Mic Pre Front-End PCB

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Rochey

Well-known member
White Market Member
Joined
Jul 2, 2004
Messages
2,697
Location
Dallas, TX - Formerly UK
Ladies and Gents,

I'm currently working on a mic pre front end with the aim of offering the PCB's to GroupDiy members.

The function is to actually sit in front of the Mic Preamp itself and offer the following functions:

1 - Full remote control, from either switches or logic (anything can that generate 3V+)
2 - Tried and tested EMI Filter
3 - Pad (-26dB)
4 - Phantom Power (on/off)
5 - Highpass filter (120Hz and above)
6 - Phase.

The Schematic can be found at: http://www.tendolla.com/design/micfrontendpcb/micfrontendschem.png

Feedback please? :)
 
Why did you choose the VR series? I'm just wondering, I usually choose the HE series if I need a polar part.

I would suggest this be spec'd as a Bipolar part.
 
[Why is this in the Black Market ?]

I'd expect those hi-pass capacitors shouldn't be 270pF. Isn't the Zin of a random mic pre a bit too unpredictable to use this kind of HPF ? Don't (some) microphones prefer to see a less frequency-dependent load ?

If you use a DPDT relay for the phantom, you could switch out the two 6k81 resistors when not needed.

Is R3 supposed to be across the mic lines even when the pad is not enabled?

JDB.
 
[quote author="Svart"]I would suggest this be spec'd as a Bipolar part.[/quote]
Any suggestions for 63V bipolars ? Highest I can find in common supply is 50V, which IMO is uncomfortably low for P48.

JDB.
 
[quote author="jdbakker"]
I'd expect those hi-pass capacitors shouldn't be 270pF. Isn't the Zin of a random mic pre a bit too unpredictable to use this kind of HPF ? Don't (some) microphones prefer to see a less frequency-dependent load ?
[/quote]

This front end was initially done by one of our Pro Audio apps engineer as a front end for the PGA2500. I added the relay drivers so that you could drive the 12V (or 24V) relays from a smaller TTL or 5V (or even 3.3V) input.

I assume the values of those "tone" caps would change by the input impedance of the following mic pre? creating a "high-pass" RC filter?

In which case, the values would be different for each mic-pre... ouch.

suggestions?
 
Having 1k series resistors at the input surely privedes an excellent RFI filter but it introduces some very serious noise.

As pointed out by another poster before, the HPF is not a happy solution as shown. It might working with a specific following circuit for low-cost applications but is not usuable for generale purpose pro-audio frontend. If nothing else the noise performance will be upright horrible with standard BJT inputs. There's no solution to the varying -3 dB frequency problem--you have to know the impedances for passive filters.

In addition to this I miss a RC filter for the 48 V (not 24 V, I'd suggest...) supply.

Samuel
 
[quote author="Rochey"][quote author="jdbakker"]
I'd expect those hi-pass capacitors shouldn't be 270pF. Isn't the Zin of a random mic pre a bit too unpredictable to use this kind of HPF ? Don't (some) microphones prefer to see a less frequency-dependent load ?
[/quote]
I assume the values of those "tone" caps would change by the input impedance of the following mic pre? creating a "high-pass" RC filter?[/quote]
Yes. The PGA2500 has a single-ended Zin of 4600 Ohm, so with a 270nF capacitor you'd get the specified rolloff.

[quote author="Rochey"]In which case, the values would be different for each mic-pre... ouch.[/quote]
Yes, and a given mic pre's Zin may even change with transistor hFE. But that's not the worst bit.

[quote author="Rochey"]suggestions?[/quote]
Don't do it this way.

Think about what happens with the mic pre current noise. If your input capacitor is large, the current noise is mostly shorted by the low impedance of the microphone. If you use this HPF trick, you end up with a lot of low-frequency noise in your mic pre output. How much ? Well, I can't give figures for the '2500 as the folks at TI didn't think their customers would be interested in input current noise figures (and I can't be bothered to try to derive it through the difference of the Z=0 and Z=150R noise graphs). For a part like the THAT1512, with a current noise of 2pA/sqrt(Hz) and voltage noise of 1nV/sqrt(Hz), terminated with 10k as per the recommended circuit in the data sheet, the noise floor at 20Hz would be ~26dB higher!

The wiring of the pad relay is still wrong. R3 should be out of circuit when the pad is disengaged.

JDB.
[IMHO, this thread should be in the Drawing Board, with a new one started in the Black Market once boards are really ready]
 
Hia Samuel,

thank you for your feedback. It's always nice to have another set of eyes look over a board :)

It's 24V because there isn't a 48V supply rail library in Eagle... but all my notes suggest it's 48V :)

RC on the 48V supply is doable, but surely that could be done on the power supply itself? would you put an Phantom RC filter on every input?

