Rectifier question

[Phil_S]

I'm one of those guys who looks to lower cost the hard way. So, I'm looking at 6X4/EZ90 rectifiers. (I know, four diodes cost $0.40; that's not the hard way.) The spec on this tube is RMSin=325-0-325, IRect=70, Cout=16.

If I use two in parallel, do I get to double the IRect and Cout ratings?

This would make a big difference it the application of this tube.

[markh]

When paralleling valve diodes, you have to derate them a bit, as the lower impedance one of the pair will conduct more current. Valve diodes are a bit more forgiving in this respect (vs. SS) due to their relatively high internal resistance.

--mark

[Phil_S]

The amp I'm thinking about is a Champ deviant and it's not expected to go above 70mA. The PT is 250-0250 so RMSin won't be an issue. I was more concerned about Cout. The first cap is 20uf. Are you saying that I'll be OK in this department? Thanks.

[Michael]

Quoting from our friend, Chemilec:

"One way to help fix this is to put a Low Value, Series Resistor in front of Each Diode. The voltage drop across this series resistor tends to Equil Out the differences in the diodes. The resistance of each series resistor should provide a voltage drop, somewhere around 0.3 to 0.6 volts at the current draw of the particular circuit it is used in."

http://www3.telus.net/chemelec/Projects ... rallel.htm

See also:

http://www.powerdesigners.com/InfoWeb/d ... iodes.shtm

[Phil_S]

Quoting from our friend, Chemilec:
"One way to help fix this is to put a Low Value, Series Resistor in front of Each Diode.

OK, I get this and thank you for the link. What is meant by "in front" for a rectifier tube? Place a 100 ohm R on each plate to soften the inrush from the power transformer? I imagine these could get pretty hot. Do we need 5W sandblock R's here?

Sorry for asking what may seem obvious. This is new turf for me. I like to put a belt and suspenders on it before I try it. It's less exciting that way and more rewarding.

[Michael]

Quoting from our friend, Chemilec:
"One way to help fix this is to put a Low Value, Series Resistor in front of Each Diode.

OK, I get this and thank you for the link. What is meant by "in front" for a rectifier tube? Place a 100 ohm R on each plate to soften the inrush from the power transformer? I imagine these could get pretty hot. Do we need 5W sandblock R's here?

Sorry for asking what may seem obvious. This is new turf for me. I like to put a belt and suspenders on it before I try it. It's less exciting that way and more rewarding.

Sorry; I missed the fact that there were tube "diodes". I'm not really sure "sharing resistors" are required, as valve diodes don't really act like diodes except for the single direction conduction. They are much more like JFETs.

To answer your questions, yes, you would put the resistor between the AC and the plates.

The size of the resistors depends on the current load of your amp. Figure an 18 watt would pull something like 100 mA or so at idle, and maybe 120 at full pull, or close to 0.36 watts per resistor (two per tube). 2 watt resistors would be plenty big with some wiggle room for surges. They would need to be mounted where they can breath a bit.

Review and correction welcome.

[tunghaichuan]

Take a look at the schematic linked at the top of this page:

http://users.rcn.com/fiddler.interport/HF86.HTM

I used to own one of these, the Eico HF-86 power amp. It uses two 6CA4/EZ81 rectifiers in parallel in the power supply.

tung

[Michael]

Tung is spot on. They way they do it is just run one side of the AC to both cathodes on one recto and the other side to both cathodes of the other.

Man, those guys are brilliant!

Here is a schematic from my 1959 HF-81, which is one sweet amp for LPs.

[Phil_S]

Boy oh boy, you really blew this one up! The dual rectifier in parallel is pure genius, but I don't understand the use of the filament winding. Anyhow, the PT I've got only has one filament winding, so it won't be possible to implement the Eico model as drawn.

Would it work (anyone tried it?) to just take the cathode outputs and tie them together in parallel without using the filament winding?

BTW, we are at odds over plate and cathode. One the 6CA4, I'd call pins 1 and 7 plates, and pin 3 cathode.

[Michael]

Not a big deal. Just tie your anodes together on each rectifier. Then run pins 3 in parallel.

The "filament" is just a heater. So if your PT has enough juice in its PT heater supply to power both rectifiers and all your other tubes, you are golden.