Ad797 preamp walkthrough

[iampoor1]

Recently Ive been reading alot of schematics, books, and articles to fill up all the missing gaps in my knowledge. Of course, its still like the grand canyon. 

I recently came across an AD797 preamp project from http://diy-tubes.com/diy-kits/

http://diy-tubes.com/image/data/manuals/ad797-v1.0-en.pdf

I have a few questions and I would like to make sure that Im right for what (little) I already know!

R1/R2 are the phantom feed resistors. 6.8k very typical

C1/C2 are the DC/Phantom blocking capacitors. How do you calculate the value of these? I have seen 1uf, 2.2uf (like this design, 10uf, 100uf etc. I am presuming that a lower value will present less phase shift, and a smaller time constant for charging..? (Not sure if that would improve transient response? Or maybe I am totally wrong here in the first place)!

C3/C4 Provide a low impedance path to ground for high frequencies. They also create a low pass filter in conjunction with the source impedance.

R3/R4 set the input impedance and limit the open loop gain so that the preamp doesnt need a microphone attached to not oscillate?

C9-C16 are the bypass capacitors per the AD797 Datasheet

R5/R6 provide an output load in addition to what the preamp is driving. They are low to reduce noise and to allow plenty of current drive.

Now, the feedback network I dont understand. The 5k Pot R24 and R9/R7 create the feedback loop with a voltage gain of around 50 or 34db right? C5/C6 in conjunction with the feedback network set the frequency for this gain? Im not sure what C8/R10 and R8/c7 do, and why they are connected after the output resistors.

Please help me fill in the blanks. Im here to learn!!

PS: per previous recommendations...Douglas Self's small signal audio design is an amazing book!

 

[PRR]

C1 C2 need to be small compared to 2K to pass your lowest bass note.

AND they need to be small compared to the mike (200r) IF the amplifier has input current (BJT input), or you get a rise in random rumble. Altho in most studios the blower rumble masks quite bad amp noise rumble.

R5 R6 isolate the delicate amp from whatever hunk of crap you attach to the output. In particular, the capacitance of long lines. If the NFB amp feeds that directly, the open-loop gain drops excessively at high freqs, the NFB loop goes unstable. The isolator should be a significant fraction of the amp's output impedance. Which is not "zero" at high frequencies.

In a bit of extra engineering, R8 R10 force the at-jack output impedance to zero despite the 33r stuck in. C6 C5 take hi-freq NFB from before the 33r. Myself, I'd let the output be 66r. I like a dumb no-tricks output, and see little value in a "zero ohm" output.

 

[ricardo]

C1 C2 need to be small compared to 2K to pass your lowest bass note.

AND they need to be small compared to the mike (200r) IF the amplifier has input current (BJT input), or you get a rise in random rumble. Altho in most studios the blower rumble masks quite bad amp noise rumble.

I think you meant to say the Reactance of C1 C2 needs to be small ..

R5 R6 isolate the delecate amp from whatever hunk of crap you attach to the output. In particular, the capacitance of long lines. If the NFB amp feeds that directly, the open-loop gain drops excessively at high freqs, the NFB loop goes unstable. The isolator should be a significant fraction of the amp's output impedance. Which is not "zero" at high frequencies.

In a bit of extra engineering, R8 R10 force the at-jack output impedance to zero despite the 33r stuck in. C6 C5 take hi-freq NFB from before the 33r. Myself, I'd let the output be 66r. I like a dumb no-tricks output, and see little value in a "zero ohm" output.

This stuff is copied straight from AD797 datasheet.

I have some 'real life' evidence that "zero ohm" output Z gives better RFI immunity over long balanced lines.  I used this on a large balanced line distribution system to active speakers for theatre sound.  Not with AD797.

I'd love to check RFI immunity with 100m Belden driven by their 45R nominal Z and 90R but can't test RFI to that sort of level this Millenium 

See http://www.proaudiodesignforum.com/forum/php/viewtopic.php?f=6&t=464 for loadsa measurements but alas, none for
'real life' RFI.

 

[iampoor1]

C1 C2 need to be small compared to 2K to pass your lowest bass note.

AND they need to be small compared to the mike (200r) IF the amplifier has input current (BJT input), or you get a rise in random rumble. Altho in most studios the blower rumble masks quite bad amp noise rumble.

R5 R6 isolate the delicate amp from whatever hunk of crap you attach to the output. In particular, the capacitance of long lines. If the NFB amp feeds that directly, the open-loop gain drops excessively at high freqs, the NFB loop goes unstable. The isolator should be a significant fraction of the amp's output impedance. Which is not "zero" at high frequencies.

In a bit of extra engineering, R8 R10 force the at-jack output impedance to zero despite the 33r stuck in. C6 C5 take hi-freq NFB from before the 33r. Myself, I'd let the output be 66r. I like a dumb no-tricks output, and see little value in a "zero ohm" output.

Thank you!

Whats the relationship between Input capacitance and amp rumble? I think I have heard amp rumble beore, it was also like a sub bass generator. I had to filter out around 15-20hz IIRC. Is the rumble basically a low frequency oscillation??

Okay. So the values arnt particularly important, as low as they are low but not zero right? Ive never had to drive long cable lines, and Ive seen everything from 10r to 220r used as output resistors. Ill need to do a bench test sometime...

I just reread the datasheet and found that section. Whats the logic with R8 being 3k? Is there a particular frequency response that you want to achieve a 0 ohmn output at? (If there is a purpose, which seems negligable for a studio setting)