Its rated for 25v and I'm not sure how long it was on the shelf. My transformer is 120VAC to 24VAC so was I wrong to assume the capacitor would never see voltages higher than 24VDC? New to electronics

A 24VAC sinisoid peaks at 34V, which is what your cap will charge to.

You want some pretty good headroom on your working voltage rating, so I would be real hesitant to use anything less than a 50V rated cap and would much prefer 63V or 80V. Even at Digi-Key prices, you'd be looking at $4 for an 80V cap versus $2 for a 25V cap. Just the aggravation you've already gone through is worth that.

 

This is just a supply I use for breadboarding. 5VDC and 12VDC outputs, so I'm assuming that a 10% ripple would be ok? As far as current goes the max I'll ever pull is about 1A MAYBE 1.5A

10% ripple is probably fine as long as your lows do not drop below the minimum voltage that your regulators need.

Let's say that you get an 18VAC output transformer rated at 2A and use that as the max load to give us some margin even over your 1.5A "maybe" max.

With a 12V regulator, let's say that we want to maintain 15V to give it good overhead. So that means we can let the input filter cap drop 3V from 18V, which is 17%. So the 10% should be a reasonable choice. That let's the voltage drop 1.8V each half cycle of the input waveform, or (1/60)/2 = 8.3ms. At a current draw of 2A, that's a total charge flow out of the cap of about

Q = A*T = 2A * 8.3ms = 16.6mC

You need that to correspond to no more than 1.8V on the capacitor, so

C = ΔQ/ΔV = 16.6mC/1.8V = 9259μF, or 10,000μF

Even if we use a 3V allowable drop and only a 1A max current, that is only going to get you down into the 3000μF range. So I think your cap size is pretty reasonable (if anything, a bit small).

 

Get a 1000uF 50v capacitor. Make sure it's a low-ESR type and is rated for a high temperature (105* C) as power supplies tend to heat up fairly quickly. Can't really go wrong with these values, provided you don't connect it backwards

Based on the numbers I just ran (and which may have an error), it would seem that 1000μF would only suffice for reasonably small current draws.

So check my reasoning here.

We are primarily concerned with the voltage sag in the cap over the course of half a period (assuming a full-wave rectifier). Ignoring the exact shape of the waveform and just using the half-period, then this means 8.33ms (for 60Hz systems). During that time, a 1000uF capacitor would sag

Q = CV
ΔV = ΔQ/C = I*T/C

ΔV/I = T/C = 8.33ms/1000uF = 8.33V/A

That seems pretty significant for anything over a couple hundred milliamps.

 

Post deleted.

 

Post deleted.

Hey Alec, you know you can actually delete the post. I believe you just go to "edit" and at the bottom there's a button for deleting the entire post.

 

Oh, thanks for that. I'd done my usual trick of only getting to the bottom of page 1 of the thread before posting, then finding my post was redundant .

 

Oh, thanks for that. I'd done my usual trick of only getting to the bottom of page 1 of the thread before posting, then finding my post was redundant .

I hear you there. I've done it more times than I care to remember!

 

So I'm using my homemade power supply to do some bread boarding and out of no where BAM! my 6800uf capacitor blows up for no reason. Luckily there was no fire or anything but any ideas on what causes this? Before I replace it I want to make sure it wasn't something I did.

PS the only load at the time was a single led so I doubt that was the cause

Thanks guys

The most LIKELY reason for capacitors exploding is having the polarity reversed. Tantalum capacitors are notorious for becoming blasting caps when you plug them in backwards.
Eric the Experienced

 

The most LIKELY reason for capacitors exploding is having the polarity reversed. Tantalum capacitors are notorious for becoming blasting caps when you plug them in backwards.
Eric the Experienced

And when they don't go BOOM! they often produce large pretty purple flames.
Bill the Burned