I'm sure this is posted before, but I can't find it...
My un-excercised (and somewhat limited anyway) electronics knowledge is failing me to come up with a relationship to give the ripple current for a given c, r and c. Is there a target max ripple recommended.?
Presumably, there is a downside to larger Rs in this case, apart from heat? I guess it increases the output resistance of the PSU seen by the amp. But again, what becomes detrimental?
My un-excercised (and somewhat limited anyway) electronics knowledge is failing me to come up with a relationship to give the ripple current for a given c, r and c. Is there a target max ripple recommended.?
Presumably, there is a downside to larger Rs in this case, apart from heat? I guess it increases the output resistance of the PSU seen by the amp. But again, what becomes detrimental?
There is free software from Ben Duncan named PSUD2: Power SUpply Designer 2. You'll learn a great deal by trying out lots of different variations in PSUD2 and scrutinizing the results.
One thing you'll learn is: in a power supply, resistors really DO obey Ohm's Law. If you put a current through a resistor, the voltage at one resistor terminal will be different than the voltage at the other terminal. And the difference "deltaV" follows Ohm's Law
* deltaV = (Current in amperes) X (Resistance in ohms)
Since Class A amplifiers draw lots of current at all times (even when the input RCA is grounded!) the term "(Current in amperes)" within Ohm's Law, will be fairly large. So the deltaV will be fairly large too.
One thing you'll learn is: in a power supply, resistors really DO obey Ohm's Law. If you put a current through a resistor, the voltage at one resistor terminal will be different than the voltage at the other terminal. And the difference "deltaV" follows Ohm's Law
* deltaV = (Current in amperes) X (Resistance in ohms)
Since Class A amplifiers draw lots of current at all times (even when the input RCA is grounded!) the term "(Current in amperes)" within Ohm's Law, will be fairly large. So the deltaV will be fairly large too.
I guess I'm missing something as I didn't really think that the resistors wouldn't obey ohms law? I expected that a larger R would drop voltage as you would think given the steady state current draw from the class A bias.
I have found a couple of simulators that give an output for the ripple of a given setup - but I guess it would be more satisfying to know the relationship, the acceptable ripple and (I'm guessing here) the tradeoff with output impedance.
Instinctively C-R-C-R_load looks like a 2nd order low pass?
I have found a couple of simulators that give an output for the ripple of a given setup - but I guess it would be more satisfying to know the relationship, the acceptable ripple and (I'm guessing here) the tradeoff with output impedance.
Instinctively C-R-C-R_load looks like a 2nd order low pass?
I routinely take a DMM and measure AC at the second C in a CRC. It’s fun to change the R and - just as Mark says - see the expected reduction in AC (ripple). For me, I’m trying to get less than 10mv in my CRC. That’s an arbitrary number, but it’s what I shoot for in my 50w SE amp (crappy PSRR).
The maximum tolerable ripple will depend on whether the amplifier is push-pull or single ended. The speaker sensitivity also is a factor. Push-pull amplifiers with their "push" and "pull" outputs will cancel power supply ripple so they will tolerate higher amounts of ripple whereas single ended amplifiers do not so they will need lower ripple power supplies.
When I build single ended amplifiers I use CLC supplies to keep the ripple level low.
As Mark mentioned, playing with PSUDII is a good way to learn how C and R affect ripple. L can also be inserted in place of R to see its effect on ripple.
When I build single ended amplifiers I use CLC supplies to keep the ripple level low.
As Mark mentioned, playing with PSUDII is a good way to learn how C and R affect ripple. L can also be inserted in place of R to see its effect on ripple.
It’s push pull - Aleph J. Probably ending up biased to up to 3A per rail (monoblock).
So if I target a certain ripple - e.g. 20-30mV the best thing to do is draw it up in a sim and play with values to get there? Presumably starting with a suitably large capacitance.
So if I target a certain ripple - e.g. 20-30mV the best thing to do is draw it up in a sim and play with values to get there? Presumably starting with a suitably large capacitance.
Of course. But my F5 with F5 PSU as per has just acceptable hum on 89dB speakers. So I was looking to calculate the ripple on my F5, set a lower target and design accordingly.
89dB isn't overly sensitive for speakers.
Power supply ripple is not the only cause of hum. An amplifier with reasonable power supply ripple can still hum if there are other issues.
Power supply ripple is not the only cause of hum. An amplifier with reasonable power supply ripple can still hum if there are other issues.
First Watt F5 Manual
"The noise of a properly laid out circuit is about 30 uV with a quiet supply, and raw supply with two 29,000 uF capacitor and 70 mV ripple will give about 100 uV noise (measured in the band from 20 to 20,000 Hertz)."
The manual also has a schematic of the power supply.
So 29,000uF - 0.12R - 29,000uF CRC and 100uV noise.
100uV is good.
First Watt Amplifers - Specs including Noise
"The noise of a properly laid out circuit is about 30 uV with a quiet supply, and raw supply with two 29,000 uF capacitor and 70 mV ripple will give about 100 uV noise (measured in the band from 20 to 20,000 Hertz)."
The manual also has a schematic of the power supply.
So 29,000uF - 0.12R - 29,000uF CRC and 100uV noise.
100uV is good.
First Watt Amplifers - Specs including Noise
