Replacing Selenium rectifiers

I'm using a 50 amp diode bridge in my battery charger.... works and is a lot cheaper then $50 to replace :-)

You can use diodes, but you need to make sure that they are cooled properly. Voltage drop on a diode bridge is about 1.4 volts, so, for example, at 50 amps it will need to dissipate 70 watts. I hope that I got my math right.

Also selenium rectifiers are not that expensive on ebay. I sold one a while ago.

i

Ignoramus17926 fired this volley in news:qK6dnRzW4fprqZbOnZ2dnUVZ snipped-for-privacy@giganews.com:

Ig, there's no good reason to re-install seleniums unless it's physically impossible to arrange for proper heat-sinking of silicon diodes.

In addition to having a higher forward resistance than silicon diodes, selenium rectifiers are somewhat 'leaky' in the reverse-biased direction, yielding a much less efficient rectification. There are lots of stud- mount silicons out there that usually suit the purpose perfectly.

However, on modern consumer battery chargers, selenium rectifiers usually have their own finned sinks, and are 'air mounted'on flying leads, or simply mechanically mounted on a cathode-common pallet (not a heat sink, itself), while silicon diodes are often bare-chip affairs that have been sweat-brazed directly to a cathode-side heat sink. So in either case, some mechanical work will be necessary to replace them.

LLoyd

This does work, but you may need a power resistor in series, depending on the overall design, which may have depended on the high resistance seleniums to limit surge currents.

Joe Gwinn

I figured, and I have some big Copper plate scraps about 6"x6"x1/4" and there's plenty of room.

How would I value it?

Exactly, selenium rectifiers have heat sinking built in. For diode bridges, you need to find or buy, and mount, a heatsink.

Hmm I was going to say you could measure the voltage drop with and without current flowing and then figure out the resistance. (But that assumes you have working Selenium rectifiers.) What's the maximum current?

George H.

I didn't do any math :-) Just added a 220 VAC fan to blow on the diodes... and O Yes, made a nice shiny aluminum plate to mount the bridge on :-)

Big silicon rectifiers -- mounted to a heat sink about the area of the original selenium plates. If you have single plate selenium rectifiers then 20V PIV is quite adequate -- and I doubt that you will find any smaller than 50 PIV so you should be fine there.

Current rating probably twice the maximum shown on the current meter just to be safe.

Are the two rectifiers on a single stack, with just three terminals? If so -- and depending on the polarity -- it might be easier to select reverse polarity diodes so both can be mounted on the same plate without insulation. Normal silicon rectifiers put the cathode on the mounting stud, so with two of them mounted to a plate, the plate will be the positive output. If the Selenium rectifier has the negative in common instead (and the center tap of the transformer connected to positive output -- perhaps through the meter), then you want diodes with the anode to the stud instead. The proper selection here allows you to mount them both on the same plate (heat sink) without needing to get the insulating kit to keep them from shorting together.

Hmm ... going to the Digikey web page, there are rectifiers with a maximum PIV as low as 5V, so let's select ... a range of voltages and currents:

So -- from the search (among others) we have these two:

S25BRGN-ND $3.16350 -- 100V 25A reverse polarity S25BGN-ND $3.60750 -- 100V 25A standard polarity

Either of these should do well for the charger (unless the current meter goes higher than 25A).

No idea why the prices are to six figures after the decimal point. :-)

Or -- there are modular half-bridges in various formats, and one of these may well do for it all.

B483B-2-ND B483B-2T-ND

Way more voltage than you need (400 V PIV) and plenty of current (35A). Mounts to the panel with two screws, three spade terminals, common cathode '+'. No common anode '-' version shown. Price not shown. (They say "call".) So -- the individual diodes above are a safer bet.

Enjoy, DoN.

I agree.

And -- when they die -- they *stink*. :-)

"selenium"? You mean "silicon", don't you? I can't imagine a selenium rectifier being used, in a modern battery charger -- consumer or industrial.

Really cheap construction.

Find or buy -- or *make*.

Enjoy, DoN.

Do you have one half of the selenium rectifiers still alive? Perhaps put the maximum current through the rectifier (based on the meter on the charger) and measure the voltage across it (assuming that you have a bench power supply capable of that).

Or -- take the open-circuit voltage from the charger (easy to measure with the rectifiers replaced), and the maximum current (from the meter on the charger) -- and divide the voltage by the current to get the resistance. Then multiply the voltage by the current to get the wattage needed in the resistor. (You can also measure the resistance of the secondary windings in the transformer, and subtract that from the total resistance calculated.) This resistor might keep from cooking the transformer.

Good Luck, DoN.

My first recommendation is to tell us more about the battery charger. Maybe the model number would let people look on the internet and see the specs. Or maybe just post the specs. Does the transformer have a center tapped secondary?

My second recommendation is to replace the selenium diodes with silicon diodes.

My third recommendation is to look on Ebay. Most all common diodes are made overseas, so it is just a question of who imports them to the states. On Ebay a 50 amp bridge is $.99 with free shipping. And unless you are talking about a charger for a forklift, that is probably as big as you need.

Dan

It would be a pretty low value. A 0.25 ohm resistor would drop about 12 volts at 50 amps. For limiting surge currents you would not need more than about 5 or 10 watt resistor. Steady state like shorting the leads would b e 600 watts provided the transformer could supply 50 amps continuous.

Dan

Does 125 amp boost count?

One on each side, single wire, about 6"x6"

[ ... ] [ ... ]

Only one wire going to each Selenium rectifier? Then the other side is bolted direct to the chassis, perhaps? Then determine which direction the rectifiers were pointed (positive to chassis or negative to chassis) and pick the proper silicon rectifier from those I found (oops -- and already snipped from this article) -- or bigger ones depending on what current you need to handle. I would not expect a Craftsman one to be that high a current. :-)

O.K. Recovered the part numbers (from Digi-Key):

======================================================================

======================================================================

Good Luck, DoN.

If it's old enough to use selenium rectifiers the Net may not help. Is this a lead-acid charger or for NiCads which require close control of charge termination?

If lead-acid, how does it reduce the current as the voltage rises? This is what a change to Silicon diodes may affect.

I recently fixed some old golf cart chargers that had ferroresonant regulation, a cheap and simple circuit that depends on clever design and component choice to work properly. Fortunately I studied magnetic amplifiers in the Army.

If it's ferroresonant there will be a large AC capacitor connected to a separate winding on the power transformer. They regulate the AC voltage from the transformer, not the DC from the rectifier.

jsw

Only if it passes through the diodes.

In my stash, acquired at a flea market, not for sale:

I couldn't resist rectifiers with leads that look like winch cable.

A megger like this can check large surplus rectifiers for reverse leakage and breakdown.

I paid $39.13 for one last June. It's a ripoff of an old Biddle megger.

jsw