COUPLED until after you understand what you are trying to do.

[klewis]

I concure with Andrew - and suggest that you build the schematic shown on the ESP site, then start tinkering. You need a stable device that you know works. Then start mods on that stable device. We've all made the same mistake of building a complex device that didn't work only to devolve it back to the know device and then build it up one addition at a time.

Ken