Amps
[Top] [All Lists]

[Amps] L4B biasing

To: <amps@contesting.com>
Subject: [Amps] L4B biasing
From: "Jim Thomson" <jim.thom@telus.net>
Date: Thu, 25 Dec 2014 15:36:47 -0800
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Date: Wed, 24 Dec 2014 22:14:01 -0800
From: "Jim W7RY" <w7ry@centurytel.net>
To: <amps@contesting.com>
Subject: Re: [Amps] L4B biasing

11 diodes (7 volts) is way too much bias voltage.

Better to have about 3 diodes (1N5408). That is what I have found while 
designing my QSK circuit board for the Drake. The L4 and L4B was designed 
with Amperex 8802 tubes. They did not require
bias because the idled at a much lower current than Eimac 3-500 tubes. And 
the old L4 was designed around the 3-400 (8163) which did not require bias 
either.

Sure, if you have 3800 volts of plate voltage then 7 volts of bias is fine. 
The SB220 used a 5.1 volt zenner and the SB220 plate voltage was around 3200 
volts.

73
Jim W7RY

##  I own FOUR drake L4B amps..and have them all modified.   Drake states in 
their manual
how much idle current  a 3-400Z  will produce..vs a eimac 3-500Z.   They list a 
half doz tubes in there,
and idle current will vary from 160ma  to  240 ma. 

##  240ma + 2650 vdc = a blast furnace.   I used 10 x  1N5408s  fro bias.   
Then a miniature SPDT..
or DPDT..with contacts in parallel  to switch the taps on the diodes.  To pull 
this off, you require a 
CENTER OFF  type of toggle.   Center off = all  diodes in the circuit,  The 
other 2 positions are used
to tap the junction of diodes 3+4.....and 6+7.    That way I get 3 x different 
bias positions.   I did this
since I have a lot of spare tubes, both eimac and also other brands, and some 
are high mu..and some are low mu. 
200mu  and 130 mu. 

##  with all 10 diodes in the circuit, and using eimac high mu  3-500Zs....idle 
is 100 ma  when on 2650 vdc no load
SSB position....and just 40 ma..when on 1900 vdc  no load CW position.   IMD is 
just fine on ssb, in either  cw or
ssb  positions.  625 watts out on 1900 vdc position...and 1285 watts out on 
2650 vdc position. 

## The diodes are wired on perf board....and installed in series with the 
yellow wire, that u see going to the oem
3pdt  tr relay.    Drakes oem bias scheme was fubar.  2 x 50 k 50 watt wire 
wounds in series with a 7kw /5 watt..
across the B+ to B-  in the outboard hv supply.    I tossed both the bleeders 
and the 7 k resistor. In the oem  scheme,
there is a +130 vdc  drop across the 7 k rsesistor..when in SSB mode...and +90 
vdc  when in CW mode..used for
cut off bias.  

##  Cut off bias is now just a 50 k..or paralleled 100 k 3 watt mof resistors 
in the ct of the fil xfnr.  Re- wire the tr relay
so the resistor is shorted on on TX.   Dead simple. 

##  End result is the outboard HV supply runs  cool.  B4 those bleeders  were 
dumping 70 watts of heat in there. They are
not required since u already have  100 k  eq resistors across each HV cap.   NO 
load B+ in either cw or ssb  position did not budge
either.    The RF deck will  run one helluva lot cooler  with 100ma idle 
current vs 230 ma. 

##  While u are at it, toss the silly grid caps and also grid chokes.....save 
the chokes as they are good ones, with only .9 ohm 
dc resistance.  Bond the grid pins to the chassis with wide cu strap.   Your 
drive power requirements will drop by 20 watts. 
By reducing the drive power by 20 watts, the typ Xcvr  will have cleaner TX  
IMD.   The drake L4B..even with 10 diodes, still
has better imd than any ham xcvr.....except for a yaesu in Class A. 

##  The little blower in the RF deck is marginal at best.  220-240 ma of idle 
current is looney tunes.   I experimented  and u can reduce the
idle current to just 5-10ma  when operating CW mode.   1900 vdc and sending CW. 
  You cant run a GG amp with no idle current, or u will
get key clix. 

Later... enjoy xmas everyone

Jim  VE7RF  
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>
  • [Amps] L4B biasing, Jim Thomson <=