How do you test a Selenium Rectifier?

[Allodoxaphobia]

On Wed, 5 Dec 2018 11:37:17 -0800 (PST), pfjw@aol.com wrote:

Someone else (whose attribution was stripped) wrote:

I remember selenuim rectifiers being used in car battery chargers,
capable of charging at 50 amps.

Did not most battery chargers use copper-oxide rectifiers? As I
understand it, they are both more rugged and more weather resistant
than selenium. On the other hand, they are very similar in appearance.

I still have, and use, an "antique", Big, Honkin' 50 amp 12/6V battery
charger -- and it uses selenium rectifiers. It's out in the bitterly
cold, dark, detached garage ATM -- or I would quote from it's metal name
tag. A hand-me-down from my father ... and maybe even, his father.

In the past I was able to start my old, dilapidated, hardly-any-
compression Jeep with no onboard battery!

Still have the charger -- not the Jeep.

Jonesy
--
Marvin L Jones | Marvin | W3DHJ.net | linux
38.238N 104.547W | @ jonz.net | Jonesy | FreeBSD
* Killfiling google & XXXXbanter.com: jonz.net/ng.htm

 

[Phil Allison]

Dave M wrote:

Actually, they ARE heat sinks. Selenium rectifiers were much less efficient
than silicon, or even germanuim rectifiers, with a forward voltage drop of
about 1V per cell (plate).

** But silicon diodes like 1N4004s or 1N4007s you see everywhere have a similar voltage drop - about 1V at 1 amp.

https://www.onsemi.com/pub/Collateral/1N4001-D.PDF

That means that if your rectifier has 16 cells,
then the rectifier, as a whole, will have AT LEAST 16 volts of forward
voltage drop. Multiply that by the current through the rectifier, and you
have the number of watts that has to be dissipated, thus, the need for the
heat sink plates.

** But the *REAL* reason is the low reverse voltage capability - 25V per diode rather than 400V or even 1000 volts.

FYI: Silicon diodes are remarkable devices - a finger nail size, 4 diode bridge is adequate for a 1kW DC supply using only the PCB foil as a heatsink.

https://au.element14.com/vishay/w10g-e4-51/no-of-phases-single-phase/dp/1497580?st=1.5 amp bridge

Think there is another reason too in that Selenium diodes cannot be allowed to run as hot as Silicon, or their life span is drastically shortened.


.... Phil

 

On Wednesday, 5 December 2018 17:18:06 UTC, mako...@yahoo.com wrote:

Each plate is only good for about
25 volts ,


wow I always thought those were just heat sinks

I seriously would not have guessed that I would learn something new about
selenium rectifiers today.

mark

The big plates are heatsinks. Further in are the smaller selenium plate rectifiers.


NT

 

On Sun, 2 Dec 2018 10:57:07 -0000 (UTC), Cursitor Doom
<curd@notformail.com> wrote:

On Sun, 02 Dec 2018 01:26:41 -0600, Fox's Mercantile wrote:

A DMM won't read correctly. No enough voltage compliance on the current
source to read ohms correctly.

True, but it *might* give some rough indication, like if the thing has
somehow shorted out. I have a few of those lying around somewhere; I'll
measure 'em with a DVM just out of curiosity and see if I get readings
like those reported by tabbs.


Just replace it with a 1N4007 diode and about 100 ohms series resistor.

Not kosher if you want to keep the internals original! Many collectors
are extremely fussy about that.

I plan to leave it on the chassis for looks, but replace it.
My voltage is real low, so it seems it's weak or worse....

 

[MOP CAP]

In a selenium rectifier the big plates are BOTH rectifiers and heat sinks!
Seen the hugh vacuum chambers in which the were made.
CP

 

On Thursday, 6 December 2018 05:43:11 UTC, tub...@myshop.com wrote:

On Sun, 2 Dec 2018 10:57:07 -0000 (UTC), Cursitor Doom
curd@notformail.com> wrote:
On Sun, 02 Dec 2018 01:26:41 -0600, Fox's Mercantile wrote:

A DMM won't read correctly. No enough voltage compliance on the current
source to read ohms correctly.

True, but it *might* give some rough indication, like if the thing has
somehow shorted out. I have a few of those lying around somewhere; I'll
measure 'em with a DVM just out of curiosity and see if I get readings
like those reported by tabbs.


Just replace it with a 1N4007 diode and about 100 ohms series resistor.

Not kosher if you want to keep the internals original! Many collectors
are extremely fussy about that.

I plan to leave it on the chassis for looks, but replace it.
My voltage is real low, so it seems it's weak or worse....

and about to go into toxic fumes mode.

 

[Big Bad Bob]

On 12/01/18 22:30, tubeguy@myshop.com wrote:

How do you test a Selenium Rectifier?

Yes, I am aware that they should be replaced, but on preliminary tests
of this 1956 Zenith tube AM-FM Radio, I want to know if it's working. My
digital multimeter shows 1meg in one direction, 2.5 meg the other way.
NOT DISCONNECTED FROM THE CIRCUIT. To me, that seems bad, but I have not
worked with one of these in many years, and never had to replace one in
the old days. My meter was climbing, probably from the filter caps.
during this test.

forward/back resistance with a basic ohm meter should be around 100:1 or
better as I recall, and you might have to use a resistance bridge of
some kind (or other measuring circuit) with a voltage potential of
several volts for the higher voltage selenium rectifiers.

Selenium rectifiers might have a voltage drop of several volts during
normal operation. So maybe a test circuit would work better...

You could drive normal AC power through it, into a (safe) resistive load
(no capacitors), and then look at the downstream waveform on an o-scope.
It should be half wave AC at a reasonable peak voltage, into a
reasonable resistive load (let's say a 10k or 20k several-watt resistor).

I assume it's for a radio that runs on 110V [or maybe 220/240V for EU
and AU radios]. So you'll see a half-wave 50/60 cycle waveform that
peaks at around 1.4 times the input voltage.

If you look at it with an o-scope and see too much on the negative
cycle, you'll know it's bad. Otherwise it should be ok to use it as
long as the resulting output voltage is correct.


--
(aka 'Bombastic Bob' in case you wondered)

'Feeling with my fingers, and thinking with my brain' - me

 

[Big Bad Bob]

On 12/05/18 21:42, tubeguy@myshop.com wrote:

On Sun, 2 Dec 2018 10:57:07 -0000 (UTC), Cursitor Doom
curd@notformail.com> wrote:

On Sun, 02 Dec 2018 01:26:41 -0600, Fox's Mercantile wrote:

A DMM won't read correctly. No enough voltage compliance on the current
source to read ohms correctly.

True, but it *might* give some rough indication, like if the thing has
somehow shorted out. I have a few of those lying around somewhere; I'll
measure 'em with a DVM just out of curiosity and see if I get readings
like those reported by tabbs.


Just replace it with a 1N4007 diode and about 100 ohms series resistor.

Not kosher if you want to keep the internals original! Many collectors
are extremely fussy about that.

I plan to leave it on the chassis for looks, but replace it.
My voltage is real low, so it seems it's weak or worse....

right they'll probably short out when they fail and blow the filter caps.

You could try a resistive load (10k-20k, several watt resistor) and
disconnect it from the filter capacitors when you do it. Then look at
the waveform. (I suggested this in a different post already, just
summarizing)


--
(aka 'Bombastic Bob' in case you wondered)

'Feeling with my fingers, and thinking with my brain' - me