Shopping for a variac

[tkamiya]

It is getting dark so I had to cut the test short.

I'm using 4 amp as load max because of 500VA and input is 120V.  Also, fuse is 5 amp.  I did go up to 4.5 amp or so and nothing worth mentioning happened.

500 watts + 150 watts bulbs made 4.0 amps at 66.6V.
During 5 minutes run, all temperatures stayed nominal.
Core is 82F
Bottom screw is 78F
Side is 84F (where the contact cone is)
Top is 80F

Ambient is 76F with slight breeze.
Each temperature, including core, fluctuated quite a bit with presence of some wind.  (above figure is when air is still)
Nothing melted
No sparks
No abnormal heating of cone was observed
Turn-it-up-and-down caused not even a small spark

Now, here's something...  a FAINT a VERY FAINT burning smell was identified but it was unclear where it was coming from.  It could be the bulb as they were dusty.  Tried to smell around the stat area, but I couldn't find the source.

Another thing to note, temperature was significantly low compared to day time test.  That means temperature RISE (delta) may be critical.  Doing the same test in a sweaty warm warehouse with no breeze may cause something.  (remember, this is Florida)

Long time test is probably warranted, but I really can't sit outside for that long, and with children around, leaving it unattended is just plain irresponsible.  As we push this harder and harder, something will happen at some point.  Thanks to someone's post of specs, it is obvious this is meant for a light duty job.  It was never meant to be used continuously at max rating in harsh environment.

I'll leave you with a photo of a test supervisor. 

[Electro Fan]

I'll keep the setup so if anyone finds better ways to test with what I have, please let me know.

Ironically, you need a second Variac to be able to vary your bulb-based load bank to test the first Variac. 

You don't happen to have any giant wirewound adjustable resistors or anything like that, eh? 

Challenging a champion finder to find something?

A second variac?  I found one!  On Amazon!!

There was never a doubt 

[rsjsouza]

I didn't read the whole thread, but years ago I bought a Mastech 2kVA and I still have it. It is a workhorse!

https://www.eevblog.com/forum/buysellwanted/mastechvolteq-variac-comments-on-quality/

[cdev]

That old equipment photographed so well almost has a "fetishistic" quality to it now. Its almost erotic.

Fwiw, my variac looks a lot like this plus two meters and two fuses and it looked like it just came new (inside and out) from the factory.

https://www.ebay.com/p/1127332453?iid=372842692756

Call the owner in that eBay link above that I first posted above (I don't know anything about the unit or seller in this link just above in this post).


Edit: Ok, here you go
http://www.shitintogold.com/VARIAC.php
http://www.shitintogold.com/

Another Edit: Owner's name is David.  Looks like he might have a different/new URL; I think the company name was / maybe still is DRCO.  I might have picked a different URL but you're buying a variac not a PR firm.

Yet another edit:
Tell David you don't want (can't afford / can't justify) the brand new cosmetics - you just want his least expensive reliable variac and see what he says.  His pricing reflects the fact that he likes to make products that operate like and look like they came new from GenRad decades ago - but I think he can cut the price down if you go for the electronics over the case condition.  Tell him that you are going to report back to EEVblog if he takes good care of you (that's his only MO anyway) and therefore this is his chance to tap into a new market - but first you need a reliable variac that you can afford.  If his least expensive unit is too expensive for your budget keep shopping elsewhere - but I think he will try to take good care of you.

Good luck!

And another PS,

Those Powerstats like Shock posted look good.  Ask David if you go for a Powerstat vs. a GenRad if that would help keep the budget in line.  For all I know the Powerstats might be the same or more or less; the good thing with David is that he just likes making and refurbing good products and he's big into variacs.  I don't think he has any axe to grind whether it's Powerstat, GenRad, Staco, or any other manufacturer, model, or design.  It's like asking Michelangelo for something in marble - but it needs to be an affordable statue.

[Shock]

I took covers off mine and made some photo and measurements.

Take a look at your variacs chassis earth/ground wire (green/yellow) and see if it's making proper physical metal to metal contact to all the chassis metalwork parts. It looks like the lug is sandwiched between two painted surfaces and the front panel is just screwed in. I see it also has the wrong mains cable coloring so the plug and socket could be wired and switched backwards for the US. The soldered connections look a bit suspect as well, did they pass them through the lugs or is it just tacked in place?

