Re: What is a Variac?

Yes -- but if that is done, it is usually built into a power supply.

Anywhere variable AC voltage is needed -- including for testing AC-powered equipment to see how well it performs at extreme voltages.

Variac was the brand name used by General Radio, who may have been the originator of the device, and certainly the maker of the best ones which I have used.

Superior Electric called their version a Powerstat.

But all are simply a variable autotransformer, wound on a toroidal core which concentrates a lot of magnetic field in a small space.

The typical General Radio Variac for house line voltage has four fixed taps, 20V, 100V and another 20V all in series, not counting the variable tap. This can be wired with the 120 VAC at the two end taps, so the variable will go from 0-120VAC, or it can be wired with neutral at one end, and at 20V from the other end, so you can vary the output voltage from 0-140 VAC (useful for some testing.) There are the two 20V taps at the two ends for one reason -- the Variac can be panel mounted, with the shaft coming in the bottom mounting plate, or it can be a bench use one sitting on the bottom mounting plate with the shaft coming in from the top. This allows you to wire for 0-140VAC in either situation, even though the knob is turning CW or CCW for higher voltage depending on viewpoint.

I say "for house use" because there are also 240VAC industrial versions, which have an additional (center) tap. You can feed in 120 VAC between one end and that center tap, and get up to 240 VAC output (at the cost of lots more current from the 120 VAC line than you feed out at 240 VAC, of course.

The Superior Electric "Powerstat" is pretty much the same, including similar voltage taps.

I've seen one such variable autotransformer (actually a Powerstat, not a Variac) set up in a recording room at a South American US Embassy, used to adjust for sometimes rather low power line voltages. (Unfortunately, it was a large one and had a knob which looked like a steering wheel, and a small kid walked up to it and spun it clockwise, burning out the rectifier in a recorder.)

Well ... I have two sizes of Variacs (2A and 7A) made for 400 Hz. Minimum frequency is 350 Hz before it saturates. (And 400 Hz is normally used to power aircraft instruments because at 400 Hz, you don't need as much weight of core.) The core is shorter but the same diameter as the same current rating for 50-60 Hz.

For three phase, there are three identical units, stacked up and using a common shaft. This gets heavy quickly. :-)

Enjoy, DoN.

The maximum output current decreases considerably as the brush moves above the line voltage. My 10A 120/240V Powerstat 236 is rated for only 4A at 120V in, 240V out. Trying to pull more current loads it down.

The large rotating plate in the back is the heatsink for the brush and is at output voltage. On some types the shaft is also at output voltage and needs a well insulated knob and clearance from a panel mounting hole.

-jsw

O.K. That makes sense. I've never had to deal with that problem. And I don't have one on hand to read the current limits from the data on the terminal block.

O.K. Found some illustrated in eBay auctions. W5H series is normally 5A at 120 VAC, and this is 2A at maximum output voltage. (The input and load currents add in the part of the winding below the

The GR (General Radio) Variacs have a steel shaft with a black Bakelite shell, so it is insulated from the rotor and brush assembly. Now if somebody replaced the original shaft with a plain steel one, all bets are off. :-)

And, of course, if you have a three phase assembly, the shaft

Here is a rather extreme 3-phase one on eBay:

Auction # 162543903031

It has two Powerstats wired in parallel per phase, for six in total. Max 90 amps out @ 240 VAC.

The shaft appears to be bare metal, so there must be insulation where each rotor connects to the shaft.

Enjoy, DoN.

The rating of an autotransformer is generally NOT in amps, but in VA (Volt Amps) All 4 of my variacs and powerstats are rated in VA or KVA. They DO also have acurrent rating, but that is the maximum INPUT current limit. I have them from 210/220va (210 on 50Hz, 220 on 60) to

If you have a 1200va powerstat it is food for 10 amps in at 120 volts and only 5 maximum at 240 volts out.

Try this. It's too big for me to check with my limited dialup bandwidth or monthly cellular data allocation.

From my download of the Powerstat catalog C7002-1:

"CONSTANT CURRENT LOAD: Output that can be carried regardless of output voltage setting.

