I want to change the power supply in my 300B SET. With a better SS rectifier.
The current Transformer is a 2x400V CT (Hammond 300BX), if I replace the Valve Rectifier I will have at least 75V excess voltage.
I need around 440-450V on the B+ side. with an SS I think I will get 550-600V on the B+Side.
The option in my mind is to put a voltage drop using resistor which is not a good solution or put a zener diode which might impact sonic quality as well. I don't want to tweak the transformer side and need a more solid unregulated B+ side.
Anybody has done this before? And what us the rcommendation?
The current Transformer is a 2x400V CT (Hammond 300BX), if I replace the Valve Rectifier I will have at least 75V excess voltage.
I need around 440-450V on the B+ side. with an SS I think I will get 550-600V on the B+Side.
The option in my mind is to put a voltage drop using resistor which is not a good solution or put a zener diode which might impact sonic quality as well. I don't want to tweak the transformer side and need a more solid unregulated B+ side.
Anybody has done this before? And what us the rcommendation?
Last edited:
Why will a solid state rectifier be better?
Or will it just be different.just different.
What is wrong with the sound that you are trying to change?
Or will it just be different.just different.
What is wrong with the sound that you are trying to change?
Hi vincenti,
Use a resistor before the first filter capacitor. This resistor will extend the life of your filter capacitors. The voltage drop will be greater than the number you might expect from using straight ohm's law. Try using PSUD II to design this section. You will be creating an R-C filter right out of the gate.
-Chris
Use a resistor before the first filter capacitor. This resistor will extend the life of your filter capacitors. The voltage drop will be greater than the number you might expect from using straight ohm's law. Try using PSUD II to design this section. You will be creating an R-C filter right out of the gate.
-Chris
Valve rectifier is inefficient, heat generating, easy to break components in the circuit. I think valve rectification is an obsolete technology and has no sonic contibution in audio system.
By keeping the half rectification I think we won't alter the OT behaviour. A bigger capacitor with bleeder in series and fast recovery ss rectifiers should give a stable, unregulated B+.
Why just get rid of that lamp and have a cool, well behave amp?
By keeping the half rectification I think we won't alter the OT behaviour. A bigger capacitor with bleeder in series and fast recovery ss rectifiers should give a stable, unregulated B+.
Why just get rid of that lamp and have a cool, well behave amp?
vincenti… there are literally hundreds of articles (here) that dive deeply into both the real and perceived differences between solid state and vacuum rectification.
The ULTRA-SHORT answer is: solid-state rectification can have surprisingly "loud" reverse recovery oscillations that get into the power supplied. Choosing ultra-fast recovery diodes is key, along with adding some modest recovery-ringing squelching components.
Yet the simplicity of the vacuum valve is remarkable: whether you view it as a bug (as you do) or a feature, the vacuum valve delivers nearly flyback-ringing free pulsed DC, and a significant forward voltage drop. All in one cute package.
So… like someone previously said, “why change? just to change itself?” (paraphrased)
IF YOU DO, there are other facts:
[1] There is no downside for using a string of modest-voltage Zeners to deliver the voltage drop, so long as you know the average amperage, and don't exceed about 60% of each Zener's working power dissipation. For instance, if the HV is running nominally at 120 ma, and the Zeners are 1 watt jobs, then you can use strings of 8 × 0.6 = 4.8 volts Zeners to deliver the ± 70 V drop. 14 in series will do it. None of them will overheat (unless they're all bundled together!).
[2] Having a very modest series inductor (post rectification), which is bypassed by a modestly low value resistor (or any other snubber circuit) is recommended. Just look up snubber, and see what a constellation of ideas comes forth.
[3] Use very high PIV diodes in general. They're identical in price as the lower PIV, and seem not to have significantly noise or other features. ERR on the side of caution. Remember, SS rectifiers fail by exploding. Tubes, by getting cranky for a few cycles.
Anyway, best of luck.
Personally, I'd stick with the tubes.
But I'm old, and cranky, and cheeky to boot.
GoatGuy
The ULTRA-SHORT answer is: solid-state rectification can have surprisingly "loud" reverse recovery oscillations that get into the power supplied. Choosing ultra-fast recovery diodes is key, along with adding some modest recovery-ringing squelching components.
Yet the simplicity of the vacuum valve is remarkable: whether you view it as a bug (as you do) or a feature, the vacuum valve delivers nearly flyback-ringing free pulsed DC, and a significant forward voltage drop. All in one cute package.
