I recently acquired a Sencore diode substitution box. Model RS-106. One of the devices you can use it for is selenium rectifiers. I bought it as I wanted to satisfy my curiosity about the real voltage drop which occurs on a working selenium rectifier, as the only ones I have come across were bad. Therefore I never could get an idea of the drop across them.

It had it's original box, and the unit is very clean. It contains silicon diodes and numerous power resistors. The resistors are 5% tolerance, and show no signs of ever being abused. I don't think this unit was ever used, as the alligator clips show almost no discoloration, the front panel shows no rub marks or dirt around the switches. This leads me to believe that the resistance measurements I took are accurate, and can be used while replacing seleniums with silicon diodes.

From the user instructions-- "This is not a critical set-up. You merely align the size of the rectifier with the number of plates and you can be sure that the forward resistance is the same as the rectifier that you are substituting for."

The rotary switch for selecting different sizes has a disk marked with plate numbers, and the front panel has markings for different size rectifier plates, so you can have several different combinations for a given switch setting.

I will list the resistance I measured added in series with the silicon diode, and then the various plate size/number of plates pairings.

5 ohms
1 of 2" (or larger) plate

8 ohms
1 of 1.5" OR
6 of 2" plates

12 ohms
1 of 1.2" or
6 of 1.5" or
8 of 2" plates

28 ohms
1 of 1"
6 of 1.2"
8 of 1.5"
12 of 2"

40 ohms
1 of .65"
6 of 1"
8 of 1.2"
12 of 1.5"
18 of 2"

66 ohms
1 of .5"
6 of .65"
8 of 1"
12 of 1.2"
18 of 1.5"

128 ohms
6 of .5"
8 of .65"
12 of 1"
18 of 1.2"

160 ohms
8 of .5"
12 of .65"
18 of 1"

197 ohms
12 of .5"
18 of .65"

258 ohms
18 of .5" plates

Of course, use common sense, and round up to common resistance values. AC outlet voltage is higher now than it was when vintage gear was designed. But this should allow people to have a rough idea of what resistance they can use IF they feel the need. I know in a lot of situations, you can just swap diodes, but I have added resistors when I found voltages were too high. I'll chalk that up to a combination of the forward resistance being lower along with higher AC outlet voltage.

John S.

Thank you for the data you gathered, The one thing that could be used
to simulate the Sencore box is a diode and a decade resistance box.

I've used a Heathkit IN-3117 decade resistance box after a diode rectifier to calculate the resistance needed to provide the correct voltage. Pretty quick way to get the right resistance.

True enough, but something I don't yet possess. So I felt this would allow a quick ballpark guess.

The first one I purchased had been modified. No longer a decade box, it had been extensively changed. I assume it was customized for a specific test application. It went back to the seller.

The second had welded switches. Used as a dummy load?!? Went back the same day.

Which is why I only buy on Ebay from sellers who have a return policy.

Someday.........


John S.

Some "rules of thumb" about selenium rectifiers:

The forward voltage drop should be about 1 volt per plate.

The reverse blocking voltage is about 25 to 30 V rms per plate.

Rectifiers for 117 V operation usually had 5 or 6 plates.

Most all designs used a current-limiting resistor in series with the selenium rectifier, usually 20 ohms or less.

A one inch square plate will handle about 100 mA. avg. DC current.

You may find rectifiers with many plates, but these usually have a tap for use as a center-tapped full-wave, as in a battery charger, or a voltage doubler, common in TV sets of the late 50s.

http://www.w3hwj.com/index_files/RBSelenium2.pdf
http://www.pmillett.com/file_downloads/selenium.pdf
Rich

Selenium rectifiers also have an AC model and their self-capacitance drops with frequency
so the amount of hash that rides in on the line cord may be a bit different if they are subbed.