Finally, regarding the high-pass capacitor values, providing you could get a range of values int he same form factor, I could specify the value for the mic pre. (.e.g for the Green Pre, use a XXuF cap etc). Or, even better maybe provide an online calculator to give the value based on a measured input impedance of your own mic pre (be it a green, an INA163 or a neve!)

how does that sound?
 
RC on the 48V supply is doable, but surely that could be done on the power supply itself? Would you put an Phantom RC filter on every input?
An RC for every channel and none on the PSU is the way to go to reduce crosstalk and noise as much as possible.

Finally, regarding the high-pass capacitor values, providing you could get a range of values in the same form factor, I could specify the value for the mic pre.
Yes but you don't address the serious noise problem that way. I suspect you'll need to skip the HP filter.

Samuel
 
[quote author="jdbakker"]
Don't do it this way.

Think about what happens with the mic pre current noise. If your input capacitor is large, the current noise is mostly shorted by the low impedance of the microphone. If you use this HPF trick, you end up with a lot of low-frequency noise in your mic pre output. How much ? Well, I can't give figures for the '2500 as the folks at TI didn't think their customers would be interested in input current noise figures (and I can't be bothered to try to derive it through the difference of the Z=0 and Z=150R noise graphs). For a part like the THAT1512, with a current noise of 2pA/sqrt(Hz) and voltage noise of 1nV/sqrt(Hz), terminated with 10k as per the recommended circuit in the data sheet, the noise floor at 20Hz would be ~26dB higher![/quote]

I've gone blank for a solution then. The only one I can think of is to put a buffered output on the front-end, with a fixed, known input impedance. but then, I add noise to the front end. It sounds to me like the high pass filter needs to be customised for every mic amp. Again, as I said in the reply to samuel, a generic RC footprint and a calculator for each type of mic pre?

The wiring of the pad relay is still wrong. R3 should be out of circuit when the pad is disengaged.
Thats because I haven't updated the schematic yet :) Removing the ground connection is done easily enough :)


[IMHO, this thread should be in the Drawing Board, with a new one started in the Black Market once boards are really ready]
MOD! Can we move this please? :)
 
Thanks for starting this thread. This pcb would be really useful to me as I'm sure it would be for others, so I hope it continues.

Shouldn't the board have an LED to indicate phantom power as well? and switches to be mounted directly to the board, to simplify building? If so, I suggest someone put together a model layout on frontpanel that can be cut-and-pasted to various projects. I would do it myself, but my layouts are terrible.

One thing to consider adding, also directly to the board, is space for a 7805 or other regulator, since often power supply pcbs you see in this forum don't have a 5vdc output already. Of course, if the supply already has such an output, then the parts need not be installed. And if there is more than one board, the part need only be installed once.

I agree that it should probably be in the Drawing Bd.
 
[quote author="blandman74"]
Shouldn't the board have an LED to indicate phantom power as well? and switches to be mounted directly to the board, to simplify building?[/quote]
I can make that into a separate board very easily.
Switches similar to: http://www.wellbuying.com.tw/pdf/ml.pdf

Also, the board pretty much take any input voltage about around 2V to drive the transistor. (that's what I've SPICE'd so far).
Say you use 12Vdc Relays, you could switch is easily using the same 12V supply to the input.


Cheers

R
 
I really do not like the way pad is configured. Why not use one of the common pads around like any othe mic pre? I like two 820R resistors shunted with a 180R resistor for around -20dB, also making the shunt resistor out of circuit when pad is not engaged...
 
[quote author="rafafredd"]I really do not like the way pad is configured. Why not use one of the common pads around like any othe mic pre? I like two 820R resistors shunted with a 180R resistor for around -20dB, also making the shunt resistor out of circuit when pad is not engaged...[/quote]

Agreed - Next edit will be the resistors themselves.

Next question - should the high-pass filter be stripped from this circuit? It sounds like each high-pass is different based on the Zin (impedance) of the mic preamp that comes after this input card. High pass could even in the recorder itself.

thoughts?
 
Leave it in, because it can be skipped with jumpers by anyone who doesn't want it, and it's easy enough to calculate the appropriate value for the caps. Just leave enough space for beefy ones, and/or multiple through-holes for nonpolarized ones, if necessary.

(Well, now I see you already had that discussion. Carry on .....)
 
there is one feature you might consider, how about a circuit that automatically turns off phantom power when a mic is unplugged? this way you can never plug a mic into a hot input. there are several ways to do it but id use one of those new TI hi-side current sense ICs. (INA-something) Ive never seen it implemented and think it would be a nice touch. perhaps more useful than ramped phantom voltage. just a thought.

mike
 
funny, i'm in the middle of designing a similar pcb for some upcoming jobs.

i personally think it should be simple and minimal, making it as universal as possible. no emi rejection network, no filters just phantom w/indicator + 20db pad + phase (optional, as it is better on the output but some circuits are quasi-balanced so i'm not sure if your standard phase flip works right..) hell, even pcb mount the switches, even easier to install.

-bryan
 
Back
Top