48V power supply (50W)

Hi,
I need a 48V power supply with 50 watt power(50W).the main voltage is 220V (rms).
can you help me with schematic?

the ripple is not important, larger than 10 volt with 50 hz frequency.

220V into a full-wave bridge rectifer, followed by a switch-mode power supply circuit to buck the 220 Vdc down to 48 Vdc.
or
220V to 76V custom-built transformer, followed by a full-wave rectifer and cap to reduce ripple.



(avg Vout) 48V / 0.636 = 75.5 Vdc (peak voltage out of rectified sine source)
220 Vrms * sqrt(2) = 311 Vpeak (AC)
311V / 75.5V = N = 4.12 turns ratio
Nice transformer/rectifier tutorial


The biggest problem is going to be finding/making a transformer that will handle 50W (plus losses). You might explore some transformers out of things like TV's or computer power supplies, to see if you can find something in your power/voltage range. If you can convert to a slightly higher voltage, you could use a linear regulator on the output to pull it down to a clean 48V (with a decent heatsink).

thanks enjunear
your link is good.
but
The cost of power supply is important for me.
50 watt transformer is expencive.
I think the very easy flyback is better. the feedback is not important because the ripple is not important for me.
but a short current protection is need.
I think a flyback with self oscillation is very cheap.
So
Is there any schematic for help me?
and the max out put voltage is 48 volt,so the rms is lower.

another question enjunear
when I need 48 volt with more than 48watt power (for example 150 watt) the better choice is computer power supply.
Is there any way to convert 12 or 5 volt from computer smps to 48 Volt???

The problem with directly using a flyback is that you are starting from AC and going to DC. If you simply rectify the 220V AC to DC, your SMPS will still need a large coil to efficiently buck it from 198 Vdc down to 48 Vdc, at 50 W of output. For your power level, your are going to have either an expensive, but efficient design (with an inductor or transformer of some flavor), or you will waste a LOT of heat in a low-cost, inefficient device like a simple linear regulator to take 198V down to 48V.

For schematics, hit google or search the foums here... there are LOTS of examples of all of these circuits. You just need to pick your poison... nothing is free.

enjunear said:

220V to 76V custom-built transformer, followed by a full-wave rectifer and cap to reduce ripple.

Click to expand...

Assuming that the transformer in your diagram is the custom-built one referred to ....

The d.c. output of the circuit in that diagram will be 1.4 x Vac, i.e. 106V

The d.c. polarity notations on the primary do nothing to help the learner/beginner.

Syncopator said:

Assuming that the transformer in your diagram is the custom-built one referred to ....

The d.c. output of the circuit in that diagram will be 1.4 x Vac, i.e. 106V

The d.c. polarity notations on the primary do nothing to help the learner/beginner.

Click to expand...

The image I linked was a simple, readily-available diagram to get the OP thinking about some options. We don't need to supply each questioner with an EXACT solution, but at least give them more information than they had before, and get them thinking about all of the available options. What I stated in my reply was some fundamental math to help the OP start thinking about some implementations.

What are your recommendations for GelecG's problem?

We don't need to supply each questioner with an EXACT solution, but at least

Click to expand...

But at least give them accurate information.

Your 220V to 76V transformer doesn't meet that criteria.

Neither does "(avg Vout) 48V / 0.636 = 75.5 Vdc (peak voltage out of rectified sine source)"

The peak of 48Vr.m.s. is 67.8 (less diode drops), and "peak voltage out" isn't "avg Vout".

What I stated in my reply was some fundamental math ...

Click to expand...

... which had a fundamental flaw.

What are your recommendations for GelecG's problem?

Click to expand...

If ripple really isn't a problem, then a 50VA transformer with a 48V secondary, or two 25VA transformers with 24V secondaries connected in series, a fullwave bridge rectifier and no smoothing.

Applying smoothing will give 67V out. Presumably much too high for the application.


Or he could use the circuit which you posted using a 50VA transformer with a 36V secondary, or two 25VA transformers with 18V secondaries connected in series. That would give about 50V out.

If that is too high, he could remove turns from the secondary winding. Not too difficult a task. If my memory serves me correctly, a mains transformer of this size will have something like 6 or 7 turns per volt.


The same reasoning applies to a 150W source. The so called "flyback" technique is more complicated than is necessary.