Main Power Cap size vs Power?

[Keith1964]

I bought a 2009 Sony STR-DA5400es AV receiver, and i noticed in the service manual that the main power caps are only 8,600uf X2 even though it's rated at 120 watts 8ohm 20-20,000. My 2016 Onkyo TX-RZ810 AV receiver has main caps that are 15,000uf x2 , and the Onkyo is rated at 130 watts.8ohm 20-20,000. And my 2004 Pioneer Elite VSX-55txi AV receiver , which has the lowest power rating of 100 watts 8ohm 20-20,000 has main caps that are 22,000uf x2. So the receiver with the lowest power rating has the largest caps. The caps in the Sony STR-DA5400es seem very small for a $2000 receiver rated at 120 watts. So what is the correlation between main cap size and output power?

 

[slipperybidness]

I bought a 2009 Sony STR-DA5400es AV receiver, and i noticed in the service manual that the main power caps are only 8,600uf X2 even though it's rated at 120 watts 8ohm 20-20,000. My 2016 Onkyo TX-RZ810 AV receiver has main caps that are 15,000uf x2 , and the Onkyo is rated at 130 watts.8ohm 20-20,000. And my 2004 Pioneer Elite VSX-55txi AV receiver , which has the lowest power rating of 100 watts 8ohm 20-20,000 has main caps that are 22,000uf x2. So the receiver with the lowest power rating has the largest caps. The caps in the Sony STR-DA5400es seem very small for a $2000 receiver rated at 120 watts. So what is the correlation between main cap size and output power?

There is nothing set in stone on this, not in the slightest. It comes down to design decisions, available space inside the chassis, and the big one being cost.

The general rule of thumb for the DIY crowd is to use 10,000uF for every 100W of continous full bandwidth power. And, it should be noted that this is typically considered as "generous filter capacitance".

EDIT--I should add details, that this is for linear non-regulated power supplies and class A or A/B amps. Once we start talking about Class D amps and SMPS, it is a completely different discussion.

 

[Keith1964]

There is nothing set in stone on this, not in the slightest. It comes down to design decisions, available space inside the chassis, and the big one being cost.

The general rule of thumb for the DIY crowd is to use 10,000uF for every 100W of continous full bandwidth power. And, it should be noted that this is typically considered as "generous filter capacitance".

EDIT--I should add details, that this is for linear non-regulated power supplies and class A or A/B amps. Once we start talking about Class D amps and SMPS, it is a completely different discussion.

Do you think caps of 8,600 is pretty small for a receiver rated at 120, especially for a receiver the originally MSRP'd for $2000. Maybe that's why Sound and Vision only bench tested the Sony STR-DA5400es at 39 watts all channel's driven, even though it's 2 channel rating is 120.

 

[Irvrobinson]

There is nothing set in stone on this, not in the slightest. It comes down to design decisions, available space inside the chassis, and the big one being cost.

The general rule of thumb for the DIY crowd is to use 10,000uF for every 100W of continous full bandwidth power. And, it should be noted that this is typically considered as "generous filter capacitance".

EDIT--I should add details, that this is for linear non-regulated power supplies and class A or A/B amps. Once we start talking about Class D amps and SMPS, it is a completely different discussion.

Actually, Class D amplification has nothing to do with the power supply smoothing (sometimes called filtering) caps. The whole capacitor size question in this thread is all about linear power supplies.

The objective of the power supply in an audio amplifier is to convert the mains AC current (120v/60Hz) to DC current at a desired voltage level. The power transformer changes the voltage level, while a full-wave rectifier converts the alternating current to direct current. Rectifiers aren't perfect, so the output current still has residual AC "ripple" in it, and what the capacitors do is smooth the voltage ripple. The size of the capacitors are indicated by a formula which includes the voltage and the magnitude of the ripple. Power supply capacitors do not determine the level of power delivered to the amplification stages; that's all up to the transformer. There is no benefit to making the capacitors larger than necessary to smooth the current to DC, as proven in section 8.1 of this article:

Linear Power Supply Design

ESP - The Audio Pages. Linear power supply design information - Learn how to design your own high performance amplifier power supply.

sound-au.com

I think @PENG and I have posted it multiple times. Required capacitor size depends on what the desired DC voltage is and what the maximum current will be from the power transformer.

