Simple power supply for DIY building breadboarding prototyping

[rmunson]

Hey Everyone -
One project that I'm working on is a simple power supply to help me understand how AC power magically gets sliced in half and smoothed out like butter. Here's where my question for the group comes in --

In some simple PS designs I see a cap coming off the diode rectifier -- sometimes its somewhat low - say 220uF other times it's quite large - say 2200uF -- is bigger always better with the filter cap? Could someone describe to how to determine how much filter cap one really needs? Sometimes I kind of feel like people are just going with the idea that bigger is automatically better --

Thanks!
Reed

 

[ConradH]

The commercial rule is something like 1000 uF per amp of output, but it can vary wildly depending on how much ripple you're willing to live with. There are power supply calculators online that will give you specifics without much work. If you go too big the rectifier gets stressed on powerup and you start to need soft start circuits. There's also the issue of what happens when the output is shorted out. Best to fuse wisely. Then there's transformer selection. Study one aspect at a time as even "simple" power supply design isn't that simple.

 

[rmunson]

Thanks Conrad -
for smooth DC, what should my goal be for peak to peak output?

 

[WBCN]

From Regulated Power Supplies by Irving M. Gottlieb:

...the output filter capacitor of the linear regulator provides few problems and often is not critical in value. This comes about because very little filtering action, per se, is needed--most is performed electronically by the series pass (or shunt) transistor.​

 

[rmunson]

for example - I'm testing a 5v phone charger power adapter -
I'm seeing approx 5 mV pk to pk

on a little regulated power supply circuit I found out there I'm seeing about the same pk to pk


does pk to pk really matter that much with audio circuits?
how smooth pk to pk is acceptable?

 

[whoaru99]

What is acceptable depends on the functional requirements/goals of the circuit. Not sure there is any one right number in abstract.

 

[EngineerNate]

From Regulated Power Supplies by Irving M. Gottlieb:

...the output filter capacitor of the linear regulator provides few problems and often is not critical in value. This comes about because very little filtering action, per se, is needed--most is performed electronically by the series pass (or shunt) transistor.​

I think OP is talking about the cap on the input side of the regulator circuit/IC that smooths the DC from the rectifier before it gets regulates.

The output cap on the regulator, at least in the case of monolithic regulator ICs, is primarily for tuning the stability of the regulator. Some don't need an output bypass cap at all, others need specific values and ESR ratings.

 

[BinaryMike]

does pk to pk really matter that much with audio circuits?
how smooth pk to pk is acceptable?

Audio circuits vary tremendously in the amount of power supply noise that is acceptable. Some of the simpler circuits pass most of the power supply noise to the output, while others reject almost all of the noise. Engineers call this parameter Power Supply Rejection Ratio. It's one of the key specifications for op-amp chips, for instance, where the figure can be well in excess of 60dB (1000:1).

 

[EngineerNate]

Audio circuits vary tremendously in the amount of power supply noise that is acceptable. Some of the simpler circuits pass most of the power supply noise to the output, while others reject almost all of the noise. Engineers call this parameter Power Supply Rejection Ratio. It's one of the key specifications for op-amp chips, for instance, where the figure can be well in excess of 60dB (1000:1).

The TL780-12 I'm retrofitting in to the power supply of a turntable has a PSSR rating of 85dB typ, 70dB minimum. It's amazing what more modern ICs can do!