why such large transformers?

[Doovieman]

Hi-
I'm trying to pick a transformer for my next GC...I was wondering why people are using 500VA transformers etc... Seems to me that 120VA-160VA should be plenty for LM3875. Any thoughts/explanations? Thanks,

Doovieman

[Franz G]

My two cents:

We are not industrial producers, we don't split every cent or we don't buy thousands of trafos.

In fact, an amplifier is DC-supplier wich you can modulate with your audio signal. So you need the DC current.

Choose an appropriate size of the trafo. And still think about regulating the PSU ...

Franz

[peranders]

This have to with feelings. I think too 500 VA is over the top for one or two LM3875!

100-225 (300) VA is reasonable, my point of view.

[Doovieman]

Cool...I thought I might be missing something here...I'm planning on using a 120VA that yields +-30V and was unclear if there was an advantage to using huge transformers. Thanks for the info,

Doovieman

[peranders]

The obvious disadvantages of huge transformers is price (if you buy them), weight, size and inrush currents. Otherwise it's good

[5th element]

Traffos when stressed will lower their output voltage, this is what the regulation factor is. If it says 5% then the output voltage will fall by 5% under max rated output power. This droop can effect the sound quality.

A larger transformer will reduce this effect two fold.

1) it will not be as stressed so therefore will not droop as much
2) larger transformers have better regulation not dropping as much under stress.

So if a DIY person can get a larger traffo for not much more money over a smaller one then they will go with the bigger one. Also as DIYer you never know when youll want to change something, say add on another two channels. If you have the bigger traffo in the first place youve always got some future proofing.

But from a technical point of view you dont need a traffo any bigger then is required for max current drawn.

It also depends on the application

My 4 channel 80watts amp has a 650VA traffo, now for max output from each channel into 4 ohm loads this is not enough, but im never going to drive each channel, full power into 4ohms plus each amp only receives a portion of the audio spectrum because its an active setup.

So if you are going to be driving the amps to max output into tough loads then a bigger traffo may come in handy. If your not a smaller one will probably be fine. Its similar with heatsinks, my sinks on the aforementioned amp are suitable for max output into 4ohms but in the application I use them in they never get hotter then slightly warm to the touch.

Matt

[johnnyx]

Transformers are rated full load. If the regulation is stated as 5%, then off load it will be 5% greater than the rated voltage. This is more noticeable on small transformers, like 5VA, such that a 12v transformer may measure 15v off load. (This makes replacing a burnt-out transformer a bit of a gamble, if a good one is measured off load). Then there is the peak-to-peak ripple voltage, which is determined by the load current and reservoir caps.

[metalman]

This is how I usually size a trafo (using the example of the GC i'll be building shortly using dual 25V secondaries).

34V Rail x 4A peak current draw = 136W instantaneous demand.
X2 Channnels = 272W for both channels.

Personally, I've found that to avoid voltage drop on a trafo at peak draw I need about a 50% overhead in the VA rating.
Thus 272 x 1.5 = 408 VA.

Now in my case, I had a great opportunity to pickup a 500VA at about the same price as a 400VA, so I chose the 500VA. That and I plan to do some playing around with some design variations which may have higher power draw.

Now, I am also ignoring that the PS smoothing capacitors will also reduce the voltage sag that occurs during transient peaks (rule of thumb: 1F = 1A draw for 1s with 1V drop), and the fact that the majority of the time all audio amps are trickling along at only 1-2W output, so it could be reasonably said that I take the overkill approach.

Just my two cents.

[morfeas]

Power rails sing along on my 3875 GC with 250VA toroid per channel.

Full power sign wave 1K I read max voltage drop 9V!!!

Go figure...

[SheldonD]

is this 9 V drop on AC input to rectifiers or DC drop at the chips?

I would guess the DC as I would not expect this on the AC with such healthy sized Xfrmr.

What is the nominal (no load) voltage DC?