So… like someone previously said, “why change? just to change itself?” (paraphrased)
IF YOU DO, there are other facts:
[1] There is no downside for using a string of modest-voltage Zeners to deliver the voltage drop, so long as you know the average amperage, and don't exceed about 60% of each Zener's working power dissipation. For instance, if the HV is running nominally at 120 ma, and the Zeners are 1 watt jobs, then you can use strings of 8 × 0.6 = 4.8 volts Zeners to deliver the ± 70 V drop. 14 in series will do it. None of them will overheat (unless they're all bundled together!).
[2] Having a very modest series inductor (post rectification), which is bypassed by a modestly low value resistor (or any other snubber circuit) is recommended. Just look up snubber, and see what a constellation of ideas comes forth.
[3] Use very high PIV diodes in general. They're identical in price as the lower PIV, and seem not to have significantly noise or other features. ERR on the side of caution. Remember, SS rectifiers fail by exploding. Tubes, by getting cranky for a few cycles.
Anyway, best of luck.
Personally, I'd stick with the tubes.
But I'm old, and cranky, and cheeky to boot.
GoatGuy
Hi azazello,
The resistor limits the peak currents and drops excess voltage when you need to do that (which would be every time you replace a tube rectifier). This damps the ringing and extends the life of all the other components. Sounds like all positive things to me. Finally, when clients actually sit and listen to the results, they are sold.
Increasing the input capacitance is the last thing you want to do. This does everything opposite to the first method. That and it decreases the conduction angle for charging, and that encourages the ringing and hash coming from the rectifier setup. That forces more expensive rectifiers and special types. Don't forget too, one of the jobs a power supply has to do is to isolate the circuits they run from outside noise on the line. Think about that for a little bit and the direction to take should become self evident.
-Chris
And why would you say that? This isn't an idea, it is the basic plan when replacing tube rectifiers. I even do this with lower voltage solid state supplies, and it works more than just fine.
The resistor limits the peak currents and drops excess voltage when you need to do that (which would be every time you replace a tube rectifier). This damps the ringing and extends the life of all the other components. Sounds like all positive things to me. Finally, when clients actually sit and listen to the results, they are sold.
Increasing the input capacitance is the last thing you want to do. This does everything opposite to the first method. That and it decreases the conduction angle for charging, and that encourages the ringing and hash coming from the rectifier setup. That forces more expensive rectifiers and special types. Don't forget too, one of the jobs a power supply has to do is to isolate the circuits they run from outside noise on the line. Think about that for a little bit and the direction to take should become self evident.
-Chris
This has to be the best way to get rid of excess voltage.
most PSU are rC and some are rCRC.
Changing that first r to R by deliberately making it bigger and much more effective is a good upgrade to any PSU where there is excess voltage.
Make that first filter actually work for you.
PSUD2 will show the differences in performance.
I did it for the Putzeys balanced volume control. Changed the 2r2 to 10r and saw what a big improvement in reduced ripple sent to the rectifiers. Then looked at changing both 2r2 to 10r. I finally adopted 20r>1mF>20r>1mF>regulators for my second board which uses standard 3pin regulators, instead of the Putzeys' special 5pin regulators.
Last edited:
I know PSUD II but I thought it won't work in modern OS but it works in Windows 10 emulation mode. It was damned easy. I tried it with Multisim but it doesn't have a ready to use model for choke.
After some simulations I think this will be a nice setup for 300B SET:
1 Just add additional 120 Ohm 25W after the diode to get 40V drop and replace 47u+47u filter capacitors with 220u+220u
2. Maybe I just add additional switch for rectification mode selector, that will be nice for comparison in front of friends.
After some simulations I think this will be a nice setup for 300B SET:
1 Just add additional 120 Ohm 25W after the diode to get 40V drop and replace 47u+47u filter capacitors with 220u+220u
2. Maybe I just add additional switch for rectification mode selector, that will be nice for comparison in front of friends.
Hi Andrew,
This is especially effective for high voltage tube circuits. Even more so with power supplies that used higher than recommended input capacitance for the rectifier tube. That maximum is 47 - 50 uF (same capacitance given the range from stated capacitance to real measured capacitance, -25% to + 50% these days).
I recently applied this to a solid state preamplifier, and the results were every bit as impressive as in tube circuits.
-Chris
This is especially effective for high voltage tube circuits. Even more so with power supplies that used higher than recommended input capacitance for the rectifier tube. That maximum is 47 - 50 uF (same capacitance given the range from stated capacitance to real measured capacitance, -25% to + 50% these days).
I recently applied this to a solid state preamplifier, and the results were every bit as impressive as in tube circuits.
-Chris
- Status
- Not open for further replies.