Power supply and voltage dividing

[Steve Blackdog]

Hello all


I am building a power supply with a 35 - 0 - 35 vac -in from the transformer. When rectified (4 x 1N4007) that pushes the DC voltage up to around 49vdc.

I need V-out to be less than +/- 40 vdc.

If I drop a 33k resistor in the V-out path with a parallel 82k to ground on the (+ve) rail,(mirrored on the (-ve) rail), V-out comes down to +/- 35 vdc.

two questions:

1). I know that this is not the preferred way to drop Voltage, but I don’t understand why not.

2). Subject to the answer to (1), I was planning to put the divider between the rectifier and the smoothing caps (4700 uF 50v).

This is to power a TDA 7294 power amp I have built. If anyone have any better ideas, I am keen to learn.

Cheers

Steve

[Enzo]

Use Ohm's Law. Imagine your circuit draws 100ma (0.1A). It comes through that 33k resistor. How much voltage will drop across that resistor?

If all you wanted was to create a 40v point from a 49v supply, sure, but as soon as you want to draw any current from it, that goes out the window.

[loudthud]

Without getting into a long winded answer, get a lower Voltage transformer, something like 25-0-25. 1N4007 diodes are not appropriate for a solid state power amp, you need something bigger with at least a 3 AMP current rating.

Study this: https://en.wikipedia.org/wiki/Th%C3%A9venin%27s_theorem

[J M Fahey]

Short answer and agreeing with above:your amplifier needs 4 or 5 AMPERES current at +/-40V ; the resistor network will allow LESS THAN 1 MILLI-AMPERE through, 5000 times less than needed.

You can drop voltage for a preamplifier which needs a few mA, but NOT for a power amp, which needs a power transformer straight giving you the proper voltage.

[Steve Blackdog]

Short answer and agreeing with above:your amplifier needs 4 or 5 AMPERES current at +/-40V ; the resistor network will allow LESS THAN 1 MILLI-AMPERE through, 5000 times less than needed.

You can drop voltage for a preamplifier which needs a few mA, but NOT for a power amp, which needs a power transformer straight giving you the proper voltage.

Thanks, for this.

does this mean I need a 5A transformer?

[Vertigorocks]

The TDA7293 May be a Better Choice (Max +/- Rails is 60Vdc). A very low current Regulator could be made using a 40V Zener Diode. I would also just recommend using a Bridge Rectifier for the High Current Vcc and -Vcc...

[loudthud]

According to the TDA7294 data sheet, the chip's power output with an 8 Ohm load and with +/- 35V supply is 70W typical. At 4 Ohms and a +/- 27V supply it is also 70W typical. These Voltages need to be what is supplied to the chip at full load. When there is no signal, the Voltages will be higher. An important thing to know is that most transformers are also rated at full load, when there is less than full power being supplied to the chip, the Voltage will be higher. Because of this, in a practical sense, 70W is a little optimistic.

For 8 Ohms, try to find a 25-0-25 secondary at 2.8A or more. A 2A transformer will run on the warm side, A 3A transformer will just be slightly warm. A 4A bridge rectifier is a little light, going higher would be safer. 4700uF capacitors per rail.

For 4 Ohms, a 20-0-20 secondary at 4A is about right. That will run just barely warm all night long. An 8A bridge is a good choice, 10000uF caps per rail.

[J M Fahey]

Thanks, for this.

does this mean I need a 5A transformer?

No, it means you need a 28+28 VAC transformer 150VA or better.
25+25V VAC as suggested above is fine too, if you can get one easier/cheaper.
Minimum VA I would be comfortable with would be 120VA, since your amplifier is almost 100W RMS

This one is perfect at $22 each:

https://www.apexjr.com/miscellaneous.html

[Helmholtz]

Maximum theoretical efficiency of a PP class B powerstage is 78%.
Real world designs are typically below 75%.
So a 100W power stage consumes at least 133W of DC power full output.
Add a few watts for idle power, drivers and rest of circuit.
Rectification and PT losses may add another 15W to AC power.
So a 150W PT seems a minimum requirement.

One might argue, though, that the amp doesn't work at full power all the time..