[tkamiya]

Good eyes!

Yes, it's a lug sandwiched between painted surfaces.  Inner side metal is tapped, and lug is making an imperfect contact.  I actually can align it perfectly concentric and tighten, and have no ground connection to the outer casings.  The lug is soldered on the back side, and properly.  All other solder connections are suspects

US color is black for live, white for neutral, and green or bare for ground.  Color coding does not appear to be consistent.  It is obviously not US or Japanese standard colors.  Maybe EU colors, mostly?  Polarity is correct.  I visually followed it.  Personally, I am not too strict on coloring wires, if it doesn't confuse myself.

I found one case of awful soldering.  This is at the main switch.  Multiple wires are there because this is its main neutral connection.  Those are DEEP cracks, gentlemen!  It appears to be cracks that formed after soldering was completed.  Looks like it was pulled a part under force.  Very dull in color indicating cold solder job, too.  This will need to be redone.

Also note, black char marks on insulation.  I think these jackets are vinyl.  I would expect at least PVC.  (or cloth covered rubber on old ones)

[tkamiya]

So....  what would YOU do to this if you owned this?  Obviously, it isn't completely useless or junk.  It functions.  Also, obviously, in current state, it's unsafe.

I would at least replace all wires with properly rated ones and redo soldering jobs.  Ground connection will be upgraded.  I'd use it at 50% duty rating.  Remove the meter and replace as it is pretty much useless.  I really would like to re-case it.  Even mine, a smaller one, is very flimsy.  I will also take it completely apart and make sure ground to live isolation is good and sound.  One thing I don't have an ability to check is winding/live to case isolation.  As this is an auto-transformer, it's critical that casing never becomes live.

[BravoV]

So....  what would YOU do to this if you owned this?  Obviously, it isn't completely useless or junk.  It functions.  Also, obviously, in current state, it's unsafe.

I would at least replace all wires with properly rated ones and redo soldering jobs.  Ground connection will be upgraded.  I'd use it at 50% duty rating.  Remove the meter and replace as it is pretty much useless.  I really would like to re-case it.  Even mine, a smaller one, is very flimsy.  I will also take it completely apart and make sure ground to live isolation is good and sound.  One thing I don't have an ability to check is winding/live to case isolation.  As this is an auto-transformer, it's critical that casing never becomes live.

Yep, tear it apart and inspect, as have friend own similar model, and the locking and securing mechanism of the heavy toroid is not properly done, and quite dangerous.

Probably they didn't fasten enough and the toroid maybe shifted or moved not centered inside the round case/chassis during shipment.

[tkamiya]

I took it apart completely.  There were few surprises.  I'm sorry, no pictures.

The shaft that starts in the knob, goes though the casing, the core (the transformer) and finally secured at bottom is not insulated anywhere.  There are fiber washers but nothing is protecting from touching the case. 

The cone shaped brush is secured to a brass plate which is secured to plastic that is secured to the shaft.  A wire goes from brush to outside after wrapping around the shaft twice.

What this means is that if anything energized touches the shaft, the whole thing will become energized.  The only ground connection to the casing is iffy.  (see my last post).  The core itself is secured by a large piece of rubber and metal disk like any toroidal transformer would be.

I would like some redundancy in protection here.  For example, use of silicone wire with abrasion protection, and some kind of insulating sleeve on the main shaft where it goes through the core.

Another surprise is, since this is a transformer, there is a beginning of the winding and the end of the winding.  This is hiding inside the core.  Inside, there is a piece of fiber washer, probably M4.  Each end of the wire goes through this once and each is loosely twisted to itself few times.  There is no means of positively securing this from comping apart.  Also, beginning and end of the wire is within 3mm of each other.  There is maximum voltage differential here.  Here, I would like to see potting, clamps, or some type of secure method.  Also, more distance.

I gave it more thought on if this is a junk or not.  It's not a junk because it functions.  It is a product that is made just enough to pass the requirement.  But every measure of cost cutting has been made to meet the price point.  That's why we couldn't find clear faults or catastrophic failures.  But, there is no redundancy for safety.  There is no robustness.  Safety seemed to have took a second seat to the price.  It works.  It sells.  It makes profit.  That's the objective.

