LM3915 LED meter circuit, do i need a buffer?

[mrclunk]

We're going to start recording to tape again but our desk lacks metering on all the channel outs so i'm looking into adding a basic LED output meter for each channel.
We'll need to know  levels going to tape but the machines VU will be out of sight.

The only place to fit them is next to the channel faders so i can either tap off directly post fader inside that module or tap off after the (post fade) direct out level control.
(But that would mean routing the signal back down the desk to the fader module.)


The LM3915 looks to be a simple solution but i'm wondering if i need an additional buffer where i'm tapping off post fader.
Its data sheet shows a buffer on the input of the chip, is this sufficient or do i need extra buffering to leave the audio untouched?

I found this api circuit on the forum which looks very close like what I'm after.
5 LEDs is perfect really.
http://www.classicapi.com/catalog/images/gallery/3124/1987-era-3124-schemo.pdf

Possibly this will be ok as is?

Attached is the circuit after the fader.

 

[mrclunk]

Here's the LM3915 diagram.
http://www.ti.com/lit/ds/symlink/lm3916.pdf

 

[SSLtech]

Here's the LM3915 diagram.
http://www.ti.com/lit/ds/symlink/lm3916.pdf

You posted a link to a pdf with MANY diagrams in it.

Without knowing which one you're referring to, it'll be hard to comment.

The 3915/3916 measures and displays DC by the way, so you'll need supporting circuitry either way... and means of calibrating it...

Have you searched for the already-implemented 3915-type iterations like the Pico-meter? I thought we've already been down this road some time ago.

 

[boji]

Since I was going to post a VU question, may I just add it to this thread? I hope that's ok.

I wanted to know if branching out an unbalanced signal to a vu meter buffer (my buffer asks for unbal in) could create noise problems in the existing audio path.

The answer is probably yes, and I'd like to know if instead choosing to tap the differential part of the signal path and going with a differential buffer input circuit (like the one mentioned on this thread, or perhaps the JLM board) would be a better idea...

I have probably answered my own question, but I'd like some secondary opinions / confirmations.

Thanks for your reply!

 

[abbey road d enfer]

The LM3915 looks to be a simple solution but i'm wondering if i need an additional buffer where i'm tapping off post fader.

Although some get away with feeding the audio directly to the LM's input, you need a rectifier to capture the envelope of the signal. Since the rectifier puts a highly variable load to the signal, you need a buffer. The API schematic does away without a buffer, taking advantage of the extra output capacity of the 2520.
Not my idea of excellence in design.
In your case, since you want to graft the meters on existing outputs, you need buffers.

 

[abbey road d enfer]

Since I was going to post a VU question, may I just add it to this thread? I hope that's ok.

I wanted to know if branching out an unbalanced signal to a vu meter buffer (my buffer asks for unbal in) could create noise problems in the existing audio path.

The answer is probably yes, and I'd like to know if instead choosing to tap the differential part of the signal path and going with a differential buffer input circuit (like the one mentioned on this thread, or perhaps the JLM board) would be a better idea...

Only a differential buffer can guarantee interference-free eavesdropping on an audio line. The cost difference between a balanced and an unbalanced buffer is about 4c.