Salas DCG3 preamp (line & headphone)

[donhughes111]

bambadoo,
Did you use the maldano build guide for your muses? I am about ready to start. Any suggestions not in the thread or in the build guide.
And, its the hookup to the I-select straight forward.
Thanks,
Don

 

[donhughes111]

reconfigure DCG3 Pre

Salas,

I will be posting pics of my build by this weekend. Went pretty well. Absolutely dead quiet even at full volume. Sounds pretty darn good for first effort. Listening test on weekend. 13 months of build. Not for the faint of heart to go that long and not hear anything. Started from ground zero. Or, more like Zero degrees Kelvin. Not sure what I was going to get when I turned everything on. Overall topology is Salas FSP, DCG3 and M2x with an SP10, EPA 100 arm and ATG33-2.
Speakers are currently Tannoy revolution XT 8F.

Reason for writing is the DCG3 pre with I-select really kicked me on the build. All on me - novice. I wanted to re do and go more like the phono pre and separate the Trafos. Got 2 pair of Flexy boards from T. I think I have the flexy figured out. Use a 6800uF cap that matches the UBiB. Run a shielded DC power cable to the pre and ground the shield to the pre box like on the phono. My real question is best way to hook up the DC power to the UBiB. I will be taking the Quasimodo stuff and the diodes off the UBiB as it will reside in other box. Do I keep the little 0R33 resistor in front of the Caps and how best to wire it. This is set up as dual mono. There are 2 +/- UBiB's

Thanks for all you do .
Don

 

[bambadoo]

The Meldano muses is pretty much straightforward.
You will figure it out 🙂

 

[pfarrell]

Wow—our systems are really similar irrespective of rooms and speakers (and tonearms). You will have to get the Hana ML unfortunately ;-) Haha. But really, you should. You've completed a full refresh of one of the decks? I can't imagine the need for a separate PSU for DCG3—quiet is quiet in my book—but IDK. Ultra FSP? Uh oh. 😀 An incredible amount of work that will surely be worth the effort, leaving you with a solid feeling of gratitude daily and with every use. Congrats!

 

[Salas]

My real question is best way to hook up the DC power to the UBiB. I will be taking the Quasimodo stuff and the diodes off the UBiB as it will reside in other box. Do I keep the little 0R33 resistor in front of the Caps and how best to wire it. This is set up as dual mono. There are 2 +/- UBiB's

Thanks for all you do .
Don

Hi

External DC power cabling to UBiB needs no special measures since there is no AC field emanating from the DC connecting wires. You may follow the phono's umbilical practice as well. Skip the bridge diodes and the Quasimodo components of course. Keep the 0.33R (RF) in series to C1 if the DC wires are thick and relatively short. But if they are long enough and kind of thinner flexible, they do offer little resistance of their own to naturally substitute RF's role.

 

[Salas]

Neat.
Finally finished the muses 72320 volume control also. Remote control works fine and also the Adafruit display.

Need a new faceplate.

Small progress... im waiting a few parts RCA -Pot Tocos 20K- Emi filter-knobs-UP68 Im playing with chinise 2sk170 - 12ma very pleased...

Nice progress in both projects.

 

[Nikos]

@nokosokey - Looking good. What did parts/values did you use for C6 and C7? These look larger than then GB kit parts.

Jaytor thanks!
Yes indeed...
I put what I had in my drawer ....they have value 220uf/100V.
https://gr.mouser.com/ProductDetail/647-UFG2A221MHM1TN/

BR Nikos

 

[bambadoo]

Quick question. The dcg3 plays very well. However a slight "hum" can be heard from the speakers when ear is close to elements.

It seems as I got some 50hz(mains), 150hz, 250hz, 350hz, 450hz etc. fundamentals on the outputs. They are pretty low in level but they are there.

Any suggestions what can cause this?

Will post pictures later.

 

[hpasternack]

Ground loops, it's always ground loops. It's fascinating that the most important signal in any audio circuit is the one with no signal (theoretically).

 

[bambadoo]

Here is a screenshot.
50hz (of course) and peaks every 100hz and smaller at 100, 200, 300 etc

Someone pointed me in the direction of rectification noise? I am using hfa08tb60 diodes on the dcstb PSU board.
https://www.vishay.com/docs/96190/vs-hfa08tb60-m3.pdf

I even disconnected the muses volume control (the psu for it) and the result remain.

Its somewhat clean build and thinking of desoldering all the inputs and just measure PSU and dcg3 out. Just to rule out the inputs and i-select board.

 

[hpasternack]

Once you've covered the basics of keeping sensitive wires away from hum and so on, you need to look at your grounding scheme. There are whole textbooks on the subject, though. I have to make money for my employer right now, so can't really go on in more detail.

 

[Salas]

Here is a screenshot.
50hz (of course) and peaks every 100hz and smaller at 100, 200, 300 etc

Loop area field sensitivity or ground loop between gear. See what happens if you lift the IEC ground from the chassis for a test. To see if there's a route through there.

 

[Salas]

I even disconnected the muses volume control (the psu for it) and the result remain.

Because the Muses has an independent PSU there could be loop area interference between its PSU zero and the general signal ground that refers to the DCSTB PSU's zero.

 

[bambadoo]

Lifting the IEC inlet gnd actually made the 50 and 100hz a bit worse. 150,250,350 etc. were unaffected of this.
Noisy kids here now. Limited time doing fault analyzes. Using focusrite solo gen3 for measurements and notice some has had some issues there also.
Hava a picoscope 12bit at hand also.

 

[bambadoo]

Was not quite easy.
Had to gnd the +-15Vdc PSU for the muses volume. I think I got rid of the 100,200hz noise. However 50,150,250 etc. is still there. I am a bit lost here.

Have gone through this thread that looked similar
150 Hz / 250 Hz / 350 Hz noise in linear PSU

BTW: The +-15Vdc PSU are actually 2pc +15Vdc which I connected to commond ground to get +15 -0- -15 voltage.
Like this in a way
Based on TPS7A4700.

Is that a good way of doing it or could that be the reason not having positive and negative regulators...

 

[MagicBus]

Lifting the IEC inlet gnd actually made the 50 and 100hz a bit worse. 150,250,350 etc. were unaffected of this.
Noisy kids here now. Limited time doing fault analyzes. Using focusrite solo gen3 for measurements and notice some has had some issues there also.
Hava a picoscope 12bit at hand also.

Better lift circuit gnd from chassis and verify it's floating. I don't know how muses volume control works but is it as usual sensitive to pot's shaft grounding?

 

[Salas]

You can also try with sourcing Raw DC from the DCSTB (+/0/-). Then LM317/337 or 7815/7915. So everything will be referring to the same system. Maybe the positive centering float no float mix brings some noise (?)

 

[Theodosis O.]

bambadoo, you do not need an extra ps for muses.

You can put 2 leds in series, 2 per rail, to drop +-17v to +-15v, so can use dcstb.

 

[Salas]

True. Easiest test.

 

[bambadoo]

Thanks for replying. I think a need a bit more explanaition regarding tapping DC and reducing voltage by using leds.