Getting a power supply from the heater windings - a long and winding road.

[R.G.]

I've seen I don't know how many posts here and other places asking for how to power X, Y or Z addons from the heater windings.

The general idea is to add some smallish something or other by using some small addon to the heaters to make the necessary DC without stealing it from the B+ or having to mess with the AC wiring.

I have come up with a way I haven't seen before that may help. Sorry - it still involves a transformer - but it's a little one, and this doesn't hurt as much as some other ways.

There are good reasons not to install a new baby power transformer running from the switched AC power line in an amp, not least of which is the safety problems this can cause. The heaters are the next choice, but they have inherent limitations in many amps because they need to be ground referenced to avoid hum. This may be by a center tap on the winding, an artificial center tap with resistors. Sometimes the whole mess is elevated. All of these make using the heaters for making a secondary DC supply very difficult.

It struck me the other day that transformers, like the opening into the time-space rift in Ghostbusters, swing both ways. This is a common dodge to get power for a tiny amplifier, using two 120Vac:6Vac transformers with the 6vac windings connected to get an isolated 120Vac (or 240Vac) on the second transformer.

But you can also get transformers with two isolated secondaries. If you had a transformer like the common dual-120Vac primary and dual 6.3Vac secondaries, you can drive one of the secondaries from the heater winding in an amp and take isolated 6Vac out of the other secondary if you're careful to open circuit the primary windings, which will have their nominal voltages on them too.

That is, you can use a dual-secondary transformer as a 6Vac to 6Vac isolation transformer by ignoring the primary. This isolates the output from whatever the real heater windings are doing, and lets you ground reference whatever you do with the isolated winding any way you want.

Granted, this is still using a transformer, and that's a PITA, but it's a *small* one, can be *very small* if you find a good toroid to do it, and it avoids mucking with the incoming AC power, so its safer. We're talking 1"/25mm cubes here for modest powers.

If you need something that's not 6vac, and can't get it by doubling the isolated 6Vac you get from this trick, you can use two 120Vac (or 240Vac) primary transformers, one with a 6V secondary and the other with whatever you need. Hook up the two primaries together, and you get 6Vac in and a custom selected voltage out. Again, because of the modest power involved this can be small, and not very expensive.

Mouser has some Tamura transformers that might work fine for this for about $5 - $8. Surplus is an option too.

It's a dodge that may make a harmonious add-on possible without messing with either the AC power or upsetting the heater arrangements.

[Wombaticus]

Rather than capping off open primaries, is there any reason not to use a small transformer with a single primary and single secondary? Hammond makes a series (over a fairly wide range of VA ratings) intended to convert from 115 to 120V. For example here's a fairly modest one that's only a couple of inches in it's longest dimension. It occurs to me that if the intention is to rectify and regulate, then the slight bump in voltage would probably be welcome to help ensure that the regulator won't drop out.

[R.G.]

Rather than capping off open primaries, is there any reason not to use a small transformer with a single primary and single secondary?

Not at all, except that you have to connect the primary to the amp's internal AC power. In fact, I've advised people to use a similar set up in the past, but the AC power connection is an issue. You simply have to do the AC power wiring to a higher standard than the secondary wiring, even with high voltage secondaries.

The intent is not to make both high voltage and heater voltages for a small amp. That's available as a single transformer, or by a number of two-transformer setups.

The intent is to make it easy to get an isolated voltage for making up any number of things needing a little DC supply in pre-existing amplifiers, and to not need AC power wiring to do it. Using a dual-secondary transformer and actively ignoring the primary(ies) to get a 6Vac to 6Vac supply is something I've never seen done before, and I thought it might be useful.

It occurs to me that if the intention is to rectify and regulate, then the slight bump in voltage would probably be welcome to help ensure that the regulator won't drop out.

It is a good idea, all right. There's a wide range of things that can be done with an isolated low voltage AC supply in an amp, and not all of them can be done well with the 7.1V or so that full wave rectifying 6Vac can get you. A LDO can get you 5V from that pretty well, but maybe you need 10V - or 12V. Or my favorite, 9Vdc, regulated, as I put in the Workhorse amplifier series.

For that, you need to either be familiar and willing to do voltage doublers or triplers for the higher DC, or to go to the two-transformer setup to get a different voltage to rectify.

Once you've bought into using a transformer at all, the question then becomes whether it's better to not connect to the amp's AC power line stuff. For many people, avoiding the safety issues is a good idea.

And this is one more illustrations of the many ways of de-frocking cats.

[FredB]

I've looked into this a good bit a while back, so I know exactly what you are talking about. The LP Standard I built used an auxiliary power transformer that was powered from an unused 5V filament winding. That Hammond transformer is a rather awkward size, and I found the triad flat-head series of transformer are a much easier fit. They would work fine as you have outlined.

