Building a symmetrical PSU B1 buffer

Never used one with the DCB1, don't know what it can do for it. If you use one let us know of any possible pros or cons you will find against a straight IEC connector's mains feed please.

Yes, please let us know.
In the Blowtorch thread quite some time ago designer John Curl wrote against such a mains filter advocating a single large 8 uF capacitor across mains as a better solution. (If trying, remember to use a dedicated capacitor for such use.)


/RK
 
line filters

FWIW: Normal line filters contain Ls and Cs that “might” cause additional ringing with the mains transformer though I have not seen any conclusive information on whether that really causes an issue. However, at least for a preamp with it’s low supply current draw, an RC or CRC type filter may actually be a good alternative to a standard line filter. Additional snubbers on secondary (should be used anyway) but also on the primary of the power transformer might also help, actually a CRC filter on the primary would act both as a line filter and snubber circuit. I would put two Rs both in the live and the neutral wire, and calculate it such that you have a reasonably small voltage drop (say 1 – 2 Volts primary).

A frequent claim is that a normal line filtet “kills” the dynamics especially of power amps (not the subject here, I know) but again it is not clear to me whether this is because of a negative impact on the power supply, say by reducing peak current, or whether it is just subjective because some high frequency hash has now been filtered out and the sound is “less dynamic”.

Lastly, Charles Hansen of Ayre states that any component that can be magnetised, e.g. ferrites, are bad and should be avoided even in the power supply. Hence he is using his own type of line filter that he says contains no magnetic materials. My “guess” is that he is using some conductive material and is inducing eddy currents in the material, thus damping high frequencies.
 
Lastly, Charles Hansen of Ayre states that any component that can be magnetised, e.g. ferrites, are bad and should be avoided even in the power supply. Hence he is using his own type of line filter that he says contains no magnetic materials. My “guess” is that he is using some conductive material and is inducing eddy currents in the material, thus damping high frequencies.
could it simply be small air cored inductors in the mains side CLC filters?
 
Hard to say, one would have to talk to the horse's mouth, presuming it wanted to talk in the first place. However, just a few windings would have very low inductance, and I believe I read something about a "patent pending" which would lead me to assume there must be more "in it" ... plus I gleaned some details from a photo showing the entrails of a Ayre USB Dac with an onboard HF filter, again something in the coils ...
 
Last edited:
I had another listen to mine last night. There is a little bit of sibilance in it, though its settled down a fair bit. Sounds very good though apart from that, and definitely better than the regular b1. And thats with the stepped pot in palce which was inferior to the lightspeed in the regular B1. Bass definitely bigger with more dynamics - I wonder is this because the stepped pot is 50k?

So tonight I hope to add in a 560k resistor across the inputs to match teh 50k pot a bit better. That sibilance might just go. Also I know the wiring of inputs needs to be tidied up a bit.

Would anyone have a link fo rhow to simply measure input impedance? I have a signal generator, old scope and a decent DMM that can measure frequency etc.

Fran
 
Characteristically sibilance free is my experience with it in several systems.

1. Change to 560k. 2. What resistors are on your pot? 3. What resistors are your 220R in the audio part? 3. Still you need more secondary voltage.
 
1. will do

2. SMD - its the cheapo stepped pot DACT-like from ebay. Its ok in use, and works fine, better than an alps blue, but not as good as the lightspeed.

3. xicon metal film 1%50ppm from mouser: 273-220-RC (nothing special I know, and deinfitely not kiwame or takman!!)

Voltage, I know, I know!!!


Fran
 
1. will do

2. SMD - its the cheapo stepped pot DACT-like from ebay. Its ok in use, and works fine, better than an alps blue, but not as good as the lightspeed.

3. xicon metal film 1%50ppm from mouser: 273-220-RC (nothing special I know, and deinfitely not kiwame or takman!!)

Voltage, I know, I know!!!


Fran

Hmm...Its a whole system and a family of parts special to your rig as anyone's. You will see what will smooth it out step by step. The base is that you should know there is smoother high range heard with that circuit.
 
Thanks Salas,

I got some 120va 15v transformers on sale at farnell so I bought them up. 14€ each, not bad!

The thing is, I can hear the clean treble extension on this thing. I can hear how quiet it is and how dynamic it is. At the same time, I don't want to overstate the sibilance I'm hearing, but it is there. I also fully realise that there is a bunch of optimisation to be done on my implementation yet - e.g. matching the input resistor to the pot, sorting out internal wiring - in particular the connectors are old, have some corrosion, and the input wiring really needs to be soldered in. The input voltage needs to be upped, and also the current through the shunt too maybe.

So theres a ways to go yet, and from what I hear so far, I think it sounds very very good. It certainly is well out in front of the regular B1. So much so, that if anyone is thinking of building a B1, then I would say skip right to the DCB1.


Fran
 
Things to watch/check out/compare for retaining smooth hf transients:

Dirty contacts.

Non shielded in/out wire, especially if silver plated copper. Compare a plain quality coax substitution effect on one input.

Proximity of transformers.

Artifacts in mains power.

Synergistic in line buffer resistors. I.e. The audio section's 220Rs and somewhat the 220Ks (560Ks if with 50K pot).

Pot quality.

Some dry soldering somewhere.

Traces of oscillation when driving some particular load / interconnects.

Mistracking due to imperfect geometry & rake angle, hard dirt & wear on needle, if the source is a TT.
 
Yes, so when Louis and Ella sing "April in Paris" you get "april in Parisssssssss"


thats an exaggeration, but you know what I mean!

Fran

Those Verve recordings have it a lot. Especially Louis in ''Autumn in New York''. It seems they had to sing songs pleasing their audiences both sides of the pond.😉

The key is to know from some systems how a recording sounds and how much it got. I prefer KNOWN headphones for that, so to avoid irregularities from amps further the chain, tweeters etc.
 
Dirty contacts.

Non shielded in/out wire, especially if silver plated copper. Compare a plain quality coax substitution effect on one input.

thats the most likely problem in my case... will also have a look in my parts stash to see if I have any other 220R resistors. Even a different cheapo type might help if clean contacts and replacing the 220K doesn't work.


I don't expect too much problems with the inrush at 120VA, although I might need to go a little higher on the fuse to avoid nuisance blows.

More later.....


Fran
 
3. xicon metal film 1%50ppm from mouser: 273-220-RC (nothing special I know, and deinfitely not kiwame or takman!!)
Fran
I've ordered my parts from Mouser yesterday and I included some Vishay/Dale RN series that another guy recommended to me, planning to use them in the DCB1 signal path. Do you think it would be worth paying some 10-15 Euro more just to get some Takmans for this role?
 
yep I am using the KOA 1watts, and I think they sweeting up the sound.
Honestly, I am not sure I like that. It's not a huge amount, and it doesn't blur things.

Takeman's sounded a bit zingy in the original B1 circuit. IMO. I would not spend that kind of money on Takemans alone.

I think DaleRN or PRP are about as good as it gets for my taste. Clear, not mellow, not spaztic. you can always change them. The PRP offers a slight improvement in resolution, if one has very resolving equipment.