BPA300 mono block finished and measured

What is C2 for? Can I use 9pF instead? What if C2 is not installed?

9pf should be fine. it will start the high frequency rolloff a bit lower but not audibly.
C2 purpose is to aid stability by, among other things, adding phase margin.
it's pretty helpful especially given R3 at 200k.
OTOH i haven't found it necessary in BPA300 or other inverting LM3886 amps if i use a T-network instead of the generic feedback arrangement from the datasheet and application notes. the generic approach forces you into high R values in an inverting topology if you want both reasonable input impedance and reasonable gain. the T-net neatly avoids that.
 
T-network instead of generic feedback? What is EMI? Sorry, but my English isn't good enough...

no worries.
EMI is elecromagnetic interference

T-network looks like this:

!
R
!
!----R----
!
R
!

it replaces the R between pins 3 and 9

the top lead of the top R goes to pin 3
the right lead of the middle R goes to pin 9
the bottom lead of the bottom R goes to Signal ground
tie together the remaining 3 leads (vertical center of the pictorial)
someplace on the forum there is a calculator to help you choose values.
perhaps someone can chime in with a link to it.
 
Last edited:
there really is no"best".
i use +/-25V for a bpa-300 (ie: bridged) into 8 ohm.
each half of the bridge sees 4ohms shared among 3 chips, effectively 12 ohm each.
the datasheet tells you how the chips behave at various voltages.
decide what behavior you want then pick a voltage accordingly.
 
Why are we assuming the model 10 has a "typical configuration"? :)
It's probably safer to assume that the people behind this amp addressed the higher frequency PSRR degradation by using the right poles and getting the greatest synergy possible from the bridge design. From the internal photos I've seen, the power supply looks adequate. The heatsinking may also be contributing to the high volume sound degradation if we accept the guidelines suggested by tomchr for handling the LM3886. By the way, I'm a big fan of your blog and have sent you an email. I look forward to hearing your response.
 
Just from visual inspections - there are indeed ways to get around the PSRR limitation but I didn't notice anything unusual in the components around the LMs when I looked (some time ago at images available via Google). Of course this is speculation on my part because the necessary components could've been on the other side of the PCB, that though seems unlikely.

In general I assume commercial stuff is designed by idiots unless I see evidence of smart-assness, having been an idiot designer myself in a former life :) No email has appeared incidentally, perhaps there's a built-in delay?
 
Yes, thank you Andrew, when I was logged out it was inaccessible, I did finally email through this board. Hope he doesn't get bored reading the same email :)

While I have your attention, I've noted that you've been a member here for a long time. I've never built a class A amp but have become recently interested in doing so. The power supply requirements have always scared me from it but with supercaps and part tolerances getting better and better, I am reconsidering. I posted my thoughts in the Aleph-X Builders thread. I'm assuming you were around to witness its development and can perhaps guide me to the worthiest class A project if that one is not it. No need to reply here since it would be off topic.