[R-390] Re: Parts Selection

[Drew Papanek]

Perry wrote:

>I champion 1% metal film resistor replacement over carbon comps for the 
>following reasons. First >they are decades more stable, have a far lower 
>noise factor, are smaller than,and are one third the >price of carbon 
>comps.

I like the "smaller than" part.  It makes repairs and restorations much 
easier.  Noise is a consideration only where the resistor is in the signal 
path.  In the R-390A RF amplifier stage, for example, the cathode resistor 
is bypassed with a capacitor (two, actually) and hence is not in the signal 
path. Carbon or metal film would make no difference.

There are very few R-390A signal path locations where carbon composition 
resistors are found. One that comes to mind is in the crystal filter 
circuit.  Another is the plate load resistor for the 12AU7 audio amp stage.  
Signal level at those points is high enough that metal film would make no 
difference in noise performance.

One caveat in any resistor replacement is to consider the voltage flashover 
rating. It is not of much concern in today's solid state low voltage 
circuits, but it is of great importance in tube equipment.  Mouser carries 
Vishay/Dale metal films; the CCF55 series is rated at 300V maximum and the 
larger CCF60 series is good to 500V.

Low level audio amp stage plate load resistors would benefit from 
appropriately voltage-rated metal film resistor replacement.  In that  
application there is 100V or more across a resistor of a couple hundred K or 
so and carbon comps fail frequently, drifting way up to sometimes over a meg 
or even becoming intermittent, causing "crackles" in the audio.

>Also some "A"’s 2.2K resistors are underrated so with metal films you can 
>double up in the same >space or opt for the 3/4 watt rated Vishay series.

There is no harm in uprating in many circuits but isolation resistors in 
plate circuit is not the place to do it.  The 2.2K  resistors in the B+ 
feeds to various R-390A stages also serve as fuses.  There always is the 
possibility of a shorted tube which will burn out the isolation resistor and 
limit current to a little less than 100 mA while doing so.  The radio's B+ 
fuse(s) might not blow fast enough to prevent damage to other components.

Metal film resisors will withstand much greater overload than carbon comps.  
In most circuit locations that would be good, but not where the resistor 
serves as a fuse.  The aforementioned 2.2K resistors should be replaced with 
carbon composition types.  I do not know how well carbon FILM types would 
work in a fuse role.  Since carbon films are more reliable, cheaper, and 
more readily available than carbon comps, I think some experimentation would 
be in order.  <Evil Grin> Time to get out the suicide cord!  Mine is the 
UL-rated version with 3 alligator clips for grounding :-).  Time to let the 
smoke out of a few resistors, 110VAC style!

As far as underrated 2.2K resistors goes, a fuse protects best when operated 
close to its limit. Maybe some of the 1/2 watt 2.2K's should actually be 1/4 
watt? (I haven't taken time to calculate.)

The resistance value of those 2.2K plate circuit isolation resistors is not 
too critical.  If used in a circuit where the current were, say, 5mA the 
voltage drop would be about 10V. If the resistor drifted to 3K, the drop 
would be 15V.  That 5 volt difference is small compared to changes due to 
tube characteristic variations and line voltage changes.  I say leave them 
alone unless they're WAY off...  There is always risk of damage (breaking a 
terminal off an irreplaceable coil, for example) when reworking and I don't 
think the small gain justifies the risk.

>Barring tube or cap shorts 50 years of operation for metal films is just 
>warm up time.

I agree.  Metal films everywhere except where burnout (fuse) characteristics 
are a concern.

Drew

_________________________________________________________________
>From predictions to trailers, check out the MSN Entertainment Guide to the 
Academy Awards® 
http://movies.msn.com/movies/oscars2007/?icid=ncoscartagline1