Designing a reliable 5V/5A PSU for BBB

[Dhanushka Jayaweera]

I am using a 5A transformer to step down 230V to 12V using diodes. After the rectification I'm getting 17V DC. First I'm using LM1084 to step down 17V to 7.5V.
Then I'm giving LM1084's output (7.5V) to 3* UCC283-5. (three UCC283-5 parallel)

Is it bad to use a linear regulator immediately after a linear regulator? (LM1084 -> UCC283-5)

I have noted that one of UCC283-5 gets heated up than the other 2 when all 3 have the same load.
I understand that this kind of a design is an overkill for safety but I'm doing it anyway because I need to ensure my load is safe at all circumstances.

I'm correctly getting 5V from all three UCC283-5. I just want to figure out why one UCC283-5 is heating up than the others.

 

[BradtheRad]

(three UCC283-5 parallel)

It is common to add a transistor (known as a 'pass transistor') when you want to assist a regulating IC, rather than to use multiple regulating IC's.

Is it bad to use a linear regulator immediately after a linear regulator? (LM1084 -> UCC283-5)

Though it is not forbidden to split the voltage drop among two devices...

There is also a line of thought which says 'The more complicated you make something, the more there is to go wrong.'

For this reason there are many pilots who prefer to fly a single-engine plane, even though we would think two engines are safer.

I have noted that one of UCC283-5 gets heated up than the other 2 when all 3 have the same load.
circumstances.

I'm correctly getting 5V from all three UCC283-5. I just want to figure out why one UCC283-5 is heating up than the others.

One of them is probably putting out a tiny bit higher voltage than the others. Consequently it is providing the major part of the current, causing it to heat up.

If you decide to keep three UCC283's in parallel, then you will need to balance the loads equally.

 

[Dhanushka Jayaweera]

Thanks a lot for ur reply.

Is it bad to use a linear regulator immediately after a linear regulator? (LM1084 -> UCC283-5)
Though it is not forbidden to split the voltage drop among two devices...

There is also a line of thought which says 'The more complicated you make something, the more there is to go wrong.'

The issue I have with LM1084 is this. It gives an output(Vout) based on the resisters. (check attached image)
And if a resister is disconnected somehow, then Vout = Vin

Therefore, there is a risk of getting 17.5V as LM1084's Vout. That's why I'm using a LM1084 first, and then using UCC283 to get fixed 5V.

 

[kam1787]

might have been simpler to have one [ic] regulator, and then given your concern, have an over-voltage circuit after it [maybe it would blow a fuse or shut-off a series-pass transistor]

as Dhamushka points out now you likely have a problem with load-sharing