Amps
[Top] [All Lists]

Re: [Amps] Plate modulation from power supply ripple?

To: Jim Thomson <jim.thom@telus.net>, "TexasRF@aol.com" <TexasRF@aol.com>, "amps@contesting.com" <amps@contesting.com>
Subject: Re: [Amps] Plate modulation from power supply ripple?
From: "Fuqua, Bill L" <wlfuqu00@uky.edu>
Date: Sat, 5 Oct 2013 14:34:56 +0000
List-post: <amps@contesting.com">mailto:amps@contesting.com>
   Small degree of modulation of the Helix supply would be bad.
The helix supply accelerates the electrons to establish a velocity that matches 
that of the wave traveling along the helix. If those velocities are not matched 
you will have less bunching of the electron beam and lower output from the twt.

73
Bill wa4lav

________________________________________
From: Amps [amps-bounces@contesting.com] on behalf of Jim Thomson 
[jim.thom@telus.net]
Sent: Saturday, October 05, 2013 8:38 AM
To: TexasRF@aol.com; amps@contesting.com
Subject: Re: [Amps] Plate modulation from power supply ripple?

###  You need that deluxe HV crowbar circuit that the OE4 fellow cooked up.  
GM3SEK et all will no doubt chime in and provide the correct url.  That will 
shut it down asap..quick like.   The V drop across a sense resistor in series 
with the
B+ is fed to the control circuit.    The cro-bar fires when plate current 
exceeds a pre-set threshold.... and is in  the cold end..... between a glitch R 
and B-.      The original design was for  I believe, a delicate tube, used for 
eme or similar,
at 1296 mhz or higher, the tube was near impossible to source.   The modules or 
sections can be stacked in series for higher B+ levels, and has been done 
already..and I think up to almost 8 kv.   It appears it could be used well > 10 
kv.

##  lets say 25% eff in class A..and 300w out.   300/.25 = 1200w dc input.    
9000v / 1200  =  .134 A plate current.   Does this sound right ?

Jim   VE7RF





From: TexasRF@aol.com
Sent: Saturday, October 05, 2013 4:43 AM
To: jim.thom@telus.net ; amps@contesting.com
Subject: Re: [Amps] Plate modulation from power supply ripple?

Jim, it is not that simple. The there is about 9000vdc and normal helix current 
is less than 5mA.

Right now there is a 12.5K ohm glitch resistor to limit any fault current but 
that is not enough to fully protect the tube in case of a flashover. More R 
could be added but then the voltage drop during normal operation becomes too 
much and changes in the helix current screws up the already marginal voltage 
regulation.

Reducing the amount of filter C would reduce the stored energy but at the 
expense of higher ripple voltage. I need to spend some time experimenting with 
the filter C value to see just how low it can be for acceptable 120 Hz hum on 
the carrier.

The thought of less hum level because of linear operation class was new to me. 
The filter C design was based on the tube spec showing .08 dB per volt change 
in gain. In that scenario 10v of ripple would cause a gain change of .8 dB. .8 
dB is a power difference of 20% which would be very noticeable if it was 
present in the form of hum modulation.

The tube runs class A and is fairly linear but not perfect by any means. It is 
run at maximum power output, around 300 watts at 10.4 GHz . Running the tube 
this way is considered to be in power saturation. That does no damage to the 
tube but is not that great for linearity. We mostly use cw or one of the 
digital modes so linearity is not an issue.

A suggestion was made to use three phase 400Hz voltage derived from three audio 
amplifiers to reduce the needed C and that would certainly work.. 9000v at 5mA 
is only 45 watts  Before taking on a major rework like that I will do some 
homework with existing power supply.

I thought  this was all figured out. Wrong!

73,
Gerald K5GW



In a message dated 10/5/2013 4:12:17 A.M. Central Daylight Time, 
jim.thom@telus.net writes:
  Date: Fri, 4 Oct 2013 08:26:13 -0400 (EDT)
  From: TexasRF@aol.com
  To: ww1c@outlook.com, garyschafer@comcast.net, amps@contesting.com
  Subject: Re: [Amps] Plate modulation from power supply ripple?


  My current interest is mainly in microwave eme operation. Traveling wave
  tubes are used to generate power. The homebrew power supplies have a lot of
  filter C in the helix supply to reduce hum on the transmitted signal. After
  this  discussion I am wondering if there may be more C than necessary.
  Normally it  would not matter but if there is a flashover in the twt the helix
  structure  could be damaged. There are trip circuits to shut the power supply
  down but most  of the stored energy would be dissipated in the helix.

  73,
  Gerald K5GW

  ##  How much B+   ??    How much plate current ??  Just put a 50 ohm glitch R
  in series with the  B+.... and wire a fast hv fuse just in front of the 50 
ohm glitch.
  Anything arcs, the glitch limits the fault current to a safer value.   The HV 
fuse interupts
  the fault current..and  will easily do that in less than 2 msecs.   But you 
have to size the
  fusing wire correctly.    Add the glitch +  HV fuse on top of your existing 
protection.

  Jim  VE7RF


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

<Prev in Thread] Current Thread [Next in Thread>