Primaluna's AC Offset Killer

ltspice tell me the the workings of the circuit.

the diodes act as a threshold for the dc component. without the diodes (rsw=1mohm 1A as the 10H coil is 1ohm).
with the diodes much less not zero , but tolerable.
 

Attachments

  • offsetkiller.jpg
    offsetkiller.jpg
    19.5 KB · Views: 1,652
  • sim.jpg
    sim.jpg
    30.6 KB · Views: 840
Yesterday I built the ver2 and installed it in the power strip for my audio equipment. Playing the same Chesky records tracks, these are after effects so far:
- the slight mechanical buzzing noise coming from my amp is gone. The amp is completely silent.
- for the same volume (28 out of 80), it's louder.
- the bass is stronger, more defined
- the soundstage is "clearer"

The most obvious change is the volume.
 
I've found another DC offset blocker. It looks very similar to Primaluna's isn't it?

DC%20Blocker.GIF

Except in this one they have huge amounts of capacitance. Effective impedance in series is 318mohm at 50Hz. The Primaluna's 100nF has an impedance of 31.8kohms and must be there only to snub the diodes. The diodes do block the DC and do have low resistance, also the 1.4V or so isn't musch compared to the mains. Perhaps that is good enough? Maybe the large capacitance isnt necessary? I must build and try it out myself.
 
I’ve built the primaluna’s ver2 (post 31) based on the basereflex’s schematic (post 21), twice:
- for the hi-fi’s power strip in my living room (int amp, dac, apple tv)
- for the power strip in my bedroom (headphone amp, tube buffer)
Same feedback: it’s louder and the bass is stronger and more defined. I also dare to say the instruments resonance has improved.

It will be interesting to see the measurements before and after.
 
Last edited:
i doubt that that will really block dc on the mains....i really do....:D

i have tried that and without that in circuit, i never noticed any difference, maybe because i was using EI's....

Diodes do block DC. Do a LTspice simulation, look at the primary current (not the voltage). The DC on most mains is less than a few tens of mV, some a few hundred mV. Toroids can hum at very small amounts of DC current, EIs can handle much more.

The hum itself is anoying, but the hum is the sound of vibrations, the vibrations go thru the entire chassis and all components. Some components are microphonic, so vibrations may mess up the sound in some way. Makes perfect sense.
 
Diodes do block DC. Do a LTspice simulation, look at the primary current (not the voltage). The DC on most mains is less than a few tens of mV, some a few hundred mV. Toroids can hum at very small amounts of DC current, EIs can handle much more.

The hum itself is anoying, but the hum is the sound of vibrations, the vibrations go thru the entire chassis and all components. Some components are microphonic, so vibrations may mess up the sound in some way. Makes perfect sense.

I used the exact same type of diode and a slightly different capacitor for $22 per circuit, board cost not included.

I have a couple of questions for you:
- why did Primaluna use this circuit on both Phase and Neutral power inputs?
- which design do you think Primaluna used for their ver2 board?
684619d1527961402-primalunas-ac-offset-killer-offsetkiller-jpg

or
DC%20Blocker.GIF
 
Last edited:
I dont know. I would think DC block on one of the two lines suffices. Perhaps the same reason some use double pole switches or fuses on both lines, gives us a nice gut feeling tho may not be necessary.
I have no idea which ckt Primaluna uses. I'd just use a single block on one phase.
 
I'm still not entirely convinced that the diodes will block DC, although they may attenuate it. You are putting an AC+DC signal through a stage with a central dead zone. The signal coming out will still have some DC. It is vaguely analogous to HF tape bias or digital dither; in either case you can get a small signal past something which might seem to block small signals.
 
  • Like
Reactions: 1 user