Current limiting resistor for valve rectifier

AVForums

Help Support AVForums:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

El Sid

AVForums Super Veteran
*
AV Forums Supporter
Joined
May 26, 2016
Messages
1,992
Reaction score
452
Location
Jhb
I have finally started to get going on building the electronics of my EL84 concept amp, and am starting with the power supply.

I am using a GZ34 rectifier, and the transformer is giving me 350-0-350V. In keeping with the original configuration of the old amp being "upcycled", I am putting in a pi filter, using the original 10H, 180mA, 200Ω inductor/choke, with capacitors yet to be determined (although I do have some 30μF motor start caps from the original that I could use). 60μF is the maximum in the datasheet.

My question relates to current limiting resistors between the transformer and the anodes. Most designs I've looked at neither have them between the tf and anodes nor between cathode and first capacitor in the pi. Is this just taking a chance, or might the transformer in these cases have sufficient resistance?

Anyway, according to the Amperex 5AR4/GZ34 datasheet, the minimum limiting resistor is 100Ω for a capacitor input filter at 350V. Measuring the Engels transformer, the resistance of the primary winding is 8Ω, secondaries 40Ω. This gives about 55Ω for winding resistance, and another 45Ω at least is required. If I put it on the cathode, assuming max current = 180mA (choke limit)  then it needs to be minimum 1.5W, so 3-5W to be safe. The Valve Wizard gives a good digestible rundown of power supplies, and if that's too lightweight for you, there is a monumental treatise on the topic in Morgan Jones.
fullwave3.jpg


I found a good idea on The Immortal Amplifier Mod, which suggests putting two diodes (1N4007 good enough) between the transformer and rectifier anodes, which will protect the caps (and transformer itself) if the rectifier fails short circuit, putting AC onto the polarised filter caps and then problems occuring. Unless someone shows me it's a bad idea I'll do that - the diodes are 40c each.
200807_techviews_1.gif


 
Top