Peak indicator led

Hi! This circuit is driven from the output of an op amp, and corresponds to +16 dBu clipping point.
In the original scheme the values of the resistances and capacitors, that I have posted, run whit +15V.
My mic pre run at +18V, and the led lights at more than +16 dBu.
Even if I move the potentiometer R1, things do not change.
What I can do to lower the sensivity of the led?
Really thanks!

Since your schematic doesn't show any junctions for overlapping lines it's hard to figure exactly what you have, but you basic sensitivity, is a base-emitter drop (.5-.6V), times the divider created by the pot, so you won't get much until near 1vp-p. Since peak indicators are generally looking for levels, a few dB below clipping that should work. 

Once you get the threshold right, you may want to add some hold, since peaks can be so brief that you don't see the LED light up.

--------
To get more sensitivity you need to either use a biased differential pair, or you can do a circuit using current mode into the emitter of common base amplifier.

Unless you are really interested in exploring discrete transistor design, you can pretty easily make a simple peak detector with something like LM339 comparators.

There should be lots of peak detector schematic on the WWW, do a search than pick a few you like to explore.  I have done this several different ways myself over the years. Have fun. 

JR

I would try to remove R2 and connect the pot directly to ground.
this way you should be able to lower input signal even more

sandro

There should also be some modest value current limiting resistor in series with the top of the pot so when turned full up, it isn't throwing a diode clamp on whatever circuitry is driving it. Which could cause distortion in the other circuitry and/or damage the transistor base-emitter junction. 

JR

Sorry for the scheme! I added all the connections.
The circuit is preceded by an operational amplifier (NE5534) used as buffer stage.
I could do amplify the op amp stage and than trimming with the pot R1.
What do you think? Could it work?

It's a little unusual but looks like it could work...

the electrolytic cap is reverse biased when the LED is off but only a half volt or so, and any leakage when reversed will be inconsequential.

I am not familiar with the beta of that transistor to know if the 680k has enough base drive to shunt off the LED but it looks like it could.

The 5534 will put out some decent current so add a small R in series with pot.

JR

Referring to post #10, with the exception of the pot and the transistor types, the circuit is exactly the same as used in many Soundcraft channel strips for peak indication. Soundcraft had a 4K7 resistor in place of the pot.

I suspect that your ground net should actually go to -18V if you are wanting to light the LED at +16dBu, which would then faithfully follow the original Soundcraft specification.

IIRC, in Soundcraft mixers the cap is connected between collector of 1st Q and ground, and of larger value (22 or 47uF) ... It may be a new version, which would improve the hold time.
I have used a very similar circuit a long time ago, it was easily usable down to +4dBu.

In a console it makes sense for the whole thing to run off the +/- supplies. Console designers like to keep LED current and other non-linear currents completely out of the local channel ground.

The OP may find out why they did that if he doesn't. But as it is currently drawn he is already potentially connecting a base emitter junction directly across his 5534 output which could lead to some more exciting effects.   

JR

Hello

I am interested in something similar to this only operating at 6.3VDC rails.
would this circuit still work at lower voltages?

Chuck

I've used window comparator circuits to do this, they detect peaks going negative or positive that way

http://www.ermicro.com/blog/wp-content/uploads/2010/03/comp_17.jpg

Make R1 & R2 1k or something, and then make VR 100k, vary this to set the threshold. Put a cap across the ouput, maybe a transistor buffer. Bobs your uncle!

Are you talking about a 6.3V rail, or rails?

The 339 schematic is designed for single supply operation, while it can be adapted to split supply by connecting the ground pins to -V.

JR