Thread: Leak Point One
View Single Post
  #4   Report Post  
Posted to sci.electronics.repair
[email protected] pfjw@aol.com is offline
external usenet poster
 
Posts: 1,625
Default Leak Point One

On Jul 11, 5:13 am, dadaderk wrote:

I only did some minor modifications in the beginning (I bridged and replaced
some capacitors). First time I put the power on in a room where it was
very cold and the GZ34 went.


I hope I am not understanding you correctly... "I bridged and replaced
some capacitors)... One should _NEVER_ bridge a cap if the goal is to
replace a cap that has failed. The old cap *MUST* be removed from the
circuit and the new cap put in its place.

If the goal is to add additional capacitance, it is much better
practice to remove the 'small' cap and replace it with a new 'larger'
cap. First, after 20 years the existing electrolytics may be tired,
and second, modern caps are typically much smaller than their older
cousins such that space is not an issue. Parallel caps _of the same
type and value_ *are* a legitimate and often used practice, however.

If you are replacing non-electrolytics, the same general logic
obtains.

In some circuits, excessive capacitance can cause excessive B+, so be
careful there.

It is poor and risky practice to run a system with the rectifier in
place and not the load... that is, remove all the tubes but the
rectifier.

Lastly (and long): The 5AR4/GZ34 is a very rugged tube unless you are
using some of the Chinese versions. Those tend to fail, often
spectacularly and often also taking much else with them. You have not
supplied and apparently do not have a schematic. Therefore you
probably do not know the correct voltages that should be coming off
the transformer and also what the correct B+ should be. Verify that
there is no more than 5V at the filament pins on the 5AR4. After that,
check all your work _VERY_ carefully. Polarity of the caps, solder-
blobs, poor connections, and so forth. Apply power through an ammeter,
preferably through a variac. Typically a tube rectifier will not pass
DC until it reaches ~70%+ of filament voltage, and the 5AR4 is more
like 75-80%. But when the needle on the ammeter starts to rise, you
will be able to learn much.

With the absolute final caveat that you are working in the dark and
without a schematic. So everything you are learning is more-or-less
guesswork.

Peter Wieck
Wyncote, PA