With regards to power supply capacitor specifications,
I always read use "low ESR", "high ripple current" but no one ever quantifies that.
What is "low ESR" and what is not "low ESR"?
Is there too low an ESR?
Same with "ripple current".
Can ripple current be too high?
Let's assume (cause this is what I am doing), a PS with a 300VA dual 36v toroidal feeding a 100 watt mono AB amp using four 10,000uf caps at 63v.
This is the PSU circuit diagram,
Octopart lists about 100 different 10,000uf 63v capacitors,
100 different 10000uf 63v capacitors
with an ESR between 12mΩ and 80mΩ and a ripple current between 3.72 amps and 17 amps, priced between $2.20 and $75.00 each.
What would be considered good ESR and ripple current?
What would be bad ESR and ripple current?
Thank you, David.
I always read use "low ESR", "high ripple current" but no one ever quantifies that.
What is "low ESR" and what is not "low ESR"?
Is there too low an ESR?
Same with "ripple current".
Can ripple current be too high?
Let's assume (cause this is what I am doing), a PS with a 300VA dual 36v toroidal feeding a 100 watt mono AB amp using four 10,000uf caps at 63v.
This is the PSU circuit diagram,
Octopart lists about 100 different 10,000uf 63v capacitors,
100 different 10000uf 63v capacitors
with an ESR between 12mΩ and 80mΩ and a ripple current between 3.72 amps and 17 amps, priced between $2.20 and $75.00 each.
What would be considered good ESR and ripple current?
What would be bad ESR and ripple current?
Thank you, David.
A lot of this is 'relative' and I doubt you will find any list of numbers where low ESR begins and not so low starts.
ESR can apparently be 'to low' for some applications such as IC voltage regulators and causes issues with stability. In other words they suggest for example a standard commercial type 10 or 47uF cap and not some specialised ultra low ESR component.
Ripple current can never be to high as it is a maximum the device the cap can tolerate, however you need to check it is specified in the same way when comparing.
Generally the bigger physically a cap is and the higher the ripple current allowable.
If the ESR is 'high' and the ripple current 'high' then the power dissipated in the cap is also 'high' and so it warms.
For your Class AB these concerns are so far off the radar as not to be an issue. If it were Class A and say pulling 10 amps continuously from the supply then you would need to look more closely at what you use. When your Class AB amp is idling the ripple current in the caps will be pretty small. Under full sine drive into say 8 ohm it will rise.
To ensure that high ESR never becomes a problem... because high ESR can limit the caps effectiveness at high frequency particularly in decoupling SMPS rails... you should also add in parallel a more consistent low ESR part such as a small film cap of say 0.22uF or a 'defined' low impedance such as a film cap in series with a low value non inductive resistor.
Rather than looking for 'good' and 'bad' ESR, try and think in terms of looking for parts that will maintain their spec over years and decades and design the system so that an increase in ESR of say 100 or 200% as things age will cause no issues... such as adding the bypass caps.
I have a simulation of a power supply and amp set up. This shows the current in one side of the power supply (the reservoir cap) and the cap is just 2200uF in size.
You can see the 1kHz audio flowing 'in' the cap and also the larger charging pulses. First image is at around 1 watt output and second image at nearer 50 watts.
ESR can apparently be 'to low' for some applications such as IC voltage regulators and causes issues with stability. In other words they suggest for example a standard commercial type 10 or 47uF cap and not some specialised ultra low ESR component.
Ripple current can never be to high as it is a maximum the device the cap can tolerate, however you need to check it is specified in the same way when comparing.
Generally the bigger physically a cap is and the higher the ripple current allowable.
If the ESR is 'high' and the ripple current 'high' then the power dissipated in the cap is also 'high' and so it warms.
For your Class AB these concerns are so far off the radar as not to be an issue. If it were Class A and say pulling 10 amps continuously from the supply then you would need to look more closely at what you use. When your Class AB amp is idling the ripple current in the caps will be pretty small. Under full sine drive into say 8 ohm it will rise.
To ensure that high ESR never becomes a problem... because high ESR can limit the caps effectiveness at high frequency particularly in decoupling SMPS rails... you should also add in parallel a more consistent low ESR part such as a small film cap of say 0.22uF or a 'defined' low impedance such as a film cap in series with a low value non inductive resistor.
Rather than looking for 'good' and 'bad' ESR, try and think in terms of looking for parts that will maintain their spec over years and decades and design the system so that an increase in ESR of say 100 or 200% as things age will cause no issues... such as adding the bypass caps.
I have a simulation of a power supply and amp set up. This shows the current in one side of the power supply (the reservoir cap) and the cap is just 2200uF in size.
You can see the 1kHz audio flowing 'in' the cap and also the larger charging pulses. First image is at around 1 watt output and second image at nearer 50 watts.
Attachments
High ripple current is always good for any power supply. But low ESR type is good only for high-frrquency dc/dc (smps). Low ESR type in low-frequency circuits can hurt sometimes.
I like this link to a PSU calculator because it shows that good or best ESR or ripple ONLY makes sense from circuit requirements.
A cap is not good or best because some auto proclaimed expert says so, it is what it should be from calculations.
A cap is not good or best because some auto proclaimed expert says so, it is what it should be from calculations.
A high ripple cap is only as good as transformer feeding it.
If the secondary saturates and limits the current then wasn't much point in a high ripple cap ! High current puts strain on transformer and rectifiers.
If the secondary saturates and limits the current then wasn't much point in a high ripple cap ! High current puts strain on transformer and rectifiers.
Any transformer does NOT saturate from excessive secondary current flow - but due to primary Vsec - product. Which is at maximum without secondary load.
There is a limit in how much current you can take from a secondary before the voltage drops to zero.
You cant take out more than you put in and you cant put more in than primary can cope with.
You cant take out more than you put in and you cant put more in than primary can cope with.