CONSTANT IMPEDANCE LOAD: loads such as incandescent lamps or resistance heaters in which the current drawn is approximately proportional to the applied voltage, increasing to maximum current at line voltage. These ratings apply only to units having maximum output voltage limited to line voltage."

Their tech support confirmed the catalog's 5A Const I, 7A Const Z rating of a Type 21 whose nameplate gave 3.75A as the limit.

When I tested a type 10 at its 3.0A Const Z rating an IR thermometer indicated the brush temperature as 100C. It was driving a 24VAC 250VA control transformer with a rectifier / capacitor output. The combination is well matched and makes a decent 10A variable power supply and battery charger. It will provide 15A briefly, until the Powerstat and transformer primary overheat.

I added the output capacitor to stabilize the readings of digital volt and amp meters.

-jsw

The core and winding's VA rating has a more comfortable margin than the brush's current rating, the one I observe since replacements are hard to find and expensive, and difficult to make because the carbon is brittle. I could only mill into an edge, milling out chipped it off.

The Powerstat chart gives both current and KVA ratings, at constant current and constant impedance, for metal panel (heatsink) and non-metallic or bracket mountings.

-jsw

Is there much difference between them? How was that handled?

This shows an 'improved' model, full 10A, which apparently took the place of the others.

Found a less expensive precision ammeter for your charging circuit, in case you're interested:

- The list of Obama administration disappointments would take three rolls of toilet paper to record. --BMF

Huh? Those ratings all apply to the -same- unit, depending on how it's used.

The article is dated 1959. I have the 5A version of that rectangular VARIAC with the handle and two-prong ungrounded outlet on top.

I buy them old, cheap and hopefully repairable at flea markets. My 3A Powerstat 10Bs were salvaged from a brand new machine (~1975) that fell off the customer's forklift and landed face-first on them. They broke in different ways and I was able to reassemble a basket of them into several good (?) ones.

The one I used to tame the buzz box welder transformer was on the front panel from a 1950's power supply.

"Input impedance: 10Euro"

Thanks for looking, but I buy little stuff mainly from Amazon to limit exposure of my credit card, and check my account afterwards.

I think a voltmeter that reads ##.## is good enough to determine the State of Charge of batteries, though I couldn't pass up that 5-1/2 digit Fluke for $25. 5% accuracy from 1ma to 10A is probably fine for current. I checked my 15-year-old truck battery at 200A this morning, it wouldn't matter if it was really 180A or 220A. At the low end knowing the key-off battery drain to within 10mA should show me if a relay sticks closed.

-jsw

That is offensively big. 57 MB total download.

Those passages do not appear in this catalog. After failing a search for the phrase "CONSTANT CURRENT LOAD" I backed off to just look for "CONSTANT" (case insensitive), and the typical references were like this example:

======================================================================

When series connected on 480 volt, 60 cycle single phase lines, the output is 0-480 volts, 4.0 amperes. When open-delta con nected from a 240 volt, 60 cycle three phase input, the output is

So -- you are allowed more current at full voltage than anywhere else on the winding. (Assuming input is also at the full voltage tap.)

And I disagree with them calling incandescent lamps and/or resistance heaters "Constant Impedance". Easy to demonstrate with an automotive lamp. Measure the resistance when cold, then attach to full voltage and measure the resistance and from that and the voltage, calculate the resistance. Lamps are commonly used for an approximation of a constant current source -- and heater elements. I've seen the latter in old tube radios to limit the surge into the filaments when starting cold, and adapting to varying line voltage.

O.K.

O.K.

Enjoy, DoN.

I think what they mean is you will get the most current at full voltage, but could draw the same 6.3A maximum current at a lower setting if necessary. You can NOT get that current above line voltage in the boost configuration, in that case the limit is 4A. I don't know if you can pull it below the input tap in the boost configuration where the volts per turn is higher.

At least that was the story when I was learning to design industrial controls. The derating curves don't reduce to a simple, easily remembered explanation or a single number on the label.

I've measured the iron loss on a few cores. It rises rapidly above the spec sheet voltage and frequency, they don't make good speaker crossover inductors.

