[Boatanchors] Re Why Twin Diodes In Xtal Sets?

[John Franke]

On page 5 of the Volume 3, No. 2, September 1993 issue of The Xtal Set 
Society, John Collins submitted a very interesting note.  He indicated that 
he had tried to parallel 4-1N82A's in an attempt to lower conduction 
threshold.  There was a similar note on pages 3 and 4 of MRL (Modern Radio 
Laboratories) Data Sheet Volume 2.  Ed. Peil of West Point, CA, noted that, 
when driving a direct coupled transistor amplifier, wiring 2 diodes in 
series increases selectivity and wiring 2 diodes in parallel increases gain 
and reduces selectivity.



Dealing first with Ed.'s note, it is easy to see how the series diode 
connection increased the resistance of the detector and therefore reduced 
loading on the tuned circuit  resulting in increased selectivity. 
Similarly, it is easy to see how the paralleled diodes reduced the detector 
resistance which increased loading and reduced selectivity.  The reduced 
resistance also increased the circuit gain.  This increase in gain was what 
was so interesting.  Maybe another way of looking at it is to say the 
paralleled diodes have less loss.  I believe that this is what  John Collins 
experienced as well.  When someone does something (like paralleling diodes) 
and the result is good, it is worth further investigation.



Does paralleling diodes reduce the conduction threshold as John Collins 
thought?  It depends on how you define conduction threshold.  If you define 
it only by such solid-state physics terms as work function or band-gap 
energy, the answer is no.  If however you define it as the voltage necessary 
to generate a specific current, the answer is yes.  Let me try to explain 
that last answer.  Lets say a diode is defined as being above its conduction 
threshold when the diode current is equal to or greater than 10 microamps. 
Lets further assume that you have a collection of identical diodes that pass 
6 microamps with a bias of 1 volt.  A single diode would have a conduction 
threshold greater than 1 volt.  But, 2 diodes in parallel would have a 
conduction threshold less than 1 volt.  By paralleling the diodes you have 
lowered the effective conduction threshold.  But, have you really gained 
anything?



It was interesting that the note from John Collins came to mind when I was 
experimenting with some spiderweb coils.  I have a slightly modified MRL 
model 2A tuner that I use as a base unit to compare other crystal sets or 
components.  The set has a microamp meter so I can make relative 
measurements to prove improvements (or losses) in efficiency for different 
detectors, capacitors, coils, etc..  It was a simple matter to just parallel 
the existing 1N277 detector with another and see if there was any change.  I 
was quite surprised to see the detector current increase by about 16%. 
Adding another diode increased the detector current by another 10%.  It was 
hard to put solid numbers on the increases because of the current variations 
caused by modulation.  So, I wired up a test circuit to better evaluate the 
effect.  I loaded the detector with a 2.2K resistor to simulate the load of 
a pair of high impedance headphones.  The radio frequency source was a TTL 
crystal oscillator operating at 1.00 MHz.  A 50 microamp meter was used to 
measure the detector current.  The input level was varied to produce a 
single diode detector current from 2 to 40 microamps in steps of 2 
microamps.  At each step, the single diode was paralleled first with one and 
then with two additional 1N277 diodes and the detector currents noted.



The percentage increase is greatest at the lower currents.  Also, adding 
more diodes would probably produce even more of a gain but only marginally 
so.



This simple experiment does not tell the whole story.  What happens to the 
selectivity?  Right now I cannot measure the selectivity, maybe by next 
year?  Maybe someone else can try this experiment and go further and include 
selectivity measurements.  I think it is a subject worth  investigating.



I can supply the graphed data off-line.



John  WA4WDL



----- Original Message ----- 
From: "Duane Fischer, W8DBF" <dfischer at usol.com>
To: <boatanchors at mailman.qth.net>
Sent: Friday, February 08, 2008 6:08 PM
Subject: [Boatanchors] Re Why Twin Diodes In Xtal Sets?


>
>
> Hello All,
>
> I have noticed that some MW and SW crystal sets from the forties and 
> fifties employ a pair of detectors instead of just one. Can you please 
> explain to me the reasons why and how such a circuit would work?
>
> Is there 'really' a benefit to using two?
>
> Although I have never personally come upon plans for one, I was told by an 
> older Ham that two diodes were used in crystal sets that received FM. Is 
> this correct? If so, what role do two diodes play in FM signal detection 
> specifically?
>
> Thank you.
>
> Duane W8DBF
>
> Duane Fischer, W8DBF/WPE8CXO
> dfischer at usol.com
>
> HHI: Halligan's Hallicrafters International
> http://www.w9wze.net
>
> HHRP: Historic Halligan Radio Project
> hhrp.w9wze.net
>
> _______________________________________________
>