power supply Ripple frequency and voltage

Thread Starter

harrison2015

Joined Apr 22, 2015
80
How do you find out the ripple frequency and ripple voltage of a linear power supply?
Do you add all the filter capacitors and the power supply load resistor
Time Constant R X C = All Filters capacitors total X load resistor = the power supply ripple frequency?

I'm looking for a very simple formula to get the ripple frequency and ripple voltage for a linear power supply but i can't find any
 

dl324

Joined Mar 30, 2015
16,846
Hi harrison2015,

Ripple frequency is determined by the rectifier circuit; 60Hz for half wave and 120Hz for full wave.

Using the formula for calculating filter cap value based on max allowed ripple:
\(Ripple = \frac{I_L}{C} * 6 * 10^{-3}\)
C in uF, Ripple V peak-peak, I amps; this holds for I of 1A or less.
 
Last edited:

dl324

Joined Mar 30, 2015
16,846
So how do you calculate the ripples Peak to Peak voltage will be?
Substitute the max current and filter cap size in the equation and it will give you ripple. It doesn't hold for max current over 1A, but it'll give you something. If you already have the supply, you can measure ripple under full load.
What is this max allowed ripple?
It's something you spec when designing a power supply. Normally you use the formula to calculate the size of the filter cap.
 

ian field

Joined Oct 27, 2012
6,536
So how do you calculate the ripples Peak to Peak voltage will be?

What is this max allowed ripple?
Mains frequency filters usually use electrolytics - and electrolytics usually have pretty wide tolerance. So if you arrive at a figure by calculation, you need to work out the worst possible case taking into account the various tolerance limits.
 

Thread Starter

harrison2015

Joined Apr 22, 2015
80
It's something you spec when designing a power supply. Normally you use the formula to calculate the size of the filter cap.
If i was designing a basic linear power supply , what formula would i use to get the size of the filter caps?

Square root of 2 x C x Eripple (P-P) x Frequency = current in amps.
But how did you get the ripple Peak to peak voltage? what formula did you use

If you already have the supply, you can measure ripple under full load.
Yes true, but once you measure the ripples peak to peak voltage of a linear power supply, what formula do you use to verify that it should be normally be and the max and min ripple voltage "range" of the linear power supply should be without going out of tolerance?
 

#12

Joined Nov 30, 2010
18,224
If i was designing a basic linear power supply , what formula would i use to get the size of the filter caps?

But how did you get the ripple Peak to peak voltage? what formula did you use

Yes true, but once you measure the ripples peak to peak voltage of a linear power supply, what formula do you use to verify that it should be normally be and the max and min ripple voltage "range" of the linear power supply should be without going out of tolerance?
1) By using the formula I gave you.
2) The same formula.
3) The same formula using the high and low limits of the capacitors you buy.
 

Thread Starter

harrison2015

Joined Apr 22, 2015
80
Square root of 2 x C x Eripple (P-P) x Frequency = current in amps.
I'm talking about calculating the ripple min and max voltage range before I build a linear power supply and take measurements

How would I know the ripples min and max voltages if i don't have a formula to get it?
And how do you know which capacitor values to use?
 

Thread Starter

harrison2015

Joined Apr 22, 2015
80
where do you get the 1.414?

When you look at schematics of linear power supplies, what formula do you use to get the ripple Peak to peak?

The fuse rating is 1 amp , do you just use the fuse ratings current as the I in the formula? the schematic doesn't tell you the current draw from the circuit so how would you know what is I current ?
 

#12

Joined Nov 30, 2010
18,224
1.414 is the square root of 2, to 3 decimal places.
I use the formula I gave you to calculate the P-P ripple voltage.
When I design a power supply, I already know what current I am designing it for. That is where I get the current.

You obviously don't know how to do math, but there is no other way to calculate the capacitance needed to achieve the ripple voltage that YOU decided to allow when the current that YOU decided to design for is flowing. I can not fix that for you.
 

Thread Starter

harrison2015

Joined Apr 22, 2015
80
When I design a power supply, I already know what current I am designing it for. That is where I get the current.
Ok I understand this

But when looking at various schematics on google of linear power supply schematics how can you calculate the ripple voltage if you don't know the current? do you just use the fuse current rating?

Vripple P-P = I/1.414 x C x F
1 amps divided by the square root of 1 amp?
 

michael8

Joined Jan 11, 2015
410
But when looking at various schematics on google of linear power supply schematics how can you calculate the ripple voltage if you don't know the current?

I couldn't. I'd look at the output regulator maximum current and use that or some other hint from the circuit.

The following assumes a capacitance input (only) filter.

The AC power flow isn't continuous (single phase, either half or full wave rectified). In the USA it's 60 Hz (50 Hz is usual in Europe). For a full-wave rectifier there are 2 pulses of power each cycle. So for 60 Hz power and full wave:

1 cycle = 1/60 -> 16.6 mS (milli-seconds) in one cycle,
full wave has two pulses so each is separated by 1/2 of 16.6 mS or 8.3 mS

Between input pulses the only source of power is that stored on the capacitor which discharges supplying the load during this time. How much does it discharge?

voltage change on capacitor = (current amps * time seconds) / capacitance farads

The ripple is the amount the voltage on the capacitor changes. So once you have the input pulse
timing, load current and filter capacitance you can calculate the ripple (peak to peak).

And you can run this the other way and calculate the capacitance needed for a specific ripple too...
 
Top