Just wondering what you have found to sound the best in your power supplies. Should the first filter cap after the diode bridge be large or small for the best sound? This can apply to both solid state and tube equipment. Some people like a large cap after the bridge and some do not, what do you recommend?
If we are talking about a typical power supply with no series resistors or chokes, then large caps. near the bridge and small caps near the audio circuit. It's not about best sound (directly) it's about good circuit design.
Neither. The best size.
Typically 2000uF at 12V, 1A, 120Hz, I think for 10% ripple. C = 2.88F * IDC / (VDC * F), where IDC is maximum DC load current in amperes, VDC is the desired DC voltage, and F is the ripple frequency (twice line frequency for full wave rectifiers).
If this is too much ripple, add additional filtering (pref. CLC), regulate (pref. switching reg at high power), or just build a better circuit. Anything with low PSRR isn't worth listening to anyway.
Big capacitors worsen power factor considerably. This invites ground loop induced hum (from the huge peak currents) and reduces power available from a given transformer or line circuit. With the above value, power factor is typically around 0.5, depending on source impedance. A 100VA transformer will only deliver 50V 1A (= 50W) within ratings, much less if a huge cap is used. A 2kW amplifier cannot operate continuously from a 120V, 15A circuit because the input will be 4kVA! This is an excellent reason to supply high-power amps from a PFC supply.
Tim
Typically 2000uF at 12V, 1A, 120Hz, I think for 10% ripple. C = 2.88F * IDC / (VDC * F), where IDC is maximum DC load current in amperes, VDC is the desired DC voltage, and F is the ripple frequency (twice line frequency for full wave rectifiers).
If this is too much ripple, add additional filtering (pref. CLC), regulate (pref. switching reg at high power), or just build a better circuit. Anything with low PSRR isn't worth listening to anyway.
Big capacitors worsen power factor considerably. This invites ground loop induced hum (from the huge peak currents) and reduces power available from a given transformer or line circuit. With the above value, power factor is typically around 0.5, depending on source impedance. A 100VA transformer will only deliver 50V 1A (= 50W) within ratings, much less if a huge cap is used. A 2kW amplifier cannot operate continuously from a 120V, 15A circuit because the input will be 4kVA! This is an excellent reason to supply high-power amps from a PFC supply.
Tim
Small cap gives low voltage, high ripple. Big cap gives large peak currents, which get into audio and overheat cap and transformer. As Sch3mat1c says, pick the right one for your design. Don't follow fashion or your favourite guru. All electronic design is a compromise.
In the past I have picked the smallest uf value that would keep the voltage across it up to near the calculated value of X 1.414, or peak charge. I have seen some designers use a small uf value for the first cap in order to keep down RFI and simulate a vacuum tube rectifier. The larger the first cap the longer the current charge pulse gets in time. There is not always a direct correlation between measured performance and the best sound, so that’s why I was asking for feedback on what sounded the best. It seems to me that the greater the value of the first filter cap the stronger and better defined the bass gets in sound.
No, a large cap means the charge pulse get shorter in time but larger in amplitude. It might help reduce droop, which in some circuits might help bass. This depends on the resistance of the transformer etc. The second capacitor would have a greater effect on the sound.
To simulate a valve rectifier you would add a series resistor. This reduces voltage, but also lengthens and reduces the height of the charge pulse. So it has a similar effect to a smaller cap, but does it in a slightly different way.
I think your usual practice of using the smallest first cap which will maintain the required voltage is probably the best option. The place to put big caps is in the second position.
To simulate a valve rectifier you would add a series resistor. This reduces voltage, but also lengthens and reduces the height of the charge pulse. So it has a similar effect to a smaller cap, but does it in a slightly different way.
I think your usual practice of using the smallest first cap which will maintain the required voltage is probably the best option. The place to put big caps is in the second position.
Hello Rick,
What kind of equipment are you thinking about: amplifier or preamplifier (dac...)?
Power factor is a measure of how much in phase the current is with the voltage in an AC circuit. Sticking a big capacitor across the mains can draw a large current but dissipate relatively low power - this means a poor power factor as there are losses in the wiring associated with the large current flow even though the load has low dissipation (power consumption). Poor power factors are bad news for suppliers as consumers by and large aren't paying for the wasted energy in the wiring.
