Building a symmetrical PSU B1 buffer

diyAudio Chief Moderator
Joined 2002
Paid Member
Nice & detailed account of your thoughts and actions. Keep us posted.

P.S. Shortening the output to ground has a toll on the audio fets, I hope that with the 220R buffer resistors the current will be just OK not to kill any ever. 50K pot for now keeps it back in sonics. A 10k-20k even cheap carbon will do better until you will optimize the LSPD.
 
I don't really see why shortening the output to ground could kill a jfet. I describe what I think happens:
Q1
  • Acts as a current source, so no issue at all, current IDS will remain at IDSS. VDS is well below limits at any time.
Q2:
  • Current through the fet (ID) is limited by the jfet design if I'm not wrong.
  • Voltage over the jfet (VDS): never more then V+, which is +10V.
  • Tension on the gate (VGS): can have full volume from the source, say between + and -2V. At 2V with Rin 220Ohm, gate-current could become 9mA which is below the spec of 10mA gate current. Although I don't see how a gate could draw such a current.
Please let me know where I'm wrong.
I could put an extra resistance to the ground during mute, but it would need to be extremely low value to protect the poweramp from DC. So in practice useless.
I checked the DC offset voltage on startup while it was not shorted to gound, there was no peak showing that both power supply sides went up simultaniously. I'll put a scope to check in detail, I'll do it with a full volume sinewave on the input.
I can as well raise the start-up delay of my poweramp, but I try to have every part self-contained and not depending on another part.
 
diyAudio Chief Moderator
Joined 2002
Paid Member
I don't really see why shortening the output to ground could kill a jfet. I describe what I think happens:
Q1
  • Acts as a current source, so no issue at all, current IDS will remain at IDSS. VDS is well below limits at any time.
Q2:
  • Current through the fet (ID) is limited by the jfet design if I'm not wrong.
  • Voltage over the jfet (VDS): never more then V+, which is +10V.
  • Tension on the gate (VGS): can have full volume from the source, say between + and -2V. At 2V with Rin 220Ohm, gate-current could become 9mA which is below the spec of 10mA gate current. Although I don't see how a gate could draw such a current.
Please let me know where I'm wrong.
I could put an extra resistance to the ground during mute, but it would need to be extremely low value to protect the poweramp from DC. So in practice useless.
I checked the DC offset voltage on startup while it was not shorted to gound, there was no peak showing that both power supply sides went up simultaniously. I'll put a scope to check in detail, I'll do it with a full volume sinewave on the input.
I can as well raise the start-up delay of my poweramp, but I try to have every part self-contained and not depending on another part.

* http://www.diyaudio.com/forums/pass...ymmetrical-psu-b1-buffer-111.html#post2115940

**The DC component rises during shut down.
 
Hello,

I'm in trouble with LED matching...

I must have missed something.:scratch2:

I've bought LEDs in v4.3 BOM (Mouser 593-VAOL-3HCE4), and measured forward voltage of each LED fed by a 9V battery and a 1K2 resistor in series (right, isn't it?).

Measures varies from 1,901V to 1,910V, in spec according to datasheet (mav 2V froward voltage) but far from the typical 1,8V expected value.:eek:

So no way of getting 9V or 5,4V (9,53V and 5,70 are best match)

Help!
 
It will be OK. The Vout values will still track with positive somewhat higher as expected and in sweet spot range. Just find 2 similar strings.

I had the same results when matching my leds (same part number). You will get a few more tens of a volt in the supply rails. Nothing to worry about.
Other people in the thread reported to have the same values with the same part number.

Thanks everyone, so it's OK :cool:

Powered ON soldering iron again...:smash:

I have red LED from different suppliers that vary between 1.55V and 1.95V.

Wow! So I suppose I've been lucky...;)
 
Well, soldering iron now is powered OFF...:D

Haargh, my R-Core transformer is still shipping from France...:headbash:

I've implemented directly last mods, I think...

Vref caps are Wima MKS2XL 2,2uF, the value is ok?

Is 10 Ohm 5W the right value to replace the 68 Ohm couple?

Thanks in advance
 

Attachments

  • intero.jpg
    intero.jpg
    116.8 KB · Views: 406