Mixer internal phantom power blocking and adding HPF

[bernatvm]

I'm doing some recap and fixing of an AH GS3 mixer. Some of the groups weren't working, some channels doesn't have phantom power. I've also been asked to design some kind of solution for adding HPF to some of the channels.

While studying the schematics I found something that I don't quite understand. In the input of the channel there are some phantom blocking RC filters and the formula Fc=1/2pi*R*C (R3,R4=4k7 and C1,C2=47uF) gives me a Cutoff Frequency of 720Hz. That's not what I'm listening so... Why are my calculations wrong?

For the HPF my idea was to avoid any active design and go for a RC filter (or two in series for 6db/octave slope).

https://www.soundonsound.com/sound-advice/q-can-add-facilities-my-mixing-desk

Would this work if the filter is only to be used in the insert of this mixer? Should I better go with an active Sallen-key with on opamp per channel?

Thanks!

 

[ruffrecords]

Your calc is wrong. On my calculator I get 0.72

Cheers

Ian

 

[abbey road d enfer]

I'm doing some recap and fixing of an AH GS3 mixer. Some of the groups weren't working, some channels doesn't have phantom power. I've also been asked to design some kind of solution for adding HPF to some of the channels.

While studying the schematics I found something that I don't quite understand. In the input of the channel there are some phantom blocking RC filters and the formula Fc=1/2pi*R*C (R3,R4=4k7 and C1,C2=47uF) gives me a Cutoff Frequency of 720Hz. That's not what I'm listening so... Why are my calculations wrong?

Because your powers of 10 are wrong... I find 0.7Hz with your values.  But you forgot to include R7 in your calculation, which result in 2.15Hz.

For the HPF my idea was to avoid any active design and go for a RC filter (or two in series for 6db/octave slope).

Two RC's in series make a 2nd-order filter with 12dB/oct asymptotic slope.

Would this work if the filter is only to be used in the insert of this mixer?

You can expect some unwanted attenuation and poor response of the filter but it may be acceptable after some tweaking.

Should I better go with an active Sallen-key with on opamp per channel?

Better yes.

 

[bernatvm]

Of course something was wrong... ok, thanks! How are you accounting for R7?

 

[abbey road d enfer]

Of course something was wrong... ok, thanks! How are you accounting for R7?

Cut the circuit in half, using an horizontal symmetry line, you end up with 4.7uF loaded by 4.7k and 2.35k in parallels.

 

[bernatvm]

Cut the circuit in half, using an horizontal symmetry line, you end up with 4.7uF loaded by 4.7k and 2.35k in parallels.

I'm sorry but don't quite understand where 2.35k comes from. Shouldn't it be 4.7uF loaded by 4.7k (R3) in parallels with 9,4k (R7 + R4)

 

[abbey road d enfer]

I'm sorry but don't quite understand where 2.35k comes from. Shouldn't it be 4.7uF loaded by 4.7k (R3) in parallels with 9,4k (R7 + R4)

Then you should account for 2 caps in series. 2.35k comes from splitting the circuit in two halves.

 

[JohnRoberts]

Those input phantom blocking capacitors look conventional i.e. well below audio passband.

To add a HPF I would consider a one pole RC... adding a passive RC at the insert point could see some interaction with the Baxandal EQ stage whose input impedance changes with boost/cut settings. You could add a cap in series with R18, but not obvious how to easily make it switchable.

I generally prefer to perform simple HPF at a higher impedance node so you can scale up resistance and use a film capacitor. At R18 (4.7k) would require a physically larger film cap. 

JR

 

[bernatvm]

Then the "best" way to do it would be a sallen key active filter on the insert. Would I need to add an opamp to buffer the signal before or after the filter? Or this should be enough?

 

[JohnRoberts]

Not sure you need a two pole active HPF but indeed you can patch anything into an insert, and experiment.

have fun...

JR

 

[abbey road d enfer]

Then the "best" way to do it would be a sallen key active filter on the insert. Would I need to add an opamp to buffer the signal before or after the filter? Or this should be enough?

This is a LPF. Is it what you want?
You don't need an additional buffer as long as the resistors are about 2+ kilohms.

 

[bernatvm]

No, I want a hpf, just found this for a quick reference and got the wrong one. Why do you say that i need 2k+ resistors?

I imagine it has something to do with the output and input impedance of the insert on the mixer but whats the math? What book should I read to understand it myself?

 

[abbey road d enfer]

I imagine it has something to do with the output and input impedance of the insert on the mixer but whats the math?

It's a ballpark figure. Most opamps are specified for minimum 2k load. The math is more complex.

What book should I read to understand it myself?

There are many. Start with opamp cookbooks, such as this one
https://www.mouser.com/pdfDocs/analogengineer%E2%80%99scircuitcookbookopamps.pdf
and
https://d17ndn7rj73mwl.cloudfront.net/wp-content/uploads/2020/04/ic-op-amp-cookbook.pdf
(page 65)

 

[bernatvm]

Thanks! Will check them for sure!