 

[Keith1964]

Actually, Class D amplification has nothing to do with the power supply smoothing (sometimes called filtering) caps. The whole capacitor size question in this thread is all about linear power supplies.

The objective of the power supply in an audio amplifier is to convert the mains AC current (120v/60Hz) to DC current at a desired voltage level. The power transformer changes the voltage level, while a full-wave rectifier converts the alternating current to direct current. Rectifiers aren't perfect, so the output current still has residual AC "ripple" in it, and what the capacitors do is smooth the voltage ripple. The size of the capacitors are indicated by a formula which includes the voltage and the magnitude of the ripple. Power supply capacitors do not determine the level of power delivered to the amplification stages; that's all up to the transformer. There is no benefit to making the capacitors larger than necessary to smooth the current to DC, as proven in section 8.1 of this article:

Linear Power Supply Design

ESP - The Audio Pages. Linear power supply design information - Learn how to design your own high performance amplifier power supply.

sound-au.com

I think @PENG and I have posted it multiple times. Required capacitor size depends on what the desired DC voltage is and what the maximum current will be from the power transformer.

Thanks for the explanation, i was just always under the assumption that the larger the capacitor size, the greater the storage space for greater headroom.

 

[slipperybidness]

Actually, Class D amplification has nothing to do with the power supply smoothing (sometimes called filtering) caps. The whole capacitor size question in this thread is all about linear power supplies.

The objective of the power supply in an audio amplifier is to convert the mains AC current (120v/60Hz) to DC current at a desired voltage level. The power transformer changes the voltage level, while a full-wave rectifier converts the alternating current to direct current. Rectifiers aren't perfect, so the output current still has residual AC "ripple" in it, and what the capacitors do is smooth the voltage ripple. The size of the capacitors are indicated by a formula which includes the voltage and the magnitude of the ripple. Power supply capacitors do not determine the level of power delivered to the amplification stages; that's all up to the transformer. There is no benefit to making the capacitors larger than necessary to smooth the current to DC, as proven in section 8.1 of this article:

Linear Power Supply Design

ESP - The Audio Pages. Linear power supply design information - Learn how to design your own high performance amplifier power supply.

sound-au.com

I think @PENG and I have posted it multiple times. Required capacitor size depends on what the desired DC voltage is and what the maximum current will be from the power transformer.

I agree with everything you posted here. I did want to be sure to clearly state that this conversation is only relevant for unregulated linear power supplies. (Usually when you see class D, it will be mated to SMPS, but that is not always the case.)

But, amp power supplies also do not require to have excellent AC ripple specs, if they did then a regulated power supply may make more sense than trying to regulate the voltage with big honkin caps. On the other hand, processors and ICs do need to have regulated power supplies (i.e. excellent AC ripple specs).

Again, I state that the general rule of thumb for the DIY hobbyist is 10,000uF for every 100W. This is taken from an amplifier construction manual by Bob Cordell. I would assume that he ran through some of the calculations that you referenced to get to that general number.

 

[slipperybidness]

Do you think caps of 8,600 is pretty small for a receiver rated at 120, especially for a receiver the originally MSRP'd for $2000. Maybe that's why Sound and Vision only bench tested the Sony STR-DA5400es at 39 watts all channel's driven, even though it's 2 channel rating is 120.

Personally, I think you are making mountains out of ant hills.

8,600uF would seem to be a bit on the low end for a $2000 piece of gear, but I certainly would not be concerned about that in the slightest. On the other hand, I tend to avoid Sony regardless.

 

[lovinthehd]

I have no idea what size caps are in any of my four avrs or seven power amps nor do I care.

 

[Keith1964]

I have no idea what size caps are in any of my four avrs or seven power amps nor do I care.

Thanks for your constructive input...

 

[lovinthehd]

Thanks for your constructive input...

Just sayin' it's not important in and of itself....

 

[Irvrobinson]

Thanks for the explanation, i was just always under the assumption that the larger the capacitor size, the greater the storage space for greater headroom.

For some reason that's what the boutique amplifier industry wants us believe. I think it's because huge transformers and capacitors as big as Coke cans look cool, and differentiate them from receivers and cheap amplifiers.

 

[Irvrobinson]

This is another great article by Rod Elliot, for the technically inclined:

amp_design.htm