Just looking for a critique/common sense check for the custom toroidal order I want to submit later this week. Sorry for the amount of info but it's complex hence asking for help.
Future requirement - to be able to use this transformer to test prototype modules for a later amp using tube based opamps etc. However for now this is V1:
V1 Amp requirements - this is the amp's requirements.
+320V at the B+ rail, 120mA total (per ch: 20mA max for cascode, 2x20mA for drivers)
-320V at the B- rail, 40mA total (per ch: 20mA for cascode only, drivers are to ground)
+200V at the output B+ rail, 480mA (per ch: 4x60mA)
-200V at the output B- rail, 480mA (per ch: 4z60mA)
Each supply will ultimately have a 9H choke then a maida regulator. I will find out the choke resistance but for now I'm using spec for a 9H toroidy choke (29ohm) and then putting ~30V drop for the maida. The supply is full wave rectified hence 1.414 voltage increase and I'm making an assumption of 0.60 current drop.
Thus the power supply requirement for the 320V rails at 160mA:
320 + 30 + (0.16*29) = 355V, then /1.414 = 251V voltage from the secondary.
160mA / 0.60 = 266mA required from the secondary.
Thus ~ 215*0.266 = ~ 60VA minimum
For the power supply for the 200V rails at 480mA:
200 + 30 + (0.48*29) = 244V, then /1.414 = 172V voltage from the secondary
480mA / 0.60 = 800mA required from the secondary.
Thus 172*0.8 = ~140VA minimum
Heater requirements
12.6V Heater referenced to +50V, 3.4A total from the following:
* 2x 3x300mA
* 2x 2x400mA
12.6V Heater referenced to -150V, 2.4A total from the following:
* 2x 1x300mA
* 2x 2x400mA
This will be regulated DC, again this will likely be a RC (5ohm) with a maida (10V drop configured), full wave rectified. Just taking the larger current at 3.4A:
12.6V + 10V + 3.4*5 = 40V, then /1.414 = 28.2V or 30V
3.4A / 0.60 = 5.7A
30 * 5.7 = 171VA
So anything I may have forgotten (perhaps 2-3V for the diode drop?) or observations?
I'd like to finalise the order later this week.
Future requirement - to be able to use this transformer to test prototype modules for a later amp using tube based opamps etc. However for now this is V1:
V1 Amp requirements - this is the amp's requirements.
+320V at the B+ rail, 120mA total (per ch: 20mA max for cascode, 2x20mA for drivers)
-320V at the B- rail, 40mA total (per ch: 20mA for cascode only, drivers are to ground)
+200V at the output B+ rail, 480mA (per ch: 4x60mA)
-200V at the output B- rail, 480mA (per ch: 4z60mA)
Each supply will ultimately have a 9H choke then a maida regulator. I will find out the choke resistance but for now I'm using spec for a 9H toroidy choke (29ohm) and then putting ~30V drop for the maida. The supply is full wave rectified hence 1.414 voltage increase and I'm making an assumption of 0.60 current drop.
Thus the power supply requirement for the 320V rails at 160mA:
320 + 30 + (0.16*29) = 355V, then /1.414 = 251V voltage from the secondary.
160mA / 0.60 = 266mA required from the secondary.
Thus ~ 215*0.266 = ~ 60VA minimum
For the power supply for the 200V rails at 480mA:
200 + 30 + (0.48*29) = 244V, then /1.414 = 172V voltage from the secondary
480mA / 0.60 = 800mA required from the secondary.
Thus 172*0.8 = ~140VA minimum
Heater requirements
12.6V Heater referenced to +50V, 3.4A total from the following:
* 2x 3x300mA
* 2x 2x400mA
12.6V Heater referenced to -150V, 2.4A total from the following:
* 2x 1x300mA
* 2x 2x400mA
This will be regulated DC, again this will likely be a RC (5ohm) with a maida (10V drop configured), full wave rectified. Just taking the larger current at 3.4A:
12.6V + 10V + 3.4*5 = 40V, then /1.414 = 28.2V or 30V
3.4A / 0.60 = 5.7A
30 * 5.7 = 171VA
So anything I may have forgotten (perhaps 2-3V for the diode drop?) or observations?
I'd like to finalise the order later this week.
So I worked through this, got some more figures from the Tiger Toroids here in the UK plus had a look at both PSUD2 and run the figures in ltspice along with the full model of the amp.
320V rails
* Running 230V ±10% with a 15 ohm primary winding
* Secondary windings are about 20-30ohms (I used 25ohms)
* 8% regulation
200V rails
* Running 230V ±10% with a 9 ohm primary winding
* Secondary windings are 10 ohms
* 6% regulation
Now one thing that I messed up in my calculation is I'm using a full wave bridge rather than a full wave rectifier.
Now the 9H choke comes in at DC resistance of 64ohms for 300mA.
For ~340V rails (gives some wiggle room for additional RC etc)
~243Vrms secondary, with 47uF input cap, gives 358V post bridge, ~352V post choke with 2x350uF caps and then ~345V post regulator.
In short a drop of 10V seems very very efficient, so I suspect that I would see a drop from 358V by about 30V, possibly down to 330V. Either way it allows me to use a resistor to get to where I need.
Once aspect looking at the design is that the 12BH7A has a maximum of 300V across it, so I will have to tune the driver circuit a little so that it has a about 240V across the tube - this means that the 18% total regulation will never go above the 300V limit. Not a biggie but something to be careful of. The cascode 12BH7As are fine as they have 150V across each.
The ecc99s are 400V max but they sit with two tubes between +200V and -200V.
The 200V rails seemed be easier:
~170Vrms seems to give 223V post choke, ~216V at the rail.
So given Stephen (TT) said the choke used 6-7watts (power=v*i), I would assume for 0.300mA that would be about 20V drop..
Also going above 300mA complicates things with a large core etc, so an option here is to use lower inductance but use two after a channel split thus each would only be using 240mA.
So I still need some time when I'm not tired to think this through more. Seems a smaller choke would be better.
320V rails
* Running 230V ±10% with a 15 ohm primary winding
* Secondary windings are about 20-30ohms (I used 25ohms)
* 8% regulation
200V rails
* Running 230V ±10% with a 9 ohm primary winding
* Secondary windings are 10 ohms
* 6% regulation
Now one thing that I messed up in my calculation is I'm using a full wave bridge rather than a full wave rectifier.
Now the 9H choke comes in at DC resistance of 64ohms for 300mA.
For ~340V rails (gives some wiggle room for additional RC etc)
~243Vrms secondary, with 47uF input cap, gives 358V post bridge, ~352V post choke with 2x350uF caps and then ~345V post regulator.
In short a drop of 10V seems very very efficient, so I suspect that I would see a drop from 358V by about 30V, possibly down to 330V. Either way it allows me to use a resistor to get to where I need.
Once aspect looking at the design is that the 12BH7A has a maximum of 300V across it, so I will have to tune the driver circuit a little so that it has a about 240V across the tube - this means that the 18% total regulation will never go above the 300V limit. Not a biggie but something to be careful of. The cascode 12BH7As are fine as they have 150V across each.
The ecc99s are 400V max but they sit with two tubes between +200V and -200V.
The 200V rails seemed be easier:
~170Vrms seems to give 223V post choke, ~216V at the rail.
So given Stephen (TT) said the choke used 6-7watts (power=v*i), I would assume for 0.300mA that would be about 20V drop..
Also going above 300mA complicates things with a large core etc, so an option here is to use lower inductance but use two after a channel split thus each would only be using 240mA.
So I still need some time when I'm not tired to think this through more. Seems a smaller choke would be better.