Electrical engineers aren't taught the subtle imperfections of components and electricians fall short on the theory. I sought to fill that gap and become able to turn a scribbled schematic into a fully documented working machine.

This is Staco's explanation:

This addresses the limits to brush current, my concern when using them for bench testing:

The brushes drop a volt or so which limits circulating current in shorted adjacent windings. When I was looking for carbons to cut down I found that brush composition should be more or less matched in resistivity to the working voltage, to reduce motor commutator sparking. Presumably the original designer chose the proper cross section to control the current density.

In the 1980's I used a 4W night light bulb as the Zener limiting resistor in a test fixture that had to operate from an input that varied from ~20V to 80V. The 80V entered the rotating assembly through slip rings so I couldn't just add an external power supply. Today I could use a cheap switching supply to drop the voltage.

-jsw

Specifically I was confirming the loosely specified time-current curve of the breaker intended to protect the Variac brush, using the short-circuit-tolerant arc welder transformer to boost the current so that 15A from the Variac put 50A through the breaker. The transformer has enough self-inductance that it needed a good fraction of line voltage to reach 50A on the shorted output.

-jsw

Oh. Ratings, yeah.

Cool!

Those were the days. Now, we have noisy, spiky grid.

Ain't Chinglish translations wunnerful? I'm wondering about the package contents: 1 meter, 1 CABLUE.

I have a separate bank account for Paypal, JIC. I don't want to put the full $11.73 balance of my main bank account at risk, either.

I thought you wanted milliamp resolution for your smaller charging texts.

Grok that. If I had my druthers, I'd want to know.

- The list of Obama administration disappointments would take three rolls of toilet paper to record. --BMF

KaBlew?

I have a separate small account at a different bank for card purchases.

The 33.00V, 3.000A / 999.9mA meters are good enough. I have several other types I didn't mention because they are more suited to experimenting or testing component specs than charging batteries.

I was really puzzled when a relay stuck closed and caused the truck battery to drain in a few days. I didn't have a throw-away HF meter in

-jsw

Probably not a referral to a Liberal from the Republik of Kalifornoa.

OK.

I've collected half a dozen throwaway HF meters and dozens of 9LED flashlights over the years with the freebie coupons in the HF ads. As you've discovered, they're perfect for iffy tasks which may or may not be over 10A. The Magic Smoke released is indicative of a higher amperage.

Audible/spark check. Disconnect battery, pull clock fuse/hood light, wait 1 minute, touch cable terminal to battery lug, listen for relay clicks and look for spark. (In your case, listen for lack of relay click. ;) I love the headlight relay in the Toyota, ensuring that I never have a dead battery from leaving the lights on when running in from a stormy day. I first saw those in Japanese import pickups in the '70s and thought they were a great idea. I think Mercenary Bends always had them, too.

(The aforementioned test is more risky nowadays with electronics in every vehicle, but I've never fried anything. And most HIFI gear has ferrite bead protection, too.)

- The list of Obama administration disappointments would take three rolls of toilet paper to record. --BMF

This neatly solves the problem:

It isn't the right choice for an electronic tech's bench meter that may need to measure low milliAmps or microAmps accurately but it's been fine for my vehicle and appliance diagnosis and repair, and solar panel, battery and inverter testing.

I don't like having battery power exposed to accidents on shunts and loads connected with stiff heavy cables, or bare banana plugs. That's why I sprung for the expensive fine strand silicone-insulated wire and lots of Andersons to connect it..

-jsw

I got one of those, and they are pretty good. Another REALLY handy tool is from Autel - MX101 and MX201. They plug into automotive fuse panel, with fuse plugged into them, and they read out the current being drawn by the circuit,The 101 is good for 30 amps with standard ATO fuse, and the 201 good for 20 amps with the mini blade fuses.

I bought the Harbor Freight version, checked that it worked, and fortunately haven't had a problem since then that needed it. I count its cost as insurance.

They misspelled Kablooey. I'd be careful.

Ayup. Sweet. Wish I'd seen those before buying the HF clamp-on.

- The list of Obama administration disappointments would take three rolls of toilet paper to record. --BMF