Simple variable AC supply?

Just for grins, I'd like to build a simple variable-voltage AC supply, an
electronic Variac. I want variable p-p voltage control, not variable
time-domain control.

Suggestions?

Sparky

 

Do you have similar longings to construct an electronic hammer for
opening peanuts?

You possibly already have the makings of such a system, in the shape
of a high power audio amplifier. You would need to add a mains "audio
input" signal and a mains transformer to bring the amplifier output up
to the required maximum output level. Adjusting the volume control will
give the required output voltage setting.


Of course, if you want to design and build your own audio amplifier
(sorry, electronic variac) -feel free.

 

You possibly already have the makings of such a system, in the shape

I have all those. Hadn't considered using them for this application... But
then that's the difference between us (c;

What would you suggest for the output transformer? What primary and secondary
specs should I look for to best match the speaker-load outputs to AC
power-inputs (typically less than 50 vac and a few hundred miliamps).

Thanks,

 

You know the rated maximum rms output power for a particular impedance
load for your amplifier, so you can work out what the rated maximum rms
output voltage is...

You know the maximum rms output voltage you want.

So you can work out what transformer you need.

Let's say your amp gives 20v rms out at full power. You want 240v rms.
So you use a 240 <> 20 volt mains transformer, connected the "wrong way"
around - with a VA rating equal to the amp power maximum, or determined
by the load, if less.


(BTW I'm not sure what you last words meant.. Do you really only need
50v rms at a few hundred mA, or is your amp only capable of giving 50v
rms at a few hundred mA - and you want more voltage?)

Which gives a clue that a mains transformer may not always be needed. If
your amp produces 50v rms and you only need 50v rms, then you can
possibly omit a transformer entirely - depending on isolation and
earthing requirements.

 

Oh yeah!

Years ago my then father-in-law wanted to "play" with servos and I picked up
an amplifier from Lafayette (now out of business.) I think it was rated
for 70 watts and I "tested" it by connecting a bell transformer to the input
and put a 60 watt lamp on the output. You need am amplifier that has high
Z outputs. I don't know the details, but rather than saying, 8 ohms, 4
ohm, 16 ohms, it will saying something like "70 volts".

 

And the problem with a $4 triac light dimmer is...?

--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email:

Please visit my GURU's LAIR web site at http://www.tinaja.com

 

And the problem with a $4 triac light dimmer is...?

Good only for non-inductive loads? Choppy, non-sine signal may not be good
for a particular circuit? RF generated by the dimmer could wreak havoc with
the DUT? Just to mention a few...

Sparky

 

Thanks for the help.

So for a general-use, a 120 v (I'm in the USA) <> 50 v (the amp is rated 200
watts into 4 ohms) transformer rated for 200 VA would be, roughly,
appropriate?

I have regular need to power some PCBs from German industrial equipment that
uses 34 vac as power. Some stuff requires up to 50 vac. Don't usually need to
provide variable full mains voltage.

The "audio amplifier as mains power source" is new to my head, so... Is
impedance matching (for the sake of preserving the amplifier's output stages)
important?

Thanks,
Sparky

 

If it is rated at 200w rms - then its rms output voltage (maximum) is
28v. So a 120v 200VA transformer with a 50v secondary would give you
0-60v ac (approx).

See above. use a 120v <>50v 200VA transformer. Or use a more powerful
amplifier..

eg a 1000W rms amplifier into 4 ohms gives over 60v rms at full output.
No need to pratt around with transformers (other than 1:1 if needed for
isolation).

Nope. The output impedance of an audio *power* amplifier is way, way
lower than the load impedance at all times in normal operation. But I
would suggest that you use an amp that is protected against speaker
cable shorts and thermal overload - especially if your power
requirements are anywhere approaching its rms full output rating.

You do need to make sure that the amp is capable of sustained running at
the load level you require.

 

A generator.
D from BC

 

The load he wants to use it with?