What I am wondering about is the difference involved between powering them from the mains or from a secondary of the power transformer. Would you elaborate on that some more?

[Gingertube]

This works well.

I have not actually used both secondaries but I have done the following:

In both HiFi Amps and Guitar Amps: wired a 115V AC to 12V AC small tranny with the 12V AC across an existing 6.3V heater winding - rectified and filtered the 115V AC (old primary) winding to generate a -70V bias supply.

You can manipulate the final voltage by choosing a transformer with a secondary of 6.3V or above (but not less) and using either 115V or 230V primaries.

For a lower power amp or a preamp - use a dual 115V AC primary to 6V secondary to power the heaters and a wire a second identical or lower VA rated transformer in reverse across the first with the dual primaries in series to generate the B+ works very well.

Again you can manipulate the final voltage by using the dual primaries of the second tranny in either series or parallel, using a straight bridge rectifier or voltage doubler and also choosing the second transformer with a different secondary voltage.

When reverse connecting a second transformer like this the ONLY limitation is that its secondary voltage must be greater than or equal to the secondary of the first transformer secondary. Also calculate the VA for the second tranny and make sure the first is specified for that extra load.

I particularly like doing this for bias supplies in HiFI Amps as it gives additional isolation from mains born noise and transformer coupled noise, particularly diode switching noise from other windings on the first tranny. I do it even if the first tranny has a bias winding, to get that noise immunity.

Cheers,
Ian

[defaced]

Rather than capping off open primaries, is there any reason not to use a small transformer with a single primary and single secondary? Hammond makes a series (over a fairly wide range of VA ratings) intended to convert from 115 to 120V. For example here's a fairly modest one that's only a couple of inches in it's longest dimension. It occurs to me that if the intention is to rectify and regulate, then the slight bump in voltage would probably be welcome to help ensure that the regulator won't drop out.

Wouldn't a line voltage rated isolation transformer perform poorly with regard to current capability at such a low voltage? Meaning, the transformer's rated VA will not be able to be reached at low voltages do to the smaller wire diameter, and subsequent reduction in the ability to source current, as compared to an identical VA transformer which was designed for 6.3v.

[R.G.]

What I am wondering about is the difference involved between powering them from the mains or from a secondary of the power transformer. Would you elaborate on that some more?

Sure.

At one level, there is no difference. While there are some details to be worked, out, The transformer doesn't care what winding is the primary or the secondary. You can use any winding as a primary, any or all of the others as secondaries. This works fine as long as the voltage and frequency you put on the winding you're calling the "primary" pumps the magnetic field in the iron up, but not so much that the iron starts into saturation and causes excessive iron losses.

Each winding on a transformer has a certain number of turns. When you put an AC voltage across it, this sets a certain number of volts per turn on that winding. What's really happening is that this sets up a certain magnetic field condition, and *every* turn of wire around the core gets (ideally) the same voltage per turn.

If you imagine that we had a transformer designed for 120Vac to 6Vac, and that the core was sized so we could use 120 turns on the primary, or 1V/turn. The volts per turn is the same for all turns, so the 6V winding has 6 turns. If we remove the power from the 120t winding and put 6Vac on the 6V winding, making for 1V/t, then the 120t winding gets 120V. If we put 5V on the six-turn winding, the 120 turn winding produces a voltage of (120/6)*5 = 100vac, because the volts per turn is lower, but the ratio of the voltages stays the same.

So the transformer, to a first order, doesn't care which winding is the "primary" and powered by an external source, and which others are "secondaries" and have power sucked out of them. There are some gotchas hiding there, but that's how to think about it first.

Given that the voltages stay in the same ratio, we only have to worry about how much current we can pull out, which gives us the power rating. This gets a little tricky. Transformer designers DO design for one winding being a primary and others being secondaries, because they allocate window area to primaries and secondaries to get the same current density in each one. The more window are you give a winding, the bigger the wire can be in that area, and the lower the resistance, and heating losses.

Standard first-try for allocating window area is half to the primary (where 100% of the power comes in) and half to all of the secondaries, where 100% of all the power goes out. This gets juggled as the design is refined, but it's a good starting point.

By going to a transformer wound with a 120Vac primary that's not being used, and using two equally-rated secondaries as a primary and secondary, you've cut the available window area for the new low-voltage "primary" in half, so the wire is not as big as it would be if you'd designed it for 6Vac to 6Vac in the first place. And there's this whole 120Vac winding that's not being used at all, so it does not heat from current losses either. The iron losses are fixed by the voltage on the "primary" and the frequency of the AC fed to it, so they stay constant.

