Tech Square

Technical and speculative discussion of amps, cables and other topics.

Return to Tech Square

Model for Selenium Rectifier?

68.13.64.77
Posted on October 5, 2009 at 17:51:34
Lee of Omaha
Dealer

Posts: 449
Location: Omaha NE
Joined: September 8, 2006
 I've looked around a bit without success for a mathematical model for selenium rectifiers. If it's a good model, I think forward drop per plate would do. I think that if the drop per plate is pretty constant, like it is for silicon rectifiers, then a good replacement would be a silicon rectifier in series with an appropriate zener; the silicon diode to do the rectification and the zener to drop the rest of the voltage (Vz = [selenium drop - 0.7V).

Anyone?

RE: Model for Selenium Rectifier?, posted on October 13, 2009 at 15:03:34
jerem69
Audiophile

Posts: 3
Location: DFW
Joined: October 13, 2009
 As far as the -vacuum tube- rectifier replacements, they were just a silicon diode with a series ballast resistor. It worked OK if the current was in the right ballpark. So, for a selenium rectifier replacement, you might try that same model (resistor in series) for a relatively stable current draw application such as a receiver or a pre-amp's power supply. For a more dynamic model, I like the zener idea, however, they would have to be capable of handling peak current/power. Any other folks have a good yet simple solution?

RE: Model for Selenium Rectifier?, posted on October 13, 2009 at 15:09:43
jerem69
Audiophile

Posts: 3
Location: DFW
Joined: October 13, 2009
 BTW, the reverse-biased condition would probably take some protection of the zeners. Perhaps placing schottky diodes in parallel with them in the opposite direction of the main rectifiers? Now that looks like a cluster-#$%@ on the schematic, but, try it and see! Good luck. Like I said, any others with ideas?

RE: Model for Selenium Rectifier?, posted on October 13, 2009 at 15:37:14
Lee of Omaha
Dealer

Posts: 449
Location: Omaha NE
Joined: September 8, 2006
 I don't think that's necessary. It seems to me that in the reverse direction (which is the forward, usually unused direction for the zener) the zener would attempt to conduct, so it would have low forward resistance, and the silicon rectifier would have the bulk of the voltage drop. Does that sound correct or am I missing something?

RE: Model for Selenium Rectifier?, posted on October 14, 2009 at 07:14:56
jerem69
Audiophile

Posts: 3
Location: DFW
Joined: October 13, 2009
 In "looking" at the dynamics of reverse bias, the capacitance of the components may allow the zeners to see spikes which are far greater than their rated forward bias current. It should be transient, but.. When designing high-voltage power supplies, many years past, there weren't many options as far as high-voltage FETs. So we used to stack them in series. But, they had to be "balanced" with shunt capacitors to help evenly distribute the charge when they were switched open. I'm erring on the side of caution. I try to think of longevity when designing. On a side note, many folks replace selenium rectifiers with a simple series resistor/diode combination. I found this on the net: http://www.w3hwj.com/index_files/RBSelenium2.pdf
Good Article, posted on October 14, 2009 at 08:47:18
Lee of Omaha
Dealer

Posts: 449
Location: Omaha NE
Joined: September 8, 2006
 I had come across and read this article in my research, and it is the best single article I had seen. However, it advocates using a dropping resistor, and I'd like to use a circuit that better models the behavior (except failure modes!) of the original selenium rectifier.

Also in my research, I came across one article that said seleniums gradually increase their forward resistance for about a year (though the article didn't say whether that was under continuous use) then stabilize or plateau until they begin to fail. When you designed equipment using seleniums, did you design for initial conditions or for greater forward drop than the devices initially have?

I like your idea about a small capacitor in parallel with the zener.

What would you think of a series combination of silicon, zener, and resistor to most accurately model a selenium? What I'm after is an engineered drop in replacement that would require no thought on the part of the installer or consideration for the circuit. That sort of product is only possible if the replacement nearly exactly mimics the behavior of the original over the original's operating space of voltage and current.

Page processed in 0.034 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2009, Audio Asylum, all rights reserved ] [ Go Green ] [ Take5 ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]