What little I think I know about power factor is to do with the phase difference between current and voltage. Industrial loads tend to be inductive and as you have said this uses current and develops current losses that are not directly related to the power consumed.
I gather that capacitors can be used to bring the phase of the current closer to that of the voltage and thereby reduce the wasted I^2R current losses that are not power.
But what is schem referring to?
That statement seems to be something quite different, and Abrax seems to be going down the same route, by saying that capacitors make the problem worse.
Explain what you are referring to.
I gather that capacitors can be used to bring the phase of the current closer to that of the voltage and thereby reduce the wasted I^2R current losses that are not power.
But what is schem referring to?
That statement seems to be something quite different, and Abrax seems to be going down the same route, by saying that capacitors make the problem worse.
Explain what you are referring to.
Last edited:
Yes, this is correct - the supply companies strongly prefer consumers to connect appliances with near to unity power factor - that means resistive loads are best.
Inductors draw current but very little power is consumed in the load. Suppliers suffer losses due to heating relative to the square of the current drawn, not relative to the power taken. They charge domestic consumers according to power, not current.
If a company has loads which are mainly inductive, putting capacitors across the mains can improve the power factor. Inductors have lagging current, capacitors leading current. With the right value added cap, the leading can cancel the lagging so the result just looks near enough a resistor.
I am not sure about this, but I believe that industrial companies pay their electric bills according to VA, not Watts. Does anyone know to confirm this? If its true, then the company will pay lower bills by correcting their power factor.
He's saying a big cap used on the secondary of a transformer makes for a poor power factor - large current flow, larger stray fields and more heating in the transformer. So he's discouraging specifying too big caps on the secondary.
Any clearer now?
Hi,
They are simply saying the short large peak rectified currents for an oversized
capacitor will cause more resistive losses than a "correctly" sized capacitor.
In a pre-amp / CD player this hardly matters, and CRC filtering can be used.
Small then Big in this case is poor due to the supply wiring AC ripple currents.
It does matter in high power applications, it not true that you "cannot
make the filtering capacitors too big", you can, causing excessive losses.
I agree with you though, power factor is not "conversion efficiency",
though this what is often meant when the term "power factor" is used.
Lots of wasteful resistive losses will actually improve "power factor",
but not "conversion efficiency" or "maximum supply power throughput".
rgds, sreten.
Last edited:
Sreten,
are you saying that the use of the phrase "Power Factor" has nothing to do with what Schem is referring to?
BTW,
I started using psud many years ago. It clearly shows the short term current peak when voltage after the rectifier is above the voltage stored in the smoothing capacitor. That never confused me.
I also know that most transformer manufacturers recommend a transformer de-rating factor when the transformer feeds a capacitor input filter. I have seen DF between ~65% and ~75% and that is what I use to determine the maximum continuous DC current that can be drawn from the smoothing capacitors. This usually comes out at ~half the maximum AC current.
are you saying that the use of the phrase "Power Factor" has nothing to do with what Schem is referring to?
BTW,
I started using psud many years ago. It clearly shows the short term current peak when voltage after the rectifier is above the voltage stored in the smoothing capacitor. That never confused me.
I also know that most transformer manufacturers recommend a transformer de-rating factor when the transformer feeds a capacitor input filter. I have seen DF between ~65% and ~75% and that is what I use to determine the maximum continuous DC current that can be drawn from the smoothing capacitors. This usually comes out at ~half the maximum AC current.
Last edited:
Somehow they seem to write about the DC power supply then about the AC power supply. It's hard to keep track.
I thought that the original question was about main reserve supply capacitors and stage decoupling capacitors. But now I'm not sure.
I thought that the original question was about main reserve supply capacitors and stage decoupling capacitors. But now I'm not sure.
I'd be surprised if it wasn't all trafo manufacturers. A capacitive load's power factor after rectification to some degree depends on the ripple voltage - higher ripple means better (higher) PF.
If by any faint chance 'they' includes me, feel free to highlight the bits in my writing where you're getting confused. I realize my writing skills leave room for improvement in clarity😀
- Status
- Not open for further replies.