Design for Power Supply with 24-0-24V 2A transformer

[alderick84]

I'm modifying an amp design to change it's power supply, the original is meant for a 0-48V 5A transformer, and I have a 24-0-24v 2A one.

I'd like to use an integrated bridge instead a four diodes, what capacity does the bridge need 2A, 3A, 6A?

How big do the capacitors need to be, would 4700uf be enough?

How wide do the tracks need to be for the ones between the transformer and the bridge, the bridge and the capacitors, and the capacitors and rest of the amp?

Thanks a lot for your answers.

 

[LLPe]

The former tranformer has a rating of 240VA and the latter 96VA : the new transformer *is far too small.* It will fail and if not protected it could eventually catch fire !

Bridge: minimum 10A

Capacitor-size: 4700u/50V maybe, it depends on the load.
My advice would be 3 X 3300u. If you look at the prices for eletrolytics, the price usually rises sharply as capacity goes up. It might even be cheaper to go for 4 X 2200u, *and* the combined ripple current capability goes up => increased life time. In the end you'll get something for less.

A bit of math to help choose capacitors:
Q = CU and Q = It, where C is capacity, U is the ripple voltage, I is the current and t is the time between capacitors get (re)charged.
Rearrange and you have
C = It/U
The current is given by the load, t is between 8 and 10ms (50Hz) and U is ripple your application can tolerate. Ex I = 2A, t = 8ms and U = 2V => C = 8mF or 8000uF.

Lars

 

[alderick84]

Thanks a lot for your reply, you're completely right about the original transformer, it wasn't 0-48V 5A but 0-24V 5A, which brings the latter transformer closer to the original in terms of VAs.

If I wanted to use four capacitor instead of two, the design in the image would be right?

When I turn that schematic into a PCB, what track width should I use?

 

[DF96]

You have shorted the negative output from the bridge.

Your aim is unclear, so how can we help you when we don't know what you are doing and why. Why are you modifying an existing design? What is wrong with it?

 

[alderick84]

Thanks for your reply, as I said, the design is meant for a 0-24V 5A transformer but I don't have one and I don't want to build it myself, what I have is a 24-0-24v 2A one, that's why I need to change that design.

The original design uses two 6A diodes and 2 4700uf 35v capacitors. I can post the design if it's relevant for my question.

P.D You're right about my mistake by shorting the negative bridge's output, it's now fixed.

Thanks a lot.

 

[Osvaldo de Banfield]

Also, a fuse in the common line is almost useless. Use one fuse in each (Both) positive and negative rails instead, and another in the transformer's primary.

Take into account to use slow blown fuses, otherwise they will fail each time powering the power supply.

Good luck.

 

[alderick84]

Also, a fuse in the common line is almost useless. Use one fuse in each (Both) positive and negative rails instead, and another in the transformer's primary.

Take into account to use slow blown fuses, otherwise they will fail each time powering the power supply.

Good luck.

Thanks for your reply, I'll remove the fuse for the center tap and I'll add fuses for VCC+ and VCC-, and a fuse between the transformer and 115 AC.

¿What amperage should I use on every case?
¿Are different fuses for AC/DC?

Thanks a lot again!!

 

[Osvaldo de Banfield]

It depends on the load demand. I use about 1.5 to twice the normal peak current expected in the circuit. Once again, slow blown. The initial peak while charging empty caps is monstrous.

 

[AndrewT]

It depends on the load demand. I use about 1.5 to twice the normal peak current expected in the circuit. Once again, slow blown. The initial peak while charging empty caps is monstrous.

Low voltage secondary fuse/s before the main smoothing capacitors have to be so monstrous to survive first charge from cold, that they will never blow when you abuse the amplifier.
The close rated mains fuse will always blow first.

 

[Osvaldo de Banfield]

Low voltage secondary fuse/s before the main smoothing capacitors have to be so monstrous to survive first charge from cold, that they will never blow when you abuse the amplifier.
The close rated mains fuse will always blow first.

Its true, except that the guy use a soft starter (a NTC for example) in the primary of the "trafito". We don't know if he is planing to use it or not.

 

[AndrewT]

We don't know if he is planing to use it or not

It is a false Safety to omit the mains fuse and use instead secondary fusing.

It is far more important to fit a mains fuse rather than secondary fusing.
In my opinion there is no option to omit mains fusing.

 

[alderick84]

Thanks a lot for all your answers, my last concern is about fuses's capacity, I'll use three of them, one for mains and two before AC enters the rectifier bridge, as you can see on the picture.

An externally hosted image should be here but it was not working when we last tested it.

The amp I'm building uses the STK4162II IC which is around 70 watts.

You've already told me that fuses' capacity would depend on the load, how can I calculate the capacity for each fuse.

Thanks a lot again!!

 

[dearchap]

You say the IC presents a load of 70 W to the power supply. That would indicate a load current max of around 3 A. For that load the peak current through diode will be 29 A !!!!. I think you are better of increasing the transformer rating to 30-0-30. That will reduce the peak current considerably.

 

[AndrewT]

Your mains transformer is rated for 96VA at the output.
Taking account of efficiency, that means the primary must be rated at >100VA
With a 115Vac supply powering a 100VA transformer the normal maximum current will be 100/115 = 870mAac
You will need a T1A mains fuse.
However this may blow on start up.
You have to decide whether to increase that fuse to allow repeated cold starts, or to fit a soft start circuit.
The added resistance would need to be around 60r. This could be a high resistance Power NTC, or a bank of resistors (three 20r 5W in series).

I would omit the secondary fuses when fitting a T1A mains fuse.
If you use T2A, or higher, to allow cold starting, then it may be wise to look at some form of protection on the secondaries.

 

[PRR]

... increasing the transformer rating to 30-0-30. That will reduce the peak current considerably.

I do not understand this.

 

[stratus46]

Your drawing of the bridge bothers me as it isn't a bridge. It's just a circle of diodes. If you try a 'discrete' bridge wired that way you'll just blow fuses.

G²

 

[dearchap]

I do not understand this.

Do you agree with the current numbers ?

 

[AndrewT]

You say the IC presents a load of 70 W to the power supply. That would indicate a load current max of around 3 A. For that load the peak current through diode will be 29 A !!!!. I think you are better of increasing the transformer rating to 30-0-30. That will reduce the peak current considerably.

Do you agree with the current numbers ?

No.

 

[dearchap]

AndrewT can you point me the error in the analysis ? You agree with the load current max ?

 

[PRR]

> error in the analysis ?

I don't see your analysis?

You assert that a higher voltage will reduce peak current. This is only true if the Power is forced to stay constant. That is, we adjust the load resistance to suit the new higher voltage.

But this case is, I *assume*, an ordinary Audio Power Amp. The amplifier will draw a current which is a function of the speaker load. True, he could change from an 8 Ohm to a 16 Ohm. However that's an unusual way to control rectifier stresses.

I also see no harm in ~~29 Amp peak currents. For a 2A transformer I would use a >2A rectifier. Looking up specs on a low-price ($1) 2A bridge, I see single peak rating 175A and 10-cycle rating 40A. After 10 cycles (with 20,000uFd) the peak currents are down below 10A, a 4:1 safety margin. I would be OK with that. In hand-work, where part-cost is small, I might spring for a 6A ($2) rectifier just to be real sure.