Calculating Capacitors for Voltage Regulator

Hi!

I'm using a 7805 5v regulator to power a board and motor driver, here's the datasheet for it.

I'm going from 9v to 5v, how do I calculate the capacitors to use on each side?

Look at the data sheet. No "calculation" needed.

Each manufacturer has their own data sheet, make sure the part you are using and the data sheet match up in this respect as different parts have different requirements.

There is no way you can calculate the value of the capacitor because you do not have enough information about the internal circuit of the chip.

Ok, thanks, this explains a lot!

Usually 10uF on the input and 1uF on the output works pretty well.

The capacitor in the unregulated side of the power supply is very important, there must be low enough ripple so that it never drops below about 8V.

If the 7805 is very close to the power supply, inches rather than feet, no additional capacitors should be needed on the input side. If it is powered from a wall wart, for instance, you should include something like a 4.7 to 10uF capacitor on the Vin side. The datasheets used to recommend tantalum, but that was because low capacitance electrolytics used to have rather high ESR as they were only meant for audio coupling. Modern aluminum electrolytics have pretty low ESR even in low values.

But the output side should have a 0.1uF ceramic capacitor to ground. Larger capacitors there are not recommended.

The datasheet you linked to shows 0.33uF on the input, 0.1uF on the output side.

Thanks, guess I'm just having trouble actually reading the datasheet, considering the sheer amount of tables in it

In particular low-dropout regulators can be very fussy about things like max and min ESR for the output capacitor - check the datasheet.

First of all the input voltage needs to be noise free, I assume it is. The caps are used on the 7805 to prevent oscillations. If your input supply is clean I always use .1 uf caps and have never had a problem. Use a .1uf on both sides.

I always use .1 uf caps and have never had a problem.

I once had to scrap over a thousand set top boxes because someone changed the voltage regulator to an other manufacturer's part ( the same part number ) and 0.33uF was not sufficient and it oscillated. This oscillation caused a tone further down the line to be not recognised. So a 0.1uF is not always the answer.

Some ex-employee?

Did s/he change to another manufacturer to save money? Sounds like it didn't work.

[quote author=Nick Gammon link=msg=2209472 date=1430295454] Did s/he change to another manufacturer to save money? Sounds like it didn't work. [/quote] No it was a supply issue, they couldn't get the specified manufacturer so they got a replacement.

Grumpy_Mike: No it was a supply issue, they couldn't get the specified manufacturer so they got a replacement.

Yes, I've worked with purchasing departments like that; "It has the same part number so we didn't think the technical department needed to test it."

Russell.

Or there sales rep assured it it was an identical compatible part.

Or the board layout didn't follow good design rules, and it was a miracle that the original part worked in the first place.

aarg: Or the board layout didn't follow good design rules, and it was a miracle that the original part worked in the first place.

No it was a solid design, that company knew what it was doing when it came to design. They were professionals and this was "only" a regulator. The board had to route transport stream around and that was going at over 200MHz so there is no way a regulator could be badly routed. And anyway I supervised the layout. ;)

In the end they made half a million and the box had the lowest returns rate of any product they had ever made.

I have to admit, this confuses me a little. There's at least three conflicting tutorials for Arduino on a breadboard that list different values of capacitor (10uF, 47uF, 100uF) without reference to the datasheet of the regulator they use.

For example, the tutorial here at this site (http://www.arduino.cc/en/Main/Standalone) looks to be using an ST L7805CV and 2x 10uF. The datasheet (http://www.allcomponents.ru/pdf/stmicroelectronics/7805cv.pdf) shows 0.33uF on the input, and 0.1uF on the output.

Newbs like me are likely to follow tutorials like the above. I had actually looked at the datasheet and ultimately decided to go with what the tutorial said to use, believing it to be tried, tested, and safe. How do I pick the wheat from the chaff when differences like this appear?

How do I pick the wheat from the chaff when differences like this appear?

By learning a bit more. In general with this sort of regulator you can not have too much capacitance so any of those tutorial values will do. That is 10uF, 47uF, 100uF will all be fine. The largest lower limit on capacitors I have come across is 3.3uF with many at 1uF or even 0.1uF.

The trick in electronics is knowing when something is critical and when it is not. Most of the time you can recognise chaff when it has the word "instructables" in the URL.