Discussion in '"Boat Anchor" & Classic Equipment' started by K1APJ, Jun 3, 2021.
just regular silicon diodes like a 1n4007???
All depends on the exact design of the TR switch, doesn't it?
73 de Jim, N2EY
False sense of happiness. Silicon diodes don't start conducting until something like 0.7V.
Germanium diodes are much better suited to de-sensitization protection--Old-time 1N34A or obsolete but obtainable 1N270
diodes--conduct at something like 0.3V.
I'd stick with the higher power diodes.
0.7v or slightly more is not going to cook the input of a typical solid state receiver.
Typically such diodes exist at the front end of many rigs (check ur schematic?)
But in case of a nasty failure or arc over, I'd want the extra power capability.
Nothing to stop one from using BOTH!
Belt AND suspenders!!
Also W8JI, iirc, looking at his analysis over states the max voltage that appears as a transient
at the output of the Johnson TR box... trying to remember, but I think he looks at the voltage
before the attenuation of the output transformer... anyhow it is actually lower in practice. Otoh,
not sure he tested it at full QRO.
My clamp uses 1N914s, and has performed well for many years. The lowly 914 is rated to handle an Amp for a second and 4 Amps for a µS, so a couple amps at a few milliseconds isn't unreasonable.
The tubes used in the popular T/R switches can't muster anywhere near that kind of current, especially with the toroids used. As noted above +/- 0.7V or so won't hurt any decent SS front end.
As for 4007s, I'd think they have a lot of junction capacity and may not work well at mid/high HF freqencies.
Here's a link to W8JI's analysis:
AB2RA's commentary is spot-on too:
particularly about the fusing and modifying the switch to reduce heat.
What it boils down to is that the EFJ design has an inherent response delay. IOW, when the RF starts, it takes a fraction of a millisecond for the cutoff bias voltage to build up, and for that tiny bit of time, significant RF (watts) can get through to the receiver.
Note that W8JI did a SPICE simulation using a 6SN7, not an actual measurement. OTOH, he used the 6SN7, which is a much less powerful tube than the 6BL7 used in the EFJ switch.
And of course in the time frame when the EFJ and other switches were being made, hams were using hollow-state receivers.
The EFJ TR switch is reviewed in QST for December 1958.
73 de Jim, N2EY
a few minor points... WRT W8JI's analysis, 300volts is not possible as the unit, iirc uses about 150vdc B+?
And I forget what the step down ratio for the output autoformer is, but I seem to recall 10:1 or better.
That puts the max signal at 15volts, not 300volts!!
Also, the max power that can be put through that output autoformer is also going to be limited by the
saturation of the ferrite core material. I don't think those were toroids but wound on ferrite rod?
Mike's caveat about junction capacitance for diodes is something to consider.
Likely there are some silicon diodes with low capacitance to be had, if one wanted to do that.
Just a quick one. Would this work if fitted between a Hammerlund HQ180A and a Johnson ranger as they don't have any switching so I'm looking at one of these inline transmit/receive units like this one.
The ranger is only 45 watts on CW.
Thanks guy for looking.
Yes, that is a typical application for the unit. Keep in mind you will need some sort of speaker/receive audio muting, simply use the Receive/transmit switch on the 180.
Does your ranger have PTT, or is it an early unit?
If you are going to do that then you might as well use a Dow-Key or reasonable facsimile to switch the muting circuit of the HQ-180.