Hello,
I am not a technician but " killing " most of the ripple with a basic RC or LC filter before " feeding " it into the regulator cannot be bad. If you have the space available a low current choke would be my choice.
Greetings,Eduard
I am not a technician but " killing " most of the ripple with a basic RC or LC filter before " feeding " it into the regulator cannot be bad. If you have the space available a low current choke would be my choice.
Greetings,Eduard
Not necessary. PSRR won't much be improved, the LM317 does all the work.
Saying this, Elvee's "de-noiser" boost all three figure of merit quotients, PSRR, noise and output impedance!
Saying this, Elvee's "de-noiser" boost all three figure of merit quotients, PSRR, noise and output impedance!
Hello,
I have used several regulated power supplies both series and shunt types . ALL more sophisticated than the lm317 and all would show an improvement after adding some kind of preregulation.
If you have a circuit that takes a low current and uses a single cap adding an extra cap plus resistor or choke could improve things. Adding a RC network is cheap and easy and if it gives an improvement probably changing from RC to LC will give an improvement too.
Of course you will only know if you try it.
IF the lm317 would be this perfect the de noiser would not be " necessary "
Greetings,Eduard
I have used several regulated power supplies both series and shunt types . ALL more sophisticated than the lm317 and all would show an improvement after adding some kind of preregulation.
If you have a circuit that takes a low current and uses a single cap adding an extra cap plus resistor or choke could improve things. Adding a RC network is cheap and easy and if it gives an improvement probably changing from RC to LC will give an improvement too.
Of course you will only know if you try it.
IF the lm317 would be this perfect the de noiser would not be " necessary "
Greetings,Eduard
All active regulation runs out of available feedback gain at high frequency; the simple 3-pin regs like LM3x7s very early, more sophisticated and discrete regulators... later.
When the voltage drop remains reasonable, I've never regretted using C-R-C ahead of the regulator, or C-RL-C - an inductor with enough (added ) resistance that the result is more than critically-damped (this matters, too, in my experience.) Even really quite small values or R (and L) are with experimenting with!
When the voltage drop remains reasonable, I've never regretted using C-R-C ahead of the regulator, or C-RL-C - an inductor with enough (added ) resistance that the result is more than critically-damped (this matters, too, in my experience.) Even really quite small values or R (and L) are with experimenting with!
Hello,
Because the original poster wants to use it for a tube preamp current will be low and the actual output voltage of the regulator can probably be let us say 10 volt lower without any problem. So it wont be difficult to add an rc or lc network.
Of course one can start with a rc network and keep some space available for a 50mA? Choke with 5 to 10 Henry.
Of course if high voltage transformer has not yet been bought one could go for a choke input. Even a tube rectifier could be used.
Greetings,Eduard
Because the original poster wants to use it for a tube preamp current will be low and the actual output voltage of the regulator can probably be let us say 10 volt lower without any problem. So it wont be difficult to add an rc or lc network.
Of course one can start with a rc network and keep some space available for a 50mA? Choke with 5 to 10 Henry.
Of course if high voltage transformer has not yet been bought one could go for a choke input. Even a tube rectifier could be used.
Greetings,Eduard
I haven't used a pi filter even with phono preamp. My power amp uses SS rectifier, 3000 uF cap and Maida for the driver stage and screens. But if you do this with a power amp, it's a good idea to periodically check ripple and be sure it's still within specs of the caps.
Last edited:
Hello,
Try simulating with Duncan psu designer software.
Do one with a single 3000 microfarad cap and do another with CRC network 220 microfarad 470 ohm and 1000 microfarad cap. These are just 3 random values but they will probably show that CRC will be better than a single big C.
Greetings,Eduard
Try simulating with Duncan psu designer software.
Do one with a single 3000 microfarad cap and do another with CRC network 220 microfarad 470 ohm and 1000 microfarad cap. These are just 3 random values but they will probably show that CRC will be better than a single big C.
Greetings,Eduard
No doubt. What is the heat wasted in the 470 ohm resistor and what is the required rating? The PSRR of the Maida regulator is good enough that you can tolerate a fair amount of ripple. The power supply has better current capability. I did it as a "Hmmm, what if?" experiment, and it works well.