To me, I was asking the wrong question.  I should have been asking, is it safe?  is it robust?  is it quality?  Is it junk is purely a personal opinion based on personal standard.  I have mine but it's different from everyone on this board.

That concludes my involvement in this experiment.  I've cut the cord, and placed pieces in the recycle bin.  That was a fun and educational experience for me.  Thanks everyone for the encouragement.

[Shock]

Looks like you didn't keep it but I'll leave this here anyway as I wrote it out.

As BravoV mentioned you need to look closely at the mechanical side of things. When I brought my Powerstat variac secondhand I had to do this, they aren't inherently safe as people tend to mess with them or they get abused or knocked around in shipping.

In your case the soldered connections need to be mechanically attached to the lug plus soldered so if the joint breaks it won't come off. Some strain relief like glue lined heat shrink and using cable ties to prevent wires flopping about will help prevent failures. The front panel needs a protective earth strap connected to it. Polarity of the plug needs to be verified to the switch and fuse.

I'm not worried about the wiring color but on cheap Chinese products you need to check it all thoroughly. See if it's a solid copper strand power cable and what current the cable is rated for, 240V cables have a typically lower current rating than 120V. I suspect the fuse may not be rated properly either, can also be fake of course.

Long and narrow brushes are the optimal shape for variacs, it allows a more reliable electrical and mechanical connection to the winding and prevents excessive interwinding shorts. The more windings shorted the less current handling and greater unnecessary heat. Something like that anyway.

[Ian.M]

tkamiya: "I've cut the cord, and placed pieces in the recycle bin."
That says it all.   Not worth fixing up or impractical to make safe enough to use, so pretty much the dictionary definition of 'junk'. The only way it could be worse and still function would be if it had an intentionally live shaft extending up into the knob!

I was surprised to hear it didn't even have a slip-ring to connect to the brush.  Pity you didn't take pictures during the final disassembly - they would have made a nice horror-show to point cheap-skate newbs at . . .

[Cubdriver]

tkamiya - thank you for the autopsy report.  That definitely sounds like it was designed to the lowest price point.

-Pat

[Electro Detective]

so basically as mentioned before here and other threads      these are another TooHungLow item UNSAFE and or DANGEROUS to run out of the box without checking the item first,
as well as be very familiar with variable auto transformers, so you know what to look for and sort out.
Yes they look simple enough to wire up (famous last words?) and even simpler to stuff up,
especially if it wasn't done right at the factory, usually a polarity mix up and a suspect earth ground to chassis connection.

fwiw the stepdown transformers from TooHungLow I've seen and had to make safe/R, but not any more ever again! the asses that import such junk to Australia should be tar and feathered on Youtube,
and off to prison for trading in dangerous goods, and false/forged/fake quality control check labeling.       

I blame the dirtbag sellers for dumping cheap half baked risk laden electronic rubbish on to cash strapped newbs, bargain hunters, and unsuspecting cheapskates,
not the clueless whipped workers in China   

How hard can it be to show the impoverished slave workers on the sweatshoppe assembly lines how to do it properly,
and pay them 1c more per hour, so they don't complain about the extra effort to pay attention and understand what it is they are assembling and why some details are IMPORTANT,
and sellers can flog a safer decent cheap crap item, and stitch the end connedsumer for $10 more

[engrguy42]

Thanks much, tkamiya. Greatly appreciated. As I suspected, there were no results that would be sufficient for the "it's cheap, red, Chinese junk" folks to ever change their minds. They'll always find something. People believe what they want to believe.

For a bunch of hobbyists to push back on something like this is hard to comprehend. Folks who spend hours and days building and restoring things can't accept a 10 minute review of wiring and maybe a little re-soldering in order to save $200? Tech experts who don't know enough to do a few tweaks (which may be unnecessary in the first place)? But hell, you'll buy a 70 year old something on eBay and spend hours cleaning and restoring it and that's fine. Or spend days designing,prototyping,soldering and testing components on PCB's. But instead of "wow, look what I found on Amazon for only $50!!! Only took a few minutes of tweaking and it works great!!", we get "Well, *I* own a Powerstat/Variac/whatever and anything that isn't a Powerstat is junk". 

Anyway, nothing new here. BTW, my ATX power supply is working great.   

