[Amps] Replacing bias string diodes in Harbach rectifier

[Jim Thomson]

Date: Wed, 11 Aug 2010 03:52:24 -0000
From: "Bill VanAlstyne W5WVO" <w5wvo at cybermesa.net>
Subject: Re: [Amps] Replacing bias string diodes in Harbach rectifier
board

Thanks for the ideas, Bill. OK, so let's say I want to clean this amp up 
once and for all. No more accidents.

For the bias string, I'm going to go with the 1N5408 diodes I already have, 
which have a max forward surge current of 200A. If I put in a glitch 
resistor of 20 ohms, that will limit the short current to 150A, which should 
be good enough to protect the 1N5408s. 

## don't count on it !   It might work a few time then even the 1N5408's will
bite the dust.    What you really need is a higher value glitch R... like a 
50 ohm-50 watt wire wound... OR   2 x  25 ohm, 25 watt wire wounds in series. 

##  With 50 ohms, then you can really limit the fault current..to just 60A. 
I'd also highly recomend using 6A10's.  They will drop right in where a 1N5408
once stood.  Same lenght, same 16 ga wire leads, [aprx .050" diam].  Only diff
is the 6A10 is doubl;e the diam of a 1N5408.    The 6A10 is rated for a 400A surge. 

## You can also parallel  1N5408's...and also 6A10's... and yes, the current will split
50-50... or very close to that. 


With a plate current of 1A (more than 
it would really run at), that would be 20 watts, which I would put up to 30W 
for margin, since I run data modes on 6m. I'm assuming since this resistor 
is below the RF choke, it is OK to use wirewound. (If not, what would be 
better? I'm probably going to end up ordering it from Mouser anyway.)

##  Any series string of diodes  for bias is in the CT of the fil xfmr... and handles
full  CATHODE current.... which is the sum of grid and plate current. 

##  I use a 50 ohm-50watt  wire  wound in series with the 2650 vdc on my
4 x Drake L4B amps.   With 800 ma of plate current, the V drop is only
40 vdc.   Diss [key down] is  32 watts CCS.   On ssb,  average plate current
is only 400ma [ processor on]... and dissipation in the 50 ohm glitch is now down
to just  8 watts.  never had any problem with it.  The L4PS  use a .82 ohm, 1 watt
reistor in the B+  lead...as a 'HV fuse'.    

## The purpose of the glitch R is to LIMIT the fault current to a safe value.  The
purpose of the B+  fuse is to  INTERUPT the flow of current.  On the HB B+ supplies,
I use surplus BUSS  HB fuses...the ones that are 5" long... and fit 3/4" fuse clips.
These are sand filled units.  They open up in < 2 msecs. 

later... Jim   VE7RF




The bypass cap in the SB-220 is 1,000 pF (0.001 ?F). If I want this R-C 
circuit to prevent spikes that could blow things up, is this enough 
capacitance? I don't want to go much more than 20 ohms on the glitch for 
regulation reasons, as well as being a waste of power. The 1,000 pF bypass 
capacitor is a 5 KV stud-mounted doorknob type. I should be able to replace 
it with something bigger if that would help. Advice, please.

Bill W5WVO