Repairing Old School Power Supply?

[SLK001]

Okay, it seems like everyone has contributed to this thread but the original poster. Still waiting for pictures of the guts of his supply.

[Gazz]

I'm sorry folks - things got a bit crazy. I will try and get pics of the inside tomorrow.

[PStechPaul]

Sometimes an old school piece of equipment or tool has a great deal of value and is worth repairing or restoring. Such an item is often well suited to its purpose, and may be more rugged and easier to repair than a more modern counterpart. Unless you really need the smaller size, lighter weight, or additional features of a more modern device, I would recommend keeping, fixing, and using what you already have.

[Paul Alciatore]

Actually their Voltage drop is about 1 Volt per layer and several layers are usually needed to build up enough reverse Voltage for AC Voltages around 100 Volts. Even for a 12 V DC supply, the AC input to the rectifier was about 18 to 20 Volts Peak due to the RMS factor. So even 12 V supplies often needed selenium rectifiers with two or more junctions. Silicon diodes have about a 0.7 Volt drop per junction and normally only one junction is needed for Voltages up to 500 or 800 VAC, at the very least. So the selenium rectifier never had a lower loss than the silicon diode.

In my youth I built projects with selenium rectifiers. But they very rapidly disappeared when silicon diodes became widely available. By the 1980s they became increasingly hard to get. This was because the silicon diodes were clearly superior. Oh, and a lot smaller and definitely less expensive too. Win, WIN, WIN!

That would do it. Probably not more efficient though, it's hard to beat a transformer with a voltage dropping regulator system.

We need some pics of that old thang to see what's inside, just for curiosity if nothing else.

Why all the hate for selenium rectifiers? They work pretty well, and on a per-junction basis, they are fairly low-loss, comparable to silicon. It's only when you need a number of junctions that a silicon diode really beats them.

There now are better solutions for most useages, but they were very practical and inexpensive semiconductors at the time. They still can be for low voltages and high currents.

[darryl]

Active rectifiers- now there's something interesting. Seems to me a good way to simplify a power supply if it needs to have regulation of sorts. Seems the active elements, whether mosfets or transistors, would need a full reverse voltage capability- without using diodes. I need to look into this more-

[J Tiers]

Actually their Voltage drop is about 1 Volt per layer and several layers are usually needed to build up enough reverse Voltage for AC Voltages around 100 Volts. Even for a 12 V DC supply, the AC input to the rectifier was about 18 to 20 Volts Peak due to the RMS factor. So even 12 V supplies often needed selenium rectifiers with two or more junctions. Silicon diodes have about a 0.7 Volt drop per junction and normally only one junction is needed for Voltages up to 500 or 800 VAC, at the very least. So the selenium rectifier never had a lower loss than the silicon diode.

In my youth I built projects with selenium rectifiers. But they very rapidly disappeared when silicon diodes became widely available. By the 1980s they became increasingly hard to get. This was because the silicon diodes were clearly superior. Oh, and a lot smaller and definitely less expensive too. Win, WIN, WIN!

Selenium was normally 20 to 25V withstand per cell. So for low volts, plating type supplies, single cells. A 25 V withstand is close for 12V, where the peak voltage may be only a little higher.

But for low voltage plating supplies at 2 to 7 volts? Sure.

Silicon theoretically is 0.7V. You will not see that in a rectifier at rated current, it will be in the 1 volt range. Sometimes more.

So, for low voltages, very comparable, as I said. Once you get several layers, no, selenium is not competitive with a single silicon junction.

[Gazz]

I finally got some pictures for you folks. Please understand that when I asked for help, I was thinking someone would provide me with a link to some kind of online tutorial and not actual personal help! I do appreciate the responses!
Apparently something has changed since my first post. I do not use a picture hosting site and all my images are on my hard drive. When I try to upload pictures now, I keep getting a message that the image source url is missing. Thanks for your help anyway.

Thanks BCRider for that help! I didn't get that option when I wrote my reply or when I click on the picture icon. At least I didn't see it.

[Dave C]

All this discussion of selenium rectifiers has got me worried about my mig welder's health. I has a big (4") square selenium rectifier in it and it has been replaced once.

[BCRider]

I finally got some pictures for you folks. Please understand that when I asked for help, I was thinking someone would provide me with a link to some kind of online tutorial and not actual personal help! I do appreciate the responses!
Apparently something has changed since my first post. I do not use a picture hosting site and all my images are on my hard drive. When I try to upload pictures now, I keep getting a message that the image source url is missing. Thanks for your help anyway.

Gazz, the pics are on your computer. So you need to click on the box below the text editor box that says "Upload Attachments". The Upload from URL only applies to pictures already out on the web.

There are a couple of steps you need to follow.

1. Click on "Upload Attachments". There are a few different file types you can upload to these posts. Picture files in a few formats are among those options.
2. A regular file selector box will open up. Navigate to where you have your pictures and use Ctl - left mouse click to pick up to four separate pictures. They will upload and from there you can click the link you want shown beside the thumbnail. I like to use "Medium". This keeps the image size reasonable for the post but folks can still click on it and see the full size in a new tab pane.
3. Note that the image will be inserted where the cursor is located. And it'll move text around as needed. So be sure it goes where you want it to go.

[darryl]

That's a pretty straightforward power supply. Looks like there's a power switch on the ac line side and not much else. The transformer either is getting powered or it's not. Do the strip of steel test against the transformer- if it vibrates, the power input circuit is working. You might even hear it humming. The first thing to find out is if it's drawing any juice.

[Bob Engelhardt]

My best guess (but I'm not familiar with a variac secondary wired that way):



Edit: no, that's not right - the variac wiper wouldn't be like that.

[J Tiers]

That's a fonkay variac. First one I have ever seen like it.

Looks like two sections, so just as the above schemo except each diode has a separate variac contact.

Looks like one big junction for each rectifier. Close to the edge, but that's how stuff was often done back then.

Fuse, switch, transformer and done for the primary.

[The Artful Bodger]

So that is interesting, just what does the knob on the front panel do? It looks like there is some sort of swinging arm so maybe contacts on the secondary, or is something happening in the magnetic circuit?

[Noitoen]

No Variac in those pictures. Those parts on the side are some kind of diodes, selenium, maybe. Bob's drawing is pretty much what I see.

[Noitoen]

So that is interesting, just what does the knob on the front panel do? It looks like there is some sort of swinging arm so maybe contacts on the secondary, or is something happening in the magnetic circuit?

Maybe hidden behind the transformer there is some high powered rheostat