[Electro Detective]

..the "it's cheap, red, Chinese junk" folks to ever change their minds. They'll always find something. People believe what they want to believe.

I always 'find something' wrong with these fire red russian roulette devices, and other stepdown transformer junk with wonky windings and internals from Wee R Sweatshop Inc. 

in fact it's a winning bet there's ALWAYS some drama with these, or will be in the future if not sorted 

I 'want to believe' what I've seen for real, too many times, and rectified,
so users don't zap their attached gear and or themselves
or their curious bench hopping cats having a sniff

[Gyro]

Thanks much, tkamiya. Greatly appreciated. As I suspected, there were no results that would be sufficient for the "it's cheap, red, Chinese junk" folks to ever change their minds. They'll always find something. People believe what they want to believe.

I hardly think that having a brand new off the shelf item and needing to dismantle and modify it to make it safe to use, counts as "They'll always find something". 

What about all the purchasers who don't know that they need to do that (assuming they have the skills)?

[Shock]

As I suspected, there were no results that would be sufficient for the "it's cheap, red, Chinese junk" folks to ever change their minds.

Go buy one then if you think it's great quality. You deserve to own one of these.

[engrguy42]

For those who might also be considering one of these cheap Chinese pieces of junk that might kill your cat, I did find a 2000VA, 20 A model on Amazon that looks identical to the others. But this comes with a digital voltage display rather than the fairly useless analog one. I'm sure the display will require some reference adjustment or whatever. And this one is only $83 with free shipping (maybe that's an Amazon Prime thing?). And that's good cuz it's 15 pounds.

And this one has a spec that says "Temperature Rise: <60℃". I'm still on the fence about whether I'll get one (since honestly I don't really need one), but I would like to learn more about them, and actually test one myself. So I can judge based on facts.

Which, BTW, I'd encourage others to consider, so that you can judge based on facts and your own tests, rather than unsupported, vague generalities.

Kinda like I did with the ATX power supplies. In spite of being bombarded by unsupported, vague generalities and name calling, I actually built one (modified my 460 watt ATX supply) and now it works great. I use it all the time. Although someone said those ATX supplies are bad because a bug can crawl in and zap the power supply, so I'm still worrying about that. 

[ArthurDent]

Post 145 listing cheap variac specs:

I've had my eye on a 20A version of the cheap red Chinese variacs, and happened on a guy's video of the exact same model. If anyone is curious about
Oh, and here's what they give for specs on Amazon. Interesting they mention 8 working hours:

Max. 2000VA
Plug Type: US Plug
Max output : 20 Amps
Input Voltage: AC 110V/60Hz
Output Voltage: AC 0-130V/60Hz
Material: Iron (shell) & Copper Coil
Humidity : ≤ 5%
Working Hours: 8 hours
Condition Temperature:<40℃
Insulated Resistance: >10MΩ
Withstand Voltage: 2KV/1 minute

...so is 8 working hours the life expectancy? 

[ArthurDent]

I took covers off mine and made some photo and measurements.

Take a look at your variacs chassis earth/ground wire (green/yellow) and see if it's making proper physical metal to metal contact to all the chassis metalwork parts. It looks like the lug is sandwiched between two painted surfaces and the front panel is just screwed in. I see it also has the wrong mains cable coloring so the plug and socket could be wired and switched backwards for the US. The soldered connections look a bit suspect as well, did they pass them through the lugs or is it just tacked in place?

There is a big difference between an older well-made variac and a new one built to be as cheap as possible.

For one thing, my Powerstat was designed for industrial use. The set screws that hold the shaft in place can be loosened to allow the base of the Powerstat to be mounted on a panel with the shaft sticking out what you’d think of as the bottom of the unit, not the top. This shaft is not connected to any live parts of the Powerstat. To allow this mounting, what was the zero output position when viewed from the top, is now max output when viewed from the bottom because of the reversed rotation when the variac is flipped like this. This means that the taps on the winding have to be symmetrical so the unit can be used with either rotation and these connections are shown in one of the attached photos.

