August, 2010

I've done some electroplating--mainly zinc and copper and a little bit of nickel.  I've made do with a little 1-amp variable power supply I built probably 30 years ago.  It works fine for small parts, but for larger parts, it just doesn't have the juice to do the job.  Also, I wanted to try aluminum anodizing, and this process can take higher voltages than my supply could manage.  Considering all this, I decided to build a larger variable supply.

Now for serious plating or anodizing, many people recommend a variable constant current supply.  (Sounds like an oxymoron, doesn't it?  It just means that when you adjust it, you are setting the current, and once set, the supply maintains that current over a range of loads.)  An adjustable constant current supply of the size I needed is not a trivial design exercize, and while I probably have the skills to do it, I was really looking for something I could get into service quicker.

Even an adjustable voltage supply of this size (0-30 volts @ 15 Amps) is not a snap to design from scratch, so I headed down a simpler path.  I decided to build a simple fixed voltage linear supply, and then feed it with a variable AC voltage.  This is not a new idea, and seemed to be the simplest and quickest approach.

A variable AC voltage is easily accomplished by using a variable autotransformer, often generically called a Variac.  I didn't have one, but Ebay had many, and soon I had one of theirs for around $30.  It was an old used Calrad unit, reportedly in good working order, rated at 500 VA--just about right for my needs.



The unit was weighty, like a lot of older stuff, but the dial seemed to bind on part of its rotation.  Fearing the worst, I opened it up.  I found that the heavy transformer unit had shifted forcefully enough, probably in shipping, to break one of its locating tabs, and bend another.  This put the core off center, bending the power switch, and making the shaft bind.



I fixed this pretty easily, and got the mechanism working smoothly.    

My only other gripe with the unit is that, being of a certain age, it did not have a 3-prong grounded power cord or recepticle.  I fixed that while I had it open.

I buttoned it back up, and it worked fine.



So mucdh for the variable AC source.  Now to the linear AC-DC conversion.

This 20+ pound bad boy is a Stancor 2012 power transformer, probably made in the 1960s.  It was in an old rack mnount power supply that I found in the closet of an apartment I rented a long time ago.  I've been moving it for decades.  The transformer is a universal type, with a lot of winding taps to give a wide range of secondary voltages.  I was able to find a data sheet for it on the web.



On the left are 1N249 40 amp diodes from the same found power supply the transformer came from.  I'll re-use the mica insulating washers.  The other picture is the parts I bought new.  The old power supply had some large electrolytics, but I worried about their age.  The new one is more compact, anyway.

 

Rather than use a hobby box to mount everything in, I decided to make a custom aluminum case that was the same height as the Variac.  I got a piece of 6061-0 sheet aluminum.  I believe the -0 means that is not tempered, and so is soft enough to bend easily.  The last picture proves once again that JB Weld can be a life saver.

             

Even though calculations showed that the enclosure itself provided enough surface area to keep the diodes cool, I added some heat sink fins just for insurance.  I also made some new insulating spacers for the diode studs in the thicker panel.  I painted the box with a gray hammertone to match the Variac, but masked where the heatsinks would mount.  The heatsinks are attached with aluminum rivets, using liberal heat sink compound to improve heat transfer.  Last picture shows plastic caps on the stud nuts.  They are electrically hot, so I don't want them shorting to anything.

   
   

Wiring is straightforward.  Its a standard bridge rectifier with a capacitive filter.  The AC is fused at 4 amps, the DC at 15.  Ammeter with builtin shunt is wired in series, voltmeter across the output.  Power switch is in the primary circuit.  There is a bleed resistor across the capacitor to drain charge when the supply is turned off.

 

Final setup seems to work OK.  Can't wait to try it.