RJM Audio Sapphire Desktop Headphone Amplifier

It wont get any better as the sapphire is not for IEMS. It took me a while to remove the hum from the sensitive 32 ohm headphones but on my Leá IEM the hum is there. To reduce it you will need to use CRC filter infront of the sapphire, shielded cables, zobel on the rca to ground and a simple output filter.
View attachment 1387994
So I fixed the mains hum somehow by just wiring the commons from the bridge rectifiers differently... Also found out I killed one of the common terminals when I soldered it which may have been part of the issue.
Anyways, this amp has an eerily quiet background now that I fixed that - even with IEMs. RJM wasn't exaggerating about it being quiet.
 
You can put an aluminum plate as seen in the photo. It is a simple and very effective method.



20241102_122104.jpg
 
After an unsuccessful attempt to build one, I gave up that pcb-s, no idea what was wrong, my transistors had the same beta, npn and pnp and I was happy abot it, but it was lower than 400....290 exactly. Could this be the reason the amplifier distorted like hell? In fact one could hear a bit of signal but the distorsions were much more higher level...somehow like a torn speaker diafragm...
Now I try to build it again, on fresh PCBs from @Julian RO , but I'm looking for pnp's, I have more than 20 BC337 with beta around 435 but all my 327's are under 300 or above 500...
Like a fate irony....I have a bunch of BC309C, NOS, Tungsram made, and many of them meassure 435...should I give them a try? I know they are not directly complementary transistors....but if the amplification factor is a must, in that case I could complete the two demandings: beta over 400 and equal pairs...
Did anyone try, not especially in this setup, to match different pnp and npn's with good results?
And my last inquiry: is that bad if my BD139-140 are higher beta than 100? Let's say about 180....200? Of course I mean the finals in the signal path, for the regulator ones I have no worry....the more is the best there...
 
Waiting for a more serious answer to post #1896 I started selecting transistors. After another order I got enough transistors to make pairs, in fact I have all BC327/337 at beta between 435....439, all NPNs are 438 and the complementaries not perfectly equal, but I think 435-439 is enough close range. The final transistors are all 4 at hFE 174, some old stock C-DIL brands. Is there e real benefit to choose transistors with higher hFE for the power supply, or it's enough to use also here pairs at hFE 174 like the output final transistors? I know in theory the higher is better, but is it worth here?
I took the liberty to edit/re-draw the schematic, I didn't change anything excepting the layout, I think it's easier to follow the circuit now, especially if someone is short sighted as I am. I hope there's no offense with this redesign...it's just graphical layout stuff.
@rjm : If there are some more changed values or some mistakes in my draw, please let me know and I'll modify it. I hope this is the latest version of this amplifier. I didn't remember the original name of the power supply design, please refresh my memory if there's some specific name to it.
 

Attachments

  • Sapphire 42b.jpg
    Sapphire 42b.jpg
    271.2 KB · Views: 83
250 could be too much, BOM says 100 for BD139-140 hFE.... I was afraid about mine, at 174....these double the rise of amplification factor. Did anyone try such high hFE transistors? What could happen or what behaviour should be expecxted with the rise of hFE, keeping in mind that the complementaries are equal?