Has anybody seen diodes like this?

[Torben]

Hi all,

I'm working on repairing an old automotive battery charger--I'll post more on that in the Repairing Consumer Electronics section. This thing has something in it I haven't seen before, either in person or in pictures. After sketching out the circuit it seems that these must be diodes, and they test like diodes. I've never seen this kind of diode before so if someone could point me to (or give me ) information on these, that would be great. (I'm talking about the things which look like binding posts, mounted to the plate on the back wall of the charger and soldered to the outer taps of the transformer's secondary.)




Cheers,

Torben

[Edit: Just after posting the above, I discovered that the negative battery clamp lead had come disconnected. So I'm going to put that back together tomorrow and test again--that's probably the problem. Should have checked that first, of course. I'm still interested in whether the devices in the photo above are diodes and if not, what they are.]

 

[mneary]

If it's more than 50 years old, they would probably be selenium rectifiers.

 

[Torben]

If it's more than 50 years old, they would probably be selenium rectifiers.

I don't think it's fifty years old, but it might be thirty. This is the front panel:

The CSA label "LR 1978Q" might mean it was certified in 1978--or it might mean something else entirely, for all I know.

Anyway, if those are in fact diodes, then the schematic ends up looking something like this:


[Edit: They would be D3 and D4--which seem to indicate that they must be diodes. I'll search on selenium diodes--thanks!]

Cheers,

Torben

[Edit 2: Fixed the schematic.]

 

[ericgibbs]

hi Torben,
Give the rectifiers a quick sniff, Selenium has a very distinctive smell.

They look like Selenium diodes.

 

[Torben]

hi Torben,
Give the rectifiers a quick sniff, Selenium has a very distinctive smell.

They look like Selenium diodes.

Thanks Eric! Having both you and mneary confirm it helps a lot.

For some reason, I think finding diodes that look like this is just about the coolest thing in the world.


Best regards,

Torben

 

[Hero999]

Does it still work?

I knew it used bi-phase with the 0V connected to chassis just by looking at the picture, I bet reverse engineering the schematic was easy.

 

[Vizier87]

hi Torben,
Give the rectifiers a quick sniff, Selenium has a very distinctive smell.

They look like Selenium diodes.

I knew that garlic contains selenium.. does it smell like garlic?

 

[ericgibbs]

I knew that garlic contains selenium.. does it smell like garlic?

Viz
No, it has a 'dark brown smell' when the rectifier gets hot.

Ask Uncle Nigel.

 

[Hero999]

I don't know, lots of foods are high in selenium: Brazil nuts, kidney, seaweed and shellfish such as lobster crab and mussels are all pretty good sources.

Perhaps the dark brown smell is what you'd have if you mashed all the aforementioned foods up and left them to rot.

 

[Nigel Goodwin]

I concur, antique selenium rectifiers - and they smell REALLY bad when they fail

 

[Boncuk]

Viz
No, it has a 'dark brown smell' when the rectifier gets hot.

Ask Uncle Nigel.

It gets close to latrine smell if you burn it up.

Selenium rectifiers were commonly used in battery chargers limiting charging current automatically due to their high impedance.

I once replaced a selenium rectifier with a silicon rectifier and the result was a burning transformer.

Rectifier means "Gleichrichter" in German language. I usually called selenium type rectifiers "Gleich riecht er".

Boncuk

 

[Nigel Goodwin]

I once replaced a selenium rectifier with a silicon rectifier and the result was a burning transformer.

Tut, Tut!

I used to replace loads of metal rectifiers (selenium) in record players - the technique was to take out the huge metal rectifier, and replace with a piece of tag strip, fitted with a silicon rectifier and a wirewound resistor.

As you say, you can't just still a silicon diode in!.

 

[shortbus=]

I've never (knowingly)seen a single plate selenium diode like those. Usually in a stack of alternating sized squares.

 

[Vizier87]

Cause it is ancient, according to the uncles and grandpas here.
we'll be the ones who tell younger guys 30 years later "I once pricked my finger cause I mishandled a IN4148 (with an ominous picture attached) !" then the younger guys would say something like "whoa! that is a diode? That thing is HUGE!!"

Grandpa Eric: ok I got it.... urgh.
Hans: I got it too: Immediately he smells!!!

 

[crutschow]

I've never (knowingly)seen a single plate selenium diode like those. Usually in a stack of alternating sized squares.

The number of plates determines the reverse voltage rating. According to wikipedia, each plate is good for about 20V piv.

 

[Torben]

Nope, no smell. The diodes check out, and although it would probably be a good idea to replace them I'm going to see how long they hang in there. I think that on paper the single stack should be good enough given that the secondary should stay below that. I did find this on chemelec's site, which seems very similar (even the board layout matches): **broken link removed** and it says this charger cannot be used as a power supply, and must have a battery attached to function. My next step is to work out how to test it properly. I suspect the SCR and I'm going to pull that and test it at some point.



Torben

 

[mneary]

Contradicting myself, if it's new enough to have an SCR I would not expect the rectifier to be selenium.

 

[Torben]

Does it still work?

Nope. That's why I'm trying to fix it.

I knew it used bi-phase with the 0V connected to chassis just by looking at the picture, I bet reverse engineering the schematic was easy.

I assume when you say "bi-phase with the 0V connected to chassis" that it's another way of saying "full-wave centre-tapped rectifier"? I didn't see that at first because it didn't occur to me that these things might be diodes. Then when I started to draw it out in my head it became pretty obvious so I checked them and they behaved diode-like, but not exactly like the diodes I'm used to. At any rate, after working out that part of the schematic it seemed that they pretty much had to be diodes. For a second I wondered why they didn't go with a full-wave bridge but then I realized it wouldn't fit.

Reverse-engineering it would probably be very easy. I did it in an evening, so I expect someone more experienced would manage it in much less time.


Regards,

Torben

 

[Torben]

Contradicting myself, if it's new enough to have an SCR I would not expect the rectifier to be selenium.

OK, now you're just going to confuse me, man. I just thought I had this figured out, too.

I'm not too sure about the history of the parts in Canada in the 70's but this thing has a 2N6504 and those things in the picture. In the context of the schematic I don't see what else they could be.


Regards,

Torben