[R-390] Variacs and solas: additional thoughts.

[Roy Morgan]

At 10:02 PM 4/29/2005 -0700, W, Li wrote:
> >(DO NOT bring it up slowly on a variac!)....
>
>I was re-reading an old e-mail of yours, is your advice
>based on the fact that an unmetered Variac could put
>over 125VAC onto the primary of the power transformer?

W. Li,

Actually, it can be much worse.  Many variacs are wired for overvoltage and 
would deliver up to 150 volts if they are ment to be run from a 115 volt 
line.  Even if a variac has a meter, it's easy to operate it incorrectly 
and not notice the meter until after some damage has been done.

I just did two experiments:

1) With a second variac (a GR W5MT3), I powered a STACO Adjust-A-Volt rated 
for 120 volts input, 7.5 amps output.  This thing is both useful and 
dangerous in that it has a two-position power switch: "120 Volts"and "140 
Volts".  I fed it 130 volts to simulate a 117 volt variac on a 124 volt 
line.  With it switched to the 140 volt output, the output was 150 volts 
running a 4+ amp halogen work lamp.

2) I re-arranged things and fed the General Radio W5MT3 with the 
STACO.  The GR unit is rated at 115 volts input, 5 amps output, 0-135 
volts. It is set up for overvoltage and has the normal ON-OFF power 
switch.  I ran it with 121 volts input (the house line voltage at the time) 
and the output at max setting was 142 volts running a 100 watt lamp.  I 
then increased the input voltage to 124, which I get as line voltage from 
time to time, and the output was 146 volts.  If I run it on normal line 
voltage of 122, it produces something like 147 volts. If I set the dial to 
"115", I get whatever is the line voltage at the time, normally 121 to 124.

So, take your pick, 146 to 150 volts input to your 50 year old radio rated 
at 115 or 117 volts. I'd expect trouble.

The danger with the STACO transformer is that you can easily push the 
switch to "140 volts' instead of "120 volts", and if the knob is set full 
up, you get 150 volts output.

A further note on variacs in general:
The fusing or circuit breaker arrangements can be important.  The GR W5MT3 
has a circuit breaker on it and the STACO has a fuse.  My guess (without 
investigating it just now), is that both of them are in the INPUT of the 
unit. This means that if there is a short cicuit or heavy overload on the 
ouput, and you start at 0 volt setting and increase the setting slowly, 
you'll be overloading the windings as to current, and the fuse or circuit 
breaker will not trip or blow.  In my opinion, variacs should be fused or 
have circuit breakers in BOTH input and output. This will protect the thing 
wherever an overload would occur.  I have had such failures burn the low 
windings of a small variac in a DC power supply, and recently got a one-amp 
GR W-200B one-amp variac with evidence of the same thing.

Incidentally, if you run into a 400 cycle variac, it can be run on 24 volts 
AC. (The allowable voltage is proportional to the frequency.) This can lead 
to a convenient low voltage power supply.  If you hook a 400 cycle variac 
to 60 cycle supply at it's rated voltage, expect smoke in a hurry.

Some notes on constant voltage "sola" transformers:
Some are the "harmonic neutralized" type and some are not.  All these 
things operate on a magnetic circuit basis with non-linearities and very 
high circulating currents.  Part of the current is harmonics of the supply 
voltage due to the non-linearities.  Thus, the output waveform may well be 
regulated to 115 or 120 volts RMS, but have very high peaks in a distorted 
waveform.  It seems to me that this could create high B+ voltages in 
rectifier power supplies.  Filaments may be very happy at the right RMS 
voltage but the B+ may rise very high due to peak rectification in the supply.

It  has been noted on the reflector that "sola" type voltage stabilizers 
create both noise and a lot of heat, especially when they are lightly 
loaded.  The good advice given was to run them well loaded at all times to 
avoid trouble.  It may be that 20 to 40 percent loading will bring down the 
waveform peaks to reasonable levels, but some experimentation is in 
order.  Installation instructions advise making sure there is plenty of air 
circulatoin to avoid overheating.  The Sperior Electric Company makes Sola 
brand regulators still, and their website has interesting reading.

I have a 230 volt input unit here that runs the photo darkroom to avoid 
troublesome changes in enlarger lamp brightness due to changes in line 
voltage.  When it is switched on, there are significant overvoltage 
transients, and I make sure that nothing but light bulbs are on the line 
when I start it up.  A surge supressor outlet strip might help protect the 
few solid state things I use (timer and enlarging densitometer), but it 
also might get a whacking big surge each time I start up.  The Oscilloscope 
will tell the truth in the matter.

By the way, two of  the three capacitors in this thing failed open at one 
point.  Each was rated at 8 uF and 660 volts ac. The output voltage 
collapsed to near zero.  I was able to locate a fairly high voltage 
(440vac) motor starting capacitor of the right capacitance that was about 
one fifth the size of the original three caps. It seems to work fine.  The 
voltage on the capacitor during operation is normally way above 440 volts 
ac, but it was all I could find, and they normally have breakdown voltages 
very much in excess of the rating.

I once had a Sola made for 50 cycle operation (for use in Europe). The 
thing did not work well at all on 60 cycles. It regulated at some 150 volts 
output, and over a  limited range of input voltage.  As I understand it, it 
is not feasible to modify the 50 cycle units to work on 60 cycles.  I sent 
it to someone in a 50-cycle country and he was happy to get it.

If you have a sola transformer that's rated at 120 volts and you want to 
get 115 volts out of it, add a voltage bucking transformer, or see if there 
is room to add a few turns of wire in series with the normal output over 
the existing winding and arrange the connection to reduce the 
output.   I've not added bucking turns to a sola but it's an attractive 
possibility.

Normally, voltage regulator transformers will operate well just a bit above 
their rated output power (they are rated in volt-amps, not current, because 
of the the way they work and the harmonic content of the output.)  Above a 
certain point, however, they collapse and go into a low-output-voltage 
condidition. This is good, because they are inherently current 
limiting.  Apparently, there is no harm to the thing and it automatically 
recovers.


Here are some questions to investigate:

Variacs:
1) What particular makes and models of "variacs" are set up for overvoltage 
as made?
2) Do small variacs behave differently than bigger ones?  (I doubt it.)
3) Are fuses or circuit breakers normally in the input as made?
4) What errors are found in the dial readings due to line voltages being 
higher than the unit was made for?
5) Are the voltmeters found on variacs at all accurate?

Constant Voltage Transformers:
1) Do small ones behave in a way similar to large ones?  (I have examples 
from 60 volt-amps up to one Kilowatt.)
2) What peak voltages come out of the harmonic neutralized type, and what 
from the non-netutralized ones, as a function of loading.
3) Does harmonic content in the output lead to high voltages in rectifier 
power suppllies?  Are choke input plate supplies affected in the same way?
4) What are the overload characteristics of these things?  Are they the 
same for small and bigger units?
5) Where above rated output do these things collapse, and do all such 
transformers behave this way?


Sooo many projects, sooo little time!

Roy





- Roy Morgan, K1LKY since 1959 - Keep 'em Glowing!
7130 Panorama Drive, Derwood MD 20855
Home: 301-330-8828 Cell 301-928-7794
Work: Voice: 301-975-3254, Fax: 301-948-6213
roy.morgan at nist.gov --