Favorite transformers for chip amp and gainclone?

[danielwritesbac]

Hi guys! I have a question.

What's your favorite transformer for chip amp and gainclone?

My fav is Stancor "NV" style, like this one.
http://www.alliedelec.com/Search/Pro...px?SKU=9288566
It works well for designs like 47 labs: ". . .where only the simplest. . ." have such limited power supplies that transformer choice is critical.

So what transformers do you hold in high regard? Let's see em'

[ratza]

I like toroids, but I would definitely choose R-core if I had enough cash. The R-core are the best transformers for any audio power amplifier which money can buy.

[danielwritesbac]

Specific?
Which toroid's do you like?
Which R-core do you like?

[ratza]

It doesn't matter, just to be built OK. The toroid which is used in my GC is a 400VA monster and even if I stick my ear into it I cannot hear anything. Now take a look at this cutie:
http://www.selectronic.fr/article.as...r=90.3223-9999

[danielwritesbac]

I did find some R-Core on ebay.

One thing that the Toroid and R-core can do better than my fav Stancor NV series is. . . the smaller toroid and R-core are so short in height that the amplifier enclo$ure cost$ much le$$.

On a recent discussion, a friend and I, who hardly ever understand each other, both agreed on a minimum 80VA capacity and that larger 120VA to 160VA capacity would be better. Neither of us illustrated all the details as for why. For me, it relates to bass response, and I do not know the details, except that the errata does exist.

Power supply boards with extra poles and sufficient caps (at each pole) can help, but they don't completely cover the issue. A "right sized" transformer gets it done more easily.

This makes for 3 amper to 4 amper EI core transfos in sizes of about 4 x 4 x 3.5 inches (plus an extra 3" clearance away from signal-in/amp); however, Toroid and R-core are sized conveniently smaller at equivilent strength.

I had assumed that I'd simply use wooden enclosures (they have plenty of room for the EI core and there's plentiful insulation); however, the timeframe for me to make a seemly-looking wooden box was unexpectedly huge, and, in fact, a barrier to completing projects.

Compare the VA capacity requirement for doing it the easy way, to the cost of enclosures, and you'll see the point.
Its this:
If you're a good carpenter, you're "home free" on saving some cool cash on your transformer. Otherwise, you'll need the R-core or Toroid that will fit into a prefabricated enclosure.
I find this out the hard way.

[ratza]

The recent discussion took place between you and AndrewT, isn't it? I'm not a carpenter at all, but my enclosure is made of wood and it costed me around $5 at a local store, including cutting at the size I specified. As you can see, it isn't _that_ expensive.

Now, regarding the sound, IMO no juice, no life. Let's say that my amp is capable of 40W output, audible undistorted. The true efficiency of AB class is around 40...60%, so I took the worst case, 40% efficiency for a LM3886. This means that at full continuous power I will draw 100W per channel, right? Two channels equals 200W. But at 30V per rail as I have, the maximum peak power can be 30^2/4 = 225W. Two channels equals 450W. Now, instead of drawing this amount of power from the filtering caps, why not to have this power available all the time? So that's how I ended with a custom made 400W toroid.

[danielwritesbac]

Originally posted by ratza
. . . I'm not a carpenter at all, but my enclosure is made of wood and it costed me around $5 at a local store, including cutting at the size I specified. As you can see, it isn't _that_ expensive.

Now, regarding the sound, IMO no juice, no life. Let's say that my amp is capable of 40W output, audible undistorted. The true efficiency of AB class is around 40...60%, so I took the worst case, 40% efficiency for a LM3886. This means that at full continuous power I will draw 100W per channel, right? Two channels equals 200W. But at 30V per rail as I have, the maximum peak power can be 30^2/4 = 225W. Two channels equals 450W. Now, instead of drawing this amount of power from the filtering caps, why not to have this power available all the time? So that's how I ended with a custom made 400W toroid.

The diodes will still drop, unless. . .
You could have quad secondaries, or a pair of dual-secondaries transformers. That would make 4 rectifiers (16 diodes). Just a thought. I'll do it someday, but the way I make power supplies, it would have 35 capacitors (only 16 big caps). Only? LOL! Okay, I have to laugh at myself sometimes. Now is good.

[AndrewT]

Originally posted by ratza
Let's say that my amp is capable of 40W output, audible undistorted. The true efficiency of AB class is around 40...60%, so I took the worst case, 40% efficiency for a LM3886. This means that at full continuous power I will draw 100W per channel, right? Two channels equals 200W. But at 30V per rail as I have, the maximum peak power can be 30^2/4 = 225W. Two channels equals 450W. Now, instead of drawing this amount of power from the filtering caps,

that's going to be difficult to justify.
A polite way of saying "rubbish".

Originally posted by ratza

why not to have this power available all the time?

because the AC waveform of the mains supply guarantees that voltage to supply the downstream circuitry is NOT available MOST of the time.

[ratza]

Rubbish or not, I honestly don't care about others' opinion when it comes to my taste. If I am satisfied with the sound produced and I don't take into account other opinions about the sound because I listen every day to it, not them.

[AndrewT]

I'm not referring to the sound.
It's your assumptions and maths that cannot be justified.