[hpsdr] Input ptotection

[FRANCIS CARCIA]

all,
  I suggested the 1N4148 in series with a transorb another name for a TVS. The problem with the transorb in the junction C. The 1N4148 is reversed biased so there is no conduction below the transorb voltage. The reverse bias of the 1N4148 reduces the Cap load on the inut. 
  Two circuits are required for positive and negative clipping. The Racal RA6830 uses 4 1n4148s in parallel in each polarity but there is provision for up to 8 or ten in parallel. 
  Small transorbs can't handle a lot of current. 1.5 KW units would be a place to start. 
  The reverse bias of the 1N4148 has to be high enough so a high level signal does not conduct the diode or there will be IMD problems. I have done a lot of transorb testing with 
  DO 160 type lightning pulses  and the small ones blow up quite easy. As they conduct current the junction voltage takes off. a while ago I had to use a 10 volt 3 KW unit to protect a RS485 XCVR. The pulse was 45 volts induced and could also induce about 40 amps on the center conductor. By the end of the pulse the junction voltage was close to 15 volts.
  I have flashed over my tuner with 1/4 inch spacing ofthe caps with a storm approaching.
  There also needs to be a high resistance across the input as a diode reference which will also blead off ESD.  Frank WA1GFZ
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20070806/be7755e8/attachment-0003.htm>