Last edited:
The advantages of using Maida in a tube preamp is not only better sound quality, but also greater power supply flexibility with an adjustable B+ and the PSRR. The output RC network is fairly critical for transient performance, and based on listening tests, I think the original 2.7 ohm 1uF is just about right. I never tried the "Precision High Voltage Regulator" in Figure 2, which has better performance. That may be a better regulator to use. For the voltage divider in the circuit, we would calculate fixed resistors and then place a 20 turn trimmer in parallel with the bottom resistor. Important in power amps operating at higher voltages, because of the power dissipation in the bottom resistor. We used a wire wound for that resistor, which also gives better stability. As the bottom resistor heats up, the resistance decreases, which changes the set voltage. You can then use a parallel trimmer of lower rating.
Hundreds of ohms and thousands of uf are a stupid place to go for HV psu ahead of active regulation.
Two aspects:
1. 1000uf with a typical valve supply ' 2-3hundred v DC' - is an immense, dangerous amount of energy storage that needs appropriate management. If you are willing to do such - you really don't need a 'maida- style regulator; it' is the least part of the problem. And you really already ought know this and so won't be reading this thread (which is why I point out the danger ...)
2. Be very , very wary of using more than only a handful of ohms supply impedance ahead of a LM3x7 type otherwise: you can run into stability issues.
Two aspects:
1. 1000uf with a typical valve supply ' 2-3hundred v DC' - is an immense, dangerous amount of energy storage that needs appropriate management. If you are willing to do such - you really don't need a 'maida- style regulator; it' is the least part of the problem. And you really already ought know this and so won't be reading this thread (which is why I point out the danger ...)
2. Be very , very wary of using more than only a handful of ohms supply impedance ahead of a LM3x7 type otherwise: you can run into stability issues.
I hear you. 375 Joules. It has been working well for over a decade. No stability issues. Admit that it does make me nervous, and I periodically monitor the health of the system. It has been rock solid. One nice safety feature of the Maida is that it drains the caps when power is off.
Hello,
OF COURSE with a high voltage supply that can store a lot of energy it is mandatory that the caps will be empty within a minute or so after switching off.
Sometimes you can read that with modern caps available with lots of microfarads there is no need to use RC or LC network. Some say no use for regulator at all. Probably for pre amp or driver stage a regulator with some pre filtering will be my choice.
I remember reading about the disadvantages of using a BIG cap right after the rectifier. Of course modern day rectifiers can take a lot of abuse and they will survive but if you see the scope pictures there will be a pulsating current much higher than the nominal current flow. The only way to solve this is using a smaller cap or use a properly designed choke input which will keep the current flow at a steady rate.
If there are resistors getting hot in the regulator circuit it is always a good idea to use something like a Mills 12 watt wirewound. If 12 watt not enough there are other brands available.
Morgan Jones wrote about a similar network at the output of a regulator. Adding parts at the output of the regulator is tricky if you dont have the right tools and knowledge to check the effects. So better just copy the people who know. At least that is what i do.
Greetings,Eduard
OF COURSE with a high voltage supply that can store a lot of energy it is mandatory that the caps will be empty within a minute or so after switching off.
Sometimes you can read that with modern caps available with lots of microfarads there is no need to use RC or LC network. Some say no use for regulator at all. Probably for pre amp or driver stage a regulator with some pre filtering will be my choice.
I remember reading about the disadvantages of using a BIG cap right after the rectifier. Of course modern day rectifiers can take a lot of abuse and they will survive but if you see the scope pictures there will be a pulsating current much higher than the nominal current flow. The only way to solve this is using a smaller cap or use a properly designed choke input which will keep the current flow at a steady rate.
If there are resistors getting hot in the regulator circuit it is always a good idea to use something like a Mills 12 watt wirewound. If 12 watt not enough there are other brands available.
Morgan Jones wrote about a similar network at the output of a regulator. Adding parts at the output of the regulator is tricky if you dont have the right tools and knowledge to check the effects. So better just copy the people who know. At least that is what i do.
Greetings,Eduard
Stability of the LM317 or Maida regulator is influenced by the ESR of the output capacitor and the load current. Resistance ahead of the regulator can burn off some excess voltage but has little to no effect upon stability.