It gets messy figuring out what power can be drawn. At a first guess, you should be able to pull as much current out of the new "secondary" as the transformer was rated for in the first place, because an equal (nearly) current will be flowing in the new low-voltage "primary" as it was designed for, so those two windings should be making the same heat that they would have in normal use. And you'll get out half of the output power that the transformer was rated, as you can only get power out of one of the secondaries, not two. But there's that whole 120Vac "secondary" that's taking up half the winding area and not doing anything at all, not making heat. So you can pull more current through the low voltage "secondary" and "primary" than you could in normal use. Probably not twice as much, but more. I'd have to dust off some very rusty math to estimate this.

So I'd feel very comfortable with an example transformer rated for 120Vac to 6Vac/0.5A + 6Vac/0.5A being used instead as 6Vac to 6Vac/0.7A. Or with taking some amount of current out of the 120Vac. As mentioned, this is a great way to get a bias supply, as you can use a weenie little 120Vac to 6Vac at maybe 200ma or so backwards to get a high voltage winding for a bias.

The ultimate test is how hot it gets, of course.

I talked around the issues quite a bit. Did any of that hit what you were wondering about?

[R.G.]

Wouldn't a line voltage rated isolation transformer perform poorly with regard to current capability at such a low voltage? Meaning, the transformer's rated VA will not be able to be reached at low voltages do to the smaller wire diameter, and subsequent reduction in the ability to source current, as compared to an identical VA transformer which was designed for 6.3v.

I think it would, and in fact that's the exact reasoning that made me realize that a two-secondary transformer was the same as a low-voltage isolating transformer plus a high voltage winding.

The thing that makes using a high voltage isolation transformer a bad deal is the winding resistances and the window area allocation I mentioned in my previous post. When you design a transformer for a certain power, you pick the core to have enough iron area to hold a big enough magnetic-field bucket, and enough window area to wind the necessary turns into. The window area is fixed, and the iron can only stand so many volts per turn without saturating. So a 120V winding design not only has many more turns than a 6V winding, the wire has to be much, much thinner to fit in the window area. If you use this winding of tiny wire at a low voltage, the resistance of the wire at the higher currents you want from a lower voltage causes BAD voltage sag and heating from the wire resistance.

So yes, using a 120Vac:120Vac transformer to transform 6Vac to 6Vac is a bad idea. You get a lot less use of the iron and copper than you need.

[FredB]

I got the impression you were stating that there are more safety concerns involved with connecting an auxiliary power transformer to the mains than there are when connecting one to a secondary of the power transformer. So, I was wondering just what the particulars of those concerns were.

[R.G.]

Ah. Got it.

Yes, there are more safety issues with the mains than with hooking to a secondary.

The big issue is that the AC mains are connected to the entire planet, and also to the signal ground of most audio gear. So if you hook an auxiliary power transformer to an isolated secondary, you can get zapped if you connect yourself across the winding. But if you wire it up to the AC mains, you only have to touch one of the AC power wires and then something that's connected to the planet or one of those other grounds to get electrocuted.

The entire world is one conductor to the AC power lines. Hooking to an isolated secondary, you have to make two mistakes at the same time to get zapped; furthermore, if what you're connecting to is a lower voltage, the zap may be much less bad or nonexistant.

[FredB]

I see. Well, after I use my two auxiliary transformers I bought to wire up to the mains, I'll have to rethink my approach again.

[R.G.]

It works either way, of course. If you've already committed to wiring up another transformer to the mains, and you can do the wiring safely, there is no problem. But there is value, sometimes a lot of value, to not having to get over the big-ish hump of having to already know how to do AC power wiring safely.

[Steve Conner]

I've used this circuit with success. http://music-electronics-forum.com/t12836/

[Dave H]

If you don't need much power you could use an isolated DCDC converter powered from the heater winding by a bridge rectifier and capacitor. I don't know how noisy it would be but it's worth trying it to find out.

[R.G.]

Yep, isolated DC-DC converters do much the same thing, providing an isolated DC source for other miscellaneous circuits, while not affecting the use and actions of the 6Vac winding, not disturbing the ground reference of the existing winding CT, and not requiring a hard CT instead of a two-resistor tap to reference the heaters, and not caring if the heaters are elevated.

Isolated DC-DC converters are handy, but I've always had problems getting them in the input/output voltages I wanted at a price that I could afford.

For the simple case of full wave rectifying a 6Vwinding, you get about 7Vdc as an input voltage to a converter, so you need a converter that takes that voltage in and produces the voltage(s) and current(s) you want.

I think it's my frustration with practicalities like that one which led my mind to the dual-secondary scenario.