[Boatanchors] Re: Copper Cap Recitifers, Zener diodes, solid-state replacement, R-390A, etc.

Barry L. Ornitz [email protected]
Thu, 15 May 2003 03:49:59 -0400


I realize that this newsgroup is for Boatanchor radios, but 
the subject of solid-state rectifier replacement has come up 
many times.

>From what I have been able to learn, the "Copper Cap 
Rectifier" is nothing more than solid state diodes with a 
resistor wired in series to approximate the plate resistance 
of the tube it replaces.  This resistance provides the 
equivalent forward voltage drop of the tube AT ONLY ONE 
PARTICULAR FORWARD CURRENT RATING.  At lower currents, this 
combination will produce higher output voltages than the tube 
does and at higher currents it will produce a lower output 
voltage than the tube does

This is easy to understand when you consider that the plate 
current of a vacuum diode is proportional to the plate voltage 
to the 3/2 power (Child's Law).  [To be completely technical, 
the filament must be operating with normal voltages here so as 
to not be temperature limited.  Also at very high currents, 
some deviation from this equation will occur.]

With a solid-state diode, such as silicon or germanium, the 
current through the diode is exponentially proportional to the 
voltage across the diode.  At very low currents, a vacuum tube 
will show a lower voltage drop than a semiconductor diode, but 
the reverse is true at higher currents.

The basic equations, neglecting non-ideal ohmic resistance, are:

Vacuum Tube:            I = a * V^1.5

Semiconductor Diode:    I = Io * exp[ V/(n * Vt) +1]

where:                  I = current through rectifier
                        V = voltage across rectifier
                        a = constant for a particular tube
                       Io = reverse leakage current through 
                            diode
                        n = a constant, usually between 1 
                            and 2 for most semiconductors
                       Vt = k*T/q = thermal voltage, about 
                            25.69 millivolts at room 
                            temperature

I have regressed these models for the 6AL5, the 26Z5 (of 
interest to R-390A folks), and the 5AR4/GZ34 (of interest to 
vacuum tube audiophools), as well as for the 1N4007 and 1N4148 
silicon diodes.  [I also have some models for copper oxide, 
selenium, and germanium rectifiers.]  These models are suitable 
for SPICE simulations and include non-ideal resistance terms.  
Contact me for more details.

I have also plotted the currents for various voltages for the 
above devices.  If you need a copy, contact me.  I can provide 
GIF or JPEG formats.

I am not sure if the so-called inrush protection of the 
"Copper Cap Rectifiers" is anything in addition to the series 
resistor they use to simulate the plate resistance.  I doubt 
if it is from what I have read.

I would certainly not use one for replacing a 6AL5 if the tube 
is used as a signal detector or for automatic volume control 
detection.  The sensitivity will be less.  But occasionally a 
6AL5 may have been used as a bias rectifier and a solid-state 
replacement would be fine here.

The 0A2 is a voltage regulator rated at 150 volts.  It is not 
a diode but rather a gas discharge tube.  It can be replaced 
with a 10 watt Zener diode like the 1N3011 (properly heat 
sinked, of course) which will provide better regulation than 
the tube.

To correct a small point, WBob wrote:

> Well that isn't too true.. All silicon diodes will zener at 
> about 2 or 3 x their voltage ratings. They don't make very 
> good zeners and it is an uncontrolled parameter so even two 
> diodes of the same name and manufacturerer will not have the 
> same zener voltage, but they will all zener somewhere.

Technically, this is not Zener operation which is a low 
voltage phenomena (less than 5 to 7 volts).  What he is 
talking about here is avalanche breakdown.  All modern silicon 
diodes made after about 1980 or so are controlled avalanche 
rectifiers, meaning that avalanche breakdown will not destroy 
the diode as long as the diode does not overheat.  This is why 
paralleling modern diodes with resistors and capacitors is no 
longer necessary when building high voltage rectifier stacks.

Glen Zook, K9STH, then wrote:

> Actually, a germanium diode can be used under certain
> circumstances as an approximately 0.3 volt zener and a
> silicon diode can be used as an approximately 0.7 volt
> zener.

I have written about this before in this mail reflector.  What 
you are doing here is using the forward voltage drop of the 
diode.  As the above equations show, this voltage is not very 
constant with changing current levels.  In the 1N4000 series 
of diodes, for example, the following data is representative:

     Current, mA       Voltage, V
          0.1            0.449
          1.0            0.565
         10.0            0.682
        100.0            0.800
       1000.0            0.931

In my opinion, it is almost always beneficial to replace 
vacuum tube power rectifiers in old gear with modern solid-
state silicon diodes.  The 10 to 15 watts you save in filament 
power puts less strain on the power transformers.  These are 
some of the most expensive components to replace, and are also 
very sensitive to temperature.  Their lifetime approximately 
halves for every 13 degrees F rise in temperature.

If you are concerned about the higher voltages resulting with 
silicon diodes, you can add series resistance like that found 
internally in the "Copper Top" products.  This generates heat, 
of course, but no more than the tube would have with its plate 
resistance (and you still save the filament heat).

You will get "instant" high voltage, of course, and without 
the other tubes warmed up and drawing current, the voltage 
will be higher than normal for a few seconds.  If you have 
properly sized electrolytic capacitors, however, this is 
normally well within their surge ratings.  If not, go up to 
the next higher voltage capacitor.  This is much cheaper than 
replacing a transformer.  Also applying high voltage to the 
plates of receiving (and typical ham transmitting) tubes 
before the filament is warmed up is not a problem.  The so-
called cathode stripping some people worry about does not 
occur at voltages below about 10 kilovolts.

     73,  Barry L. Ornitz     WA4VZQ     [email protected]

              Copyright 2003 B. L. Ornitz
      Do not post to other groups without permission.