Retro Amp 50W Single Supply

I have been listening to this all morning. It is really nice sounding. It just amazes me how many different designs can produce great music reproduction. I A/B'd it against a PeeCeeBee and it sounds just as good. Thanks Mile for another great design!
Blessings, Terry
Nice you're done & happy. It is amazing how good 6 transistors & 70 wires can sound, at up to 70W/ch @ 8 ohms (measured, 5 seconds). I have 69 v power supply @ 6.5 amps.
Good point KayPirinha about needing 3 diode stack for non zero OT idle current. I determine that experimentally but nice to know why. I set the 200 ohm pot series 2 diodes for 40 mv across .51 ohms emitter resistor, then clamped pot with another diode.
 
Last edited:
Here are the gerbers for the apex ax-6 single supply amp. yes, i designed it for to-247/264 type output devices. I think i have spare pcbs but no shipping in sight...
the above quoted post gives the schema , pcb, stuffing and bom.
prasi

Dear friend prasi, please, Gerber Viewer say drill size null (there is no hole), can not identify the board outline...
Thanks your precious work!
 
Hello Ades,
there is drill file also (yellow) and outline file (blue) also. Please change your gerber viewer to something like gerberlogix (see screenshot, I use it always and never problem with any board house).

regards
prasi
 

Attachments

  • Image 2.png
    Image 2.png
    84.3 KB · Views: 524
Last edited:
Nice you're done & happy. It is amazing how good 6 transistors & 70 wires can sound, at up to 70W/ch @ 8 ohms (measured, 5 seconds). I have 69 v power supply @ 6.5 amps.
Good point KayPirinha about needing 3 diode stack for non zero OT idle current. I determine that experimentally but nice to know why. I set the 200 ohm pot series 2 diodes for 40 mv across .51 ohms emitter resistor, then clamped pot with another diode.

I have mine set to 7mV across 0R5 resistor. I let it play for about 4 hours and it was at 72mV. That's good enough for me. I didn't try it at lower bias as the heat is not bad where it's at. I want to finish the build off with one of the PSUs that Mile has suggested.
 
I have mine set to 7mV across 0R5 resistor. I let it play for about 4 hours and it was at 72mV. That's good enough for me. I didn't try it at lower bias as the heat is not bad where it's at. I want to finish the build off with one of the PSUs that Mile has suggested.

Sorry, that should have read 70mV across 0R5 resistor.


On reading your 1st posting I was thinking this huge thermal drift is caused by the upper driver that isn't compensated thermally. But a drift from 140 to 144 mA, as per your 2nd contribution, within four hours is more than acceptable indeed!
Best regards!
 
I really don't have any intention to capture this thread. Anyway, if I'm infringing possible forum rules, I'd ask the moderators to concieve just another one.


What I'm going to say is that the genuine Dutch magazine ELEKTOR has published an amplifier in one of it's very first issues that is very similar to these shown here. It's the EQUA. Differences are the true complementary output section, consisting of two Sziklai pairs (no 2EF's), and the current limiter arrangement.

I've built over a dozen of EQUAs in my youth. They sound amazingly, but the design has some drawbacks: There's an immense popping and humming noise on powering on (which surely could be avoided by adding a delay unit) and the current limiters simply don't work. Any short circuit at the output fries the power devices, the drivers and several of those BC107/177 small signal transistors almost immediately.

Replacing the MJ2955/2N3055 devices by the beefier MJ15003/15004 helps somewhat, but still isn't any warranty against failure at all.

In the original article ELEKTOR says that, depending on availability, the BD139/140 driver transistors could be replaced by BD241B/242B's or TIP31/32's if both Baxandall diode arrangements are adapted. I wouldn't recommed that, though, because these are much too slow, though being somewhat beefier than the 139/140's.

For those who might be interested, a schematic is given in this thread of mine: EQUA issues.

Best regards!
 
At least elektor equa has a speaker cap. Worth noting, has it's own thread, please discuss over there. It has 10 transistors and a lot more wires, not worth building point to point when AX6 sounds so good with ~70 wires. My AX6 has a very mild thump at turn on, fades away gently at turn off.
BD139/140 from fairchild (only ones stocked in US) don't have Ft specs like original phillips. MJE15028/29 do, 30 mhz, and sound great as drivers on high frequency sources like top octave solo piano. I can confirm TIP31c/32c are too slow. TIP31c/32c sounded poor in AX6 on those high frequency tracks compared to 1970 RCA 40409/40410 on the other (ST120 PC15) channel. You can't buy 40409/40410 anymore, RCA fab is now a shopping mall or something more useful.
 
Last edited:
Hi Guys,

I etched and built a couple of PSU\10R supplies. First I built the one that Prasi designed on post 671 Retro Amp 50W Single Supply with a slight redraw to accommodate the caps I had on hand and include a bridge rectifier. I tried it out but the transformer I wanted to use is a 30-0-30ACV which rectifies to 83VDC which of course popped one of my filter caps. So I had to use my 21-0-21VAC transformer. With this transformer the PSU10 will only output +55VDC which plays the AX6 fine but I wanted more. ;)

So I redrew the layout to accommodate a 36mm Cap so I could use a 10,000uF / 100V cap I have. This one handles the higher voltage just fine and I am able to reduce the output of the PSU10R down to the +63V I wanted. I used a MJL21195 output transistor for both PSUs.
 

Attachments

  • PSU10R and PSU10R HV.jpg
    PSU10R and PSU10R HV.jpg
    920 KB · Views: 543
  • PSU10R HV with AX6.jpg
    PSU10R HV with AX6.jpg
    844 KB · Views: 529
  • PSU10R 63V.jpg
    PSU10R 63V.jpg
    323 KB · Views: 547
  • PSU10R 100V.jpg
    PSU10R 100V.jpg
    305 KB · Views: 526
  • PSU10R HV (1).jpg
    PSU10R HV (1).jpg
    668.8 KB · Views: 432