low voltage high current power supply

[mr_byte31]

Hi All,

I am planning to build a 5v power supply which can provide at least 50A. I found many project on internet that use MOT (Microwave oven transformer)
it can produce a lot of current. I am focusing now on the losses and low ripple DC voltage.

I am planning to use Pi-Filter to remove the AC ripple voltage (Forget about values in figure, I just add it as reference):


I will configure the secondary windings to have 6.4 v AC to overcome the losses on the rectifier

I have two points here

losses:
-------
The bridge rectifies would reduce waste 2x0.7v = 1.4  v
The losses would be 1.4 v x 50 A = 70W
the remaining voltage would be (6.4-1.4=5 v), the power here would be = 5 v x 50 A = 250 W
we get a efficiency of approx = 1- 70/250 = 72% !! I think this is low
any idea how to improve this ?

AC ripples:
-----------
is it possible to have a very low AC ripples at this high current using Pi-filter? (forget about the C and L values in figure)

[FTL]

Don't forget, the transformer will not be 100% efficient either. Don't count on more than 95%, maybe as low as 85%?

You could consider MOSFET's with very low Rdson (i.e. 0.01 ohm or less), but you would need to actively control them. At .01 ohm, MOSFETs would have less than half the loss of the diodes. They would peak at a voltage drop of 0.5V, but would vary from 0.0 to 0.5 as the waveform changes.

[titi]

Hi mr_byte31,

using power Schottky diodes for rectifying can help to reduce the losses.
But for 50A it is not so easy to find.
Best regards.

[LithiumOverdosE]

How about using active rectification?

You can see one implementation here http://www.eeweb.com/blog/extreme_circuits/power-mosfet-bridge-rectifier.

[crunx]

One important aspect is dimensioning of the mains transformer: When you rectify the raw mains AC waveform to capacitor, the current is flowing in short pulses on highest part of the voltage amplitude. Therefore the transformer has to be dimensioned to be much larger than what a simple wattage calculation would predict. So you may well end up with >500 VA transformer. So be prepared to have a lot of iron and copper in your proposed configuration!

Also, the schematics you have drawn will not produce very precise nor stable voltage. The mains voltage tolerances are pretty wide, and the voltage will also vary much depending on the load.

Many years ago I built something similar, 5V 40A. However, I had also requirement of stability and accuracy within +/-200 mV. I had to have a voltage regulator on the output for that. And the transformer was enormous. So was heat dissipation, too. I remember the total efficiency was below 50%.

I would consider a careful analysis and design considering voltage stability and accuracy and also total weight and cost of materials before starting. In these days one would use a primary-side switched design, which then will have a small transformer (due to high frequency), and also pretty good efficiency. Making one from scratch requires quite much expertise, and parts would be not easy to shop to low cost and in small quantities. Therefore I would either buy one ready-made module, or use a salvaged power supply. I have myself found a nice powerful module providing more than you need from a old data-center-class server rack, for free! (Had to do the dumpster diving and have just a bit of luck)

[Checksum8]

Can you tell us what your load will be or what this is being used for?

Thanks

[motox]

How about using active rectification?

You can see one implementation here http://www.eeweb.com/blog/extreme_circuits/power-mosfet-bridge-rectifier.

I use this circuit in my homemade benchtop power supply. Input voltage is around 24V@3A and the losses are very small.
It’s very important to use 0.1% resistors or the mosfets will not turn on/turn off at the correct times, resulting in more losses. At first I used 1% resistors and the mosfets heated a lot, I found that they were short-circuiting the power rails during the transitions.
By the way, this circuit is original from Elektor.

[mr_byte31]

The 5v supply would be a part of the circuit that has several speakers connected in parallel (4 Ohm each)as load .  Thats why I should have almost no ripples !

I doubt if i will be able to have 0/1% tolerance resistors ! This make the idea of active bridge so hard to implement.

I am aware that the transformer would have losses but I neglect it in my efficiency calculations. I saw several projects on Youtube that used MOT to generate 200A AC !

[crunx]

What about going up with voltages and that way reducing currents and rectifying losses?  As you know, diode losses are nearly constant, so from higher voltage the percentage of losses is lower.

Also, to get high power out of a speaker, you have to get into it a high power, too. Assuming that the speaker's impedance is 4 ohms, and you want to get, say just 10 watts out of it, the voltage should be about 6,3 volts RMS / 8,8V peak. The current would be about 1.58 A. For 100 W to 4 ohms, you need 20V RMS and 5 A RMS. And Peak-to-peak voltage for 20 V RMS is 56 volts - for a single-ended push-pull thus the supply should in practice be >60V to get that much swing. Music, with irregular waveform, would be even then clipped and distorted. For differential mode (speaker placed between two identical, but opposite-phased outputs) half of the supply voltage (30V) would be just barely enough for the same result.

Therefore I wonder why you target so low supply voltage if the purpose is to feed speakers?

[vern]

Hi,
I wouldn't bother building a power supply myself.
I would buy one, like a MeanWell HRP-300-5, it's about $90, 5VDC and 60Amp. There might be other even cheaper ones.
If your application is very critical with noise you might have to add some filtering, but otherwise it's ready to go, no heat issues and no safety problems.

[mr_byte31]

What about going up with voltages and that way reducing currents and rectifying losses?  As you know, diode losses are nearly constant, so from higher voltage the percentage of losses is lower.

Also, to get high power out of a speaker, you have to get into it a high power, too. Assuming that the speaker's impedance is 4 ohms, and you want to get, say just 10 watts out of it, the voltage should be about 6,3 volts RMS / 8,8V peak. The current would be about 1.58 A. For 100 W to 4 ohms, you need 20V RMS and 5 A RMS. And Peak-to-peak voltage for 20 V RMS is 56 volts - for a single-ended push-pull thus the supply should in practice be >60V to get that much swing. Music, with irregular waveform, would be even then clipped and distorted. For differential mode (speaker placed between two identical, but opposite-phased outputs) half of the supply voltage (30V) would be just barely enough for the same result.

Therefore I wonder why you target so low supply voltage if the purpose is to feed speakers?

The speakers are 100W each and I have parallel speakers, almost 10 speakers on parallel !

MOT transformer and other stuff are scrap things , I didnt pay a cent for any of them. I am not in favour to buy $90 power supply that would be a part of a project !