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

Hi SparkyGuy,
I posted my first thoughts about how to implement an electronic variable AC
supply.
See alt.binaires.schematics.electronic
Mike

 

Ignoring for the moment the issue of damage to the amplifier,
Consider the cases of zero ohms and infinity ohms. If your load
is zero Ohms (dead short) you will get maximum current through
the load, but no voltage across it. If your load is infinity
Ohms (open circuit) you will get maximum voltage accross the
load but no current through it. Power is voltage times current,
so both of those cases will give you zero watts of output power.
Somewhere in between is a load that maximizes power. This is
usually somewhere close to the lowest ohms speaker that the
amplifier is rated to drive.

Keep in mind that audio is "peaky": it spends most of the time
at a lower-than-peak power level. Your steady signal will
generate more heat, so be sure to monitor the amplifier
temperature as you slowly raise the output level.

 

Very definately not for an audio power amplifier. The output impedance
is very, very low - much much lower than the load impedance will ever
be. It is basically a constant voltage supply.

Easily proven, if you have an amplifier running at full power output
into its designed load impedance and measure the output voltage, then
double the load impedance, the output *voltage* will hardly alter.

The process you are describing above is that of matching a load to the
output impedance of a power source. Do that for an audio power amplifier
and the result would be a tiny fraction of an ohm.

The starting point for a power amplifier design is the design output
power and the design load impedance. From this comes the design output
voltage swing.

Or put otherwise, a 1000W rms power amplifier may only be a 200W rms
power amplifier with a 1000W rms transient capability. Very true. But
that's marketing for you.

 

There have been many suggestions but I don;t think this idea was one
of them:

With an array of transformers with each one having three times the
turns ratio of the other, you don't need very many steps to cover a
wide range. The secondaries can be wired in series and the primaries
connected up by means of triacs. The triacs either wire them as
adding, shorted or subtracting.

Starting with a 1V secondary, you need 1, 3, 9, 27, and 81 to get
zero to 121Vac.
That would be 5 transformers.

Since transformers are cheaper than semiconductors, you could use
transformers with split primaries to get more options per core. You
could have parallel subtracting, series subtracting, shorted, series
adding, and parallel adding. This increases the base to 5 so you only
need 3 transformers to get from 2V to 120V.

0, 2, 4 < - 0ne transformer

10-4 .. 10+4 < - Two
20-4 .. 20+4

50-20-4 .. 50+20+4 <- Three
100-20-4 .. 100+20+4

 

No it isn't. The only "process" I described is matching the load
to the rated load found in the specifications. The rest was an
explanation as to why the impedance matters.

Again, my suggestion was to match the load to the lowest
impedance rated load found in the specifications. You are
arguing against something I never wrote.

I didn't want to confuse SparkyGuy with details that won't help
him/her, but for the record:

Most power amplifiers use negative feedback to obtain an output
impedance near zero. This confuses some technicians who had
previously learned to match the load to the output impedance on
the old tube amps. With a an output impedance near zero, the
voltage doesn't change much as you change the load, and thus a
naive technician might calculate that the lower the impedance
of the load, the higher the current. That is, until he tries to
calculate a fixed voltage driving a dead short and gets infinite
power as the answer... In real life, the amplifier will have a
maximum output current, and if you make the impedance of the
load too low the amplifier output will most definitely *not*
remain a constant voltage with a near-zero output impedance.
It may shut down or catch on fire, or it may clip, or it may
kick in a limiter (many Carvin amps do this). But one way or
the other the output voltage *will* drop. So, how to avoid
hitting the current limit? Same advice as before: match the
load to the lowest impedance rated load found in the
specifications.

If you reply, please check your post to see whether I actually
wrote what you are disagreeing with. Thanks!

 

The most simple method is to *use* a variac...

 

There are also amps, as used for US public address systems, that are
"voltage matched" instead of "impedance matched". Nominal voltages are
25 and 70. May be simpler.

For the use stated, a small Variac would be a lot cheaper.

 

On Sun, 22 Jul 2007 07:28:29 -0700 in sci.electronics.design, Don

Have you ever heard of a sine wave?