How do you test a Selenium Rectifier?

[tub...@myshop.com]

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.

[Fox's Mercantile]

A DMM won't read correctly. No enough voltage compliance on the current
source to read ohms correctly.
Just replace it with a 1N4007 diode and about 100 ohms series resistor.


--
"I am a river to my people."
Jeff-1.0
WA6FWi
http:foxsmercantile.com

[tabb...@gmail.com]

Selenium stacks have high Vf & high reverse leakage. DMMs don't apply enough voltage. Probably the simplest way to test is to power it up. They fail by Vf increasing, causing overheating & lots of toxic selenium fumes, so best not tested indoors.

Replacement is probably a wise option. But add a resistor & fuse, seleniums act as current limits. There are lots of old scopes still running with selenium EHT rectifiers in them.


NT

[Cursitor Doom]

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.



--
This message may be freely reproduced without limit or charge only via
the Usenet protocol. Reproduction in whole or part through other
protocols, whether for profit or not, is conditional upon a charge of
GBP10.00 per reproduction. Publication in this manner via non-Usenet
protocols constitutes acceptance of this condition.

[tabb...@gmail.com]

On Sunday, 2 December 2018 10:57:10 UTC, Cursitor Doom 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 replaced some parts on my 1934 set, but you'd never know by looking inside. The new parts are inside the old, which are repacked exactly as original. With selenium you could possible hide the diode somewhere not noticeable, keeping the selenium in place.


NT

[Adrian Caspersz]

On 02/12/2018 13:03, tabb...@gmail.com wrote:

> I replaced some parts on my 1934 set, but you'd never know by looking inside. The new parts are inside the old, which are repacked exactly as original. With selenium you could possible hide the diode somewhere not noticeable, keeping the selenium in place.

To maintain the authentic smell?

--
Adrian C

[Ralph Mowery]

In article <80v60e9fe9680kjb2...@4ax.com>,
tub...@myshop.com says...

The same way as with any other diode. However the selenium drop about a
volt for each plate that is on it. Each plate is only good for about
25 volts , so you may have 5 to 10 volts to overcome. Therefore most
meters will not have enough voltage to over come that drop.


Get a voltage source of around 10 volts or more (maybe even a 9 volt
battery will be enough) use your meter on the voltage scale and put hte
battery , diode , voltmeter in series. Then turn the voltage or diode
around and measure again. There should be a couple of volts dropped one
way and many more (almost all) the other.

[Cursitor Doom]

On Sun, 02 Dec 2018 05:03:03 -0800, tabbypurr wrote:

> I replaced some parts on my 1934 set, but you'd never know by looking
> inside. The new parts are inside the old, which are repacked exactly as
> original. With selenium you could possible hide the diode somewhere not
> noticeable, keeping the selenium in place.

Yes, I'm aware of this practice, but for some super-fussy types, it's
still not good enough.

[Cursitor Doom]

On Sun, 02 Dec 2018 10:12:01 -0500, Ralph Mowery wrote:

> Get a voltage source of around 10 volts or more (maybe even a 9 volt
> battery will be enough) use your meter on the voltage scale and put hte
> battery , diode , voltmeter in series. Then turn the voltage or diode
> around and measure again. There should be a couple of volts dropped one
> way and many more (almost all) the other.

+1

[Phil Hobbs]

And watch out for confetti. ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC / Hobbs ElectroOptics
Optics, Electro-optics, Photonics, Analog Electronics
Briarcliff Manor NY 10510

http://electrooptical.net
http://hobbs-eo.com

[Phil Allison]

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

** Most DMMs have separate ohms and diode tests ranges, the ohms ranges
are for " in circuit" resistor testing and will not make a good Silicon or Selenium diode conduct.

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

** Really ? Only the wimpiest rectifier has 100 ohms series resistance, like a Germanium detector diode.

The OP's Selenium rectifier powers an AM / FM radio,( minus less the tube filaments ) so ought to be good to half an amp - so maybe try 22 ohms @ 5watt.


.... Phil

[tabb...@gmail.com]

On Sunday, 2 December 2018 13:27:47 UTC, Adrian Caspersz wrote:

> On 02/12/2018 13:03, tabbypurr wrote:

> > I replaced some parts on my 1934 set, but you'd never know by looking inside. The new parts are inside the old, which are repacked exactly as original. With selenium you could possible hide the diode somewhere not noticeable, keeping the selenium in place.
>
> To maintain the authentic smell?

obviously there won't be any smell from a selenium rectifier that isn't seeing power.


NT

[tabb...@gmail.com]

On Sunday, 2 December 2018 16:15:37 UTC, Cursitor Doom wrote:
> On Sun, 02 Dec 2018 05:03:03 -0800, tabbypurr wrote:

> > I replaced some parts on my 1934 set, but you'd never know by looking
> > inside. The new parts are inside the old, which are repacked exactly as
> > original. With selenium you could possible hide the diode somewhere not
> > noticeable, keeping the selenium in place.
>
> Yes, I'm aware of this practice, but for some super-fussy types, it's
> still not good enough.

I like my kit to work. If some folk want theirs dead, I guess that's their call. There was no chance of it working without repair. FWIW having old kit that doesn't work is missing much of the experience.

With the selenium rectifier, you could still use it if it still works, but its toxicity is a problem that needs addressing somehow. That gets difficult, though not impossible.


NT

[MOP CAP]

Seleniums have a shelf life. By their very nature the go bad over
time. Thus any selenium out there is bad as they have not [at least in
the US} been made in a shelf life time.
CP

[Mike Coon]

In article <373715b3-d4e5-467b...@googlegroups.com>,
tabb...@gmail.com says...
> ... & lots of toxic selenium fumes, so best not tested indoors.

And, IIRC, a horrible smell of cabbage!

Mike.

[John-Del]

Cabbage that's already been digested...

[mako...@yahoo.com]

> 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

[Dave M]

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). 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.

I remember selenuim rectifiers being used in car battery chargers, capable
of charging at 50 amps. The plates, as I remember, were about 8" square,
and had to have a fan inside to cool the rectifier since there was so much
heat to be removed.

Cheers,
Dave M

[peterw...@gmail.com]

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.

Peter Wieck
Melrose Park, PA

[Ralph Mowery]

In article <G8ydnbcn5sPXgZXB...@giganews.com>,
dgmi...@mediacombb.net says...

>
> 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). 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.
>
>

Yes, the actual plates are heat sinks, but I was loose with my words and
just using that as a refference as to how many sections there are. That
is each plate coresponds to a diode type junction in series. So if you
count 5 plates it is similar to having 5 dioides in series. Each one
good for about 25 volts of inverse voltage and dropping about a volt.
That would make around 5 volts forward drop and good for about 125 piv.

Using the term plate was to me just a quick way of estimating how much
voltage drop and PIV the stack could handle.

[Allodoxaphobia]

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

[tabb...@gmail.com]

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


NT

[tub...@myshop.com]

On Sun, 2 Dec 2018 10:57:07 -0000 (UTC), Cursitor Doom
<cu...@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

[tabb...@gmail.com]

and about to go into toxic fumes mode.

[Big Bad Bob]

On 12/01/18 22:30, tub...@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]

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)