Hi!
We are talking here about a preamp with PSU in external chassis, so you can basically put the power transformer as far away from your preamp as needed or as long as your umbilical is. The beauty of separate PSU is that you do not need to worry about the orientation of OPT and power transformers anymore.
Within the PSU the effect of coupling between the power transformers is much less critical. Since there is the subsequent filtering chains this gets eliminated.
Different if you want to put all in one chassis. Then orientation and distance of power supply parts from signal handling parts is critical.
Going external PSU is the easy way out of this.
Best regards
Thomas
We are talking here about a preamp with PSU in external chassis, so you can basically put the power transformer as far away from your preamp as needed or as long as your umbilical is. The beauty of separate PSU is that you do not need to worry about the orientation of OPT and power transformers anymore.
Within the PSU the effect of coupling between the power transformers is much less critical. Since there is the subsequent filtering chains this gets eliminated.
Different if you want to put all in one chassis. Then orientation and distance of power supply parts from signal handling parts is critical.
Going external PSU is the easy way out of this.
Best regards
Thomas
The coupling problem with 2 mains transformers has much to do with the susceptibility of the filament supply.
With the B+ supply, you can attach capacitors from B+ to ground, and wire ground to chassis .
But with filaments, you may (if you have fixed bias) ground one side of the filament, but connecting capacitors across the filament (to filter noise) will destroy the sound quality - the music signal will circulate around the capacitor, and reduce dynamic impedance.
OTOH, if you use my regulators, BOTH filament terminals are buffered from the raw dc supply, so they should be immune to this problem, BUT I still don't like the idea of noise and/or music currents, even at low levels, circulating around wires and chassis. So I avoid connecting caps between the filament raw-dc and chassis for my own amps..
It's always IME better to avoid picking up noise in the first place, rather than filter later, and so I think the best solution is to measure, and design your power supplies - component choice, AND placement - to get the lowest noise.
With the B+ supply, you can attach capacitors from B+ to ground, and wire ground to chassis .
But with filaments, you may (if you have fixed bias) ground one side of the filament, but connecting capacitors across the filament (to filter noise) will destroy the sound quality - the music signal will circulate around the capacitor, and reduce dynamic impedance.
OTOH, if you use my regulators, BOTH filament terminals are buffered from the raw dc supply, so they should be immune to this problem, BUT I still don't like the idea of noise and/or music currents, even at low levels, circulating around wires and chassis. So I avoid connecting caps between the filament raw-dc and chassis for my own amps..
It's always IME better to avoid picking up noise in the first place, rather than filter later, and so I think the best solution is to measure, and design your power supplies - component choice, AND placement - to get the lowest noise.
I don't suppose electrostatic induction is much of an issue so we're probably talking aboult LF electromagnetic induction?
The proposed solution is good and rational from what I remember of field theory...
The psu casing will take the first hit and reduce the EM field, the distance to the main case will reduce the field further and the main case will act as a 3rd barrier offering further reduction.
As usual the thing that peaks my curiosity is when to leave well enough alone?
Would a 1/4" aluminum casing around the transformers be good enough to allow for an in-case build? How about the psu in a separate case underneath the main case? They can look identical and give the look of a matching set and so on...
I like pushing the boundaries, it's a flaw on mine.
These are details seldom discussed and that imho makes it interesting.
The proposed solution is good and rational from what I remember of field theory...
The psu casing will take the first hit and reduce the EM field, the distance to the main case will reduce the field further and the main case will act as a 3rd barrier offering further reduction.
As usual the thing that peaks my curiosity is when to leave well enough alone?
Would a 1/4" aluminum casing around the transformers be good enough to allow for an in-case build? How about the psu in a separate case underneath the main case? They can look identical and give the look of a matching set and so on...
I like pushing the boundaries, it's a flaw on mine.
These are details seldom discussed and that imho makes it interesting.
At low frequency - 50 or 60Hz mains alternations - the problem is mostly magnetic field (B field). But, keep in mind that mains supplies carry a lot of noise, especially in the city, or near industrial works. Some of this may emerge a electrostatic (E field). Noise in a mains supply is likely to be common-mode, so shunting it to safety earth may not help.
Alu case will make a big difference to E field, provided there are no big holes, but B field will march right through Alu (of normal chassis/case thickness).
Your best friend is distance. B field decreases rapidly with distance from its source, so positioning your trafo/rectifier/caps 1 or 2 metre from the sensitive amplifier circuit will help you.
If you use a case to shield the supply, and add some ferrites on the cable (where it exits the case) you can reduce HF E field, too.
To get the best, take the time to measure. you could get some coaxial cable, and connect one end to some enamelled wire - wound into a coil. Buffer the coil with a battery powered opamp, add level control, and connect to your sound card.....( or build Pete's sound card interface).
Now you can see for yourself where the relative positions of field are, using your new sniffer coil.
Alu case will make a big difference to E field, provided there are no big holes, but B field will march right through Alu (of normal chassis/case thickness).
Your best friend is distance. B field decreases rapidly with distance from its source, so positioning your trafo/rectifier/caps 1 or 2 metre from the sensitive amplifier circuit will help you.
If you use a case to shield the supply, and add some ferrites on the cable (where it exits the case) you can reduce HF E field, too.
To get the best, take the time to measure. you could get some coaxial cable, and connect one end to some enamelled wire - wound into a coil. Buffer the coil with a battery powered opamp, add level control, and connect to your sound card.....( or build Pete's sound card interface).
Now you can see for yourself where the relative positions of field are, using your new sniffer coil.
the point is: cheap Iron need more love
Hey Manu,
Just curious why you changed from 45 to 58....
Anything to hide maybe?
PSU heaters
@Rod
I have several questions for PSU heaters:
-I have on hand Pana THA 15.000uF 63V 105ºC 4.92A can I use them?
-Can I do full-wave rectifier with my dual secondary 0-6V 50Va transformers?
-If can do full-wave rect. instead R2 & R3 can I use 2mH 55mohms + 1 resistor 0,47R or other resistor value?
@Rod
I have several questions for PSU heaters:
-I have on hand Pana THA 15.000uF 63V 105ºC 4.92A can I use them?
-Can I do full-wave rectifier with my dual secondary 0-6V 50Va transformers?
-If can do full-wave rect. instead R2 & R3 can I use 2mH 55mohms + 1 resistor 0,47R or other resistor value?
0,5W is OK.
For the very best, I recommend Arcol Carbon Composition in this position. Minimized inductance for best damping of pulses.
RS stock code 386-345
farnell 1265066
These 100 ohm compo are the best type for power valve grid-stoppers (again, minimal inductance), so stock up if you are buying anything else.
For the very best, I recommend Arcol Carbon Composition in this position. Minimized inductance for best damping of pulses.
RS stock code 386-345
farnell 1265066
These 100 ohm compo are the best type for power valve grid-stoppers (again, minimal inductance), so stock up if you are buying anything else.
Ok I will try tants, now I have to wait to arrives RS order.
For B+ I want to use Cree rectifiers Schottky Diode,Z-Rec,SiC,4A,600V,TO-220 (it's necesary the snubber):
Compre Rectificadores Schottky Diode,Z-Rec,SiC,4A,600V,TO-220 Cree C3D04060A en RS Online y lo recibir en 24/48 horas.
For B+ I want to use Cree rectifiers Schottky Diode,Z-Rec,SiC,4A,600V,TO-220 (it's necesary the snubber):
Compre Rectificadores Schottky Diode,Z-Rec,SiC,4A,600V,TO-220 Cree C3D04060A en RS Online y lo recibir en 24/48 horas.