The narrow width of the brushes on a variac has been mentioned several times but the brush has to be wide enough to bridge at least 2 turns to prevent arcing. The graphite brush is resistive and doesn’t directly short any turns. You can see how the brush in mine ‘bridges’, not shorts, turns on the coil. If you count the number of turns in a variac and divide the input voltage by the number of turns you will find that there is a fraction of a volt between turns that will be distributed across the brush surface. That is why you don’t see metal brushs used in variacs because they would actually short out adjacent turns.

On this Powerstat all the wires connected to the coil are brought out to a panel where your connections are made. You can see in the photo below where the connections are symmetrical to allow reverse rotation depending on how the Powerstat is mounted.

Where this was an older bench variac, it had a 2 wire cord and 2 wire outlet which I changed. As I mentioned in a previous post I added a second fuse holder on the output wiper to properly protect it. As has been mentioned previously you can also put a smaller rated fuse in the output if you only have a small load you want to limit the maximum current to. I also added a 3mm LED to give me a visual indication that the power was on.  I had to be careful when I added this extra ‘stuff’ but there was just enough room to allow everything to fit nicely.

If you check the rating plate you will see that this is actually a 230 VAC 3.5 Amp Powerstat but it works fine on 120 VAC, perhaps even running a little cooler.   

[Cubdriver]

Post 145 listing cheap variac specs:

I've had my eye on a 20A version of the cheap red Chinese variacs, and happened on a guy's video of the exact same model. If anyone is curious about
Oh, and here's what they give for specs on Amazon. Interesting they mention 8 working hours:

Max. 2000VA
Plug Type: US Plug
Max output : 20 Amps
Input Voltage: AC 110V/60Hz
Output Voltage: AC 0-130V/60Hz
Material: Iron (shell) & Copper Coil
Humidity : ≤ 5%
Working Hours: 8 hours
Condition Temperature:<40℃
Insulated Resistance: >10MΩ
Withstand Voltage: 2KV/1 minute

...so is 8 working hours the life expectancy? 

I was wondering what that meant, too.  For something that small, it seems an odd number for a duty cycle - it would certainly reach max temp in far less than 8 hours, and if it WERE temp related it should be indicated and have some operational limit below which it can run continuously.  Things like arc welders are typically rated at their max output with a ten minute window (say, 20% at max power,  meaning you can weld for two minutes out of every ten if you're running it full bitch, and a continuous rating that might be around 40% of max, at which you can operate continuously, and a curve between the two for the intermediate values - perhaps at 80% power you can go 40% - 4 minutes out of every ten.)

I guess that's asking too much from a red scariac that you need to rework right out of the box.

-Pat

[Electro Detective]

Hey, I spotted the 'Working Hours: 8 hours' thing ages ago 

Note the other specs are just as informative for discerning spendthrifts like me

So if no real new, used or abused variable AC units were around anymore on Earth,
I would chew the bullet and go for a red 20 amp unit, if 5 continuous amps is actually what I needed

opening up and securing the internal snafus first..

and buy 2 units if anticipating a long 16 hour session at the workbench
or if one decides on a life change as a toaster

 

[tkamiya]

It really sounds like a case of something written in Chinese was auto-translated or manually translated by someone without technical knowledge.  I used to have to back translate and re-translate manuals originated from Japan.

It probably means it only works from 9am to 5pm with mandatory lunch break at noon.  Overtime is at reduced rate only.      I think, somehow, it means non-continuous duty.  But I have no idea what the de-rating curve is.  I wonder if someone can get a hold of the original Chinese manual?  I know we have members here from China....

Humidity of less than 5%?  I have no idea what this means.  These things are usually fine at non-condensating environment.  It may have meant 95% or less. 

[cdev]

I basically have two unused, "new old stock" US made, very small variacs - just the raw variacs, no case. They are just the raw autotransformer, no additional hardware. I forget the brand, they are buried deep in a closet. Probably the wattage is okay for some very small electronics, but nothing of any size. I will get them out and photograph the winding of the one that is open and post it here.

[Cubdriver]

I basically have two unused, "new old stock" US made, very small variacs - just the raw variacs, no case. They are just the raw autotransformer, no additional hardware. I forget the brand, they are buried deep in a closet. Probably the wattage is okay for some very small electronics, but nothing of any size. I will get them out and photograph the winding of the one that is open and post it here.

I'll save you the trouble.  I'm pretty sure I have a few little ones fairly readily accessible.  Could be a fun macro project.

-Pat