Mic Fet booster with stepped HPF and line drive

[AnalogJoe]

In through-hole:

I clipped this circuit some time ago thinking it might be useful. It is a +12 to -48V DC/DC converter for telecom applications that could be easily modified to be a +48V supply for phantom power use.


-48V Power Supply, "DC/DC Converter for Telecom Applications," Charles Hill, Darin Kincaid, ST Microelectronics Application Note.

Full text of -48V Power Supply Using SG3524

There may be more elegant and simple solutions using modern switchers, but the SG3524 is widely available from many, many sources and the inductor requirements allow off-the-shelf components to be used.

To provide +48V for phantom, D22 and C42 could be reversed along with pins 1 and 2 which are the inverting and non-inverting error amplifier connections.

Q3, shown as a 2N4033 or MPSA56, is not drawn particularly well, having no emitter arrow. The emitter goes to +12V.

The switching frequency is a little low, 44 kHz, but it could be raised or sync'd to Fsample.

I think they mean that U19 pins 13 and 14 should float, not U19.

SG3524 Block Diagram


SG3524 Block Diagram

In TH there's also the MC34063.

I have a lot of phantom supply circuits using SMPS ICs that are SMT.

Not sure how the opto-isolator benefits unless one needs to galvanically isolate the low voltage input from the +48V output.
I see optos and TL431 combinations in a lot of line-powered switchers using transformers for isolation.

ST SG3524 Datasheet

Very nice. I love positive to negative upward converters. They always involve pesky coils, but they are great.

[Tubetec]

There was a huge storm passed over the Irish sea a few days ago , wreaked havoc at the Port in south Wales where the post is handled ,resulting in nearly 350,000 packages from the UK to Ireland being delayed ,
Chances are my unit was caught up in it ,
The UK postal service hires in many casual workers over the season , a lot of stuff goes missing ,
So I'll just have to wait and see .

Switching power often spoils the noise figures if you measure beyond the usual 20khz limits ,
Using 9V alkaline cells would cost around a Euro an hour , so thats not an option either .
I might try a bank of sealed lead acid batteries to get a range of voltages upto 48V ,
I often see SMPS batteries in good working condition at the recycle , one fails and they replace the lot ,
I carry a cheap multimeter and a car headlight bulb as a dummy load ,

Its really the audio part of the circuit Im interested in ,
I have the idea for the buffer circuit in my head but I didnt commit it to pen and paper yet ,

[Tubetec]

The Canford unit finally arrived .
Two new alkaline cells put out 9.73v each ,
12v and 48v lines measure correctly off load ,
On load both the MKH416P48 and the 415T work as they should ,

Its all through hole components , with mostly tantalum caps in the signal path ,
The passive HPF element works very effectively , with the 150 or 400 hz settings generally being preferable on close miked voice ,
Works equally well on dynamic mics , without any batteries ,

Id been looking around for a good overview of the Neve Ba283 line driver stage ,
sure enough this popped up ,
https://groupdiy.com/threads/questions- ... a283.6425/
Unfortunately the link to PRR's schematic is missing from post 13 , so context is lost .
If someone has the document stored locally maybe it can be put back in its rightful place over on GDIY.
No sign on way back machine either ,
https://web.archive.org/web/20240000000 ... 74f09b.gif

[Tubetec]

Ive been looking for ages for a filter and attenuator calculator for balanced signals ,
https://ethanwiner.com/gadgets.html
above has a link to a useful network calculator for balanced audio lines ,
One minor improvement would be to show diagramatically each type of filter along with the list of components .
its really fast and easy to use , thanks Mr Winer .

[mediatechnology]

Thanks for posting "Mike Pads and Other Small Gadgets."

[Tubetec]

Some other handy calculators and audio related stuff here ,
https://sengpielaudio.com/Calculations03.htm

https://sengpielaudio.com/calculator-db-volt.htm

[emrr]

Thanks for posting "Mike Pads and Other Small Gadgets."

Yes, hadn’t seen that.

[Tubetec]

I ran the soundcard cal test in REW a few times with the Canford unit in the loopback path from the headphone out(configured differentially) to mic in .
The top four traces are the HPF curves ,
the lower 3 are flat with attenuation settings of 10,20,30db ,
The -10db setting knocks off nearly 14 db ,while the -20 and -30 are more or less right ,

I tried it out with a few different mics , both dynamic and phantom powered condenser ,
The first three HPF postions are very usable , the fourth is probably a step too far ,

The unlabeled resistor in the HPF network is 1200 ohms ,

[Tubetec]

C5,C8 0.82uf ,1.1uf
C6,C9 4.5uf . 4.6uf
C7,C10 12uf ,12uf

I added a 1uf in parrallel with c5 AND C6 ,as the bass cut was to extreme , by selection I was able to reduce the difference in value to .1uf .
A 1.2k resistor cuts in when the HPF is engaged , it tends to load things up a bit , shifts the action a bit too much to the right and bunches the curves up in the low end .

The orange trace is the original value C5,C6 , the lower green trace is with a value of 2uf ,

I tried 3.3k in place of the 1.2k HPF load resistor ,it helped a bit , in the end i just disconnected it ,so instead just relying on the 1.2k ohms on the mic input of the interface .
It shifts the filter into a more usable range to combat proximity effect , with less effect on the mids ,

[mediatechnology]

Not sure what they were thinking based on the original curves.
Your mod looks to be a lot more useable.