Question to any EE's around here:

[digital desire]

If you were designing a conventional (not class A) power amp, and had a 1000VA toroid for a power supply, what would be a reasonable guess for it's max rated output rms? If it makes a difference, assume a 4 ohm load.
Thanks!

[JonMarsh]

Sizing of power amplifier transformers relationship to output power is dependent on whether it's a capacitor input rectifier or choke input rectifier. Choke input rectifier circuits have much higher transformer utilization, and you can get about 90-95% of the transformer VA to the DC supply. That said, a class AB amplifier is not very efficient...

In the case of conventional rectifier capacitor input supplies, the current is pulled out of the transformer only at the very peak of the AC line, so transformer utilization is poor- about 50-55%. For a 1000VA transformer, you have the makings for roughly 200W/ch amplifier IF you assume it must be capable of FCC preheating and continuous power operation. It's always possible to run a 60 Hz transformer in some overload for a while, as long as it has been designed not to saturate the core, and until it overheats, which can take some minutes or more, you'll be OK, though your rails will usually be sagging. Rating a transformer VA is a bit tricky anyway, and different manufacturer's will have different derating for regulation, core heating, winding heating, etc. If a transformer seems unusually small for the rated VA, it's probably not really in the same ballpark performance wise as larger transformers. Iron and copper rule. Last, the power supply rails should be set appropriately for the intended load impedance and power levels you want, considering also the output stage voltage drop headroom needed for saturation losses and ripple on the capacitors- you can reduce the allowance for ripple droop by using larger filter capacitors, but besides the cost, you have to come up with inrush surge control circuits.

[digital desire]

Interesting, thanks. I found a 3 channel amp with a 850VA toroid, but rated 300 watts into 4 ohms/channel, all channels driven, and I wonder if somehow it is possible. Seems a little stretched.

[Carrotman]

Note that music signals have a significantly lower energy content than pure sine waves (cf. the IEC268-5 test signal).

If you are only playing test tones your PSU output will have be (conceptually):
P(PSU)= P(out)/(n(PSU)*n(Amp)) (Respectively PSU output power, amp output power, PSU efficiency and amp efficiency).

However, if you use the amp for normal music signals (or only play test tones for a short time ), you can get by with between 1/8th (normal consumer use) and 1/3 (demanding PA applications) of this requirement, which is what any commercial manufacturer will use to design their products. Now, I am not saying that there aren't sonic benefits to using larger supplies, but from an engineering point of view it is not required and therefore this can also be used to save the average DIY'er quite a bit of cash when purchasing PSU components for amp projects 8)


/U.

[Mike B]

My A500's are rated 500 watt output and the input, 120 VAC @ 6.5 A yields 780 VA for 500 watt.

Sounds about right, ~64% efficiency.