Something else to keep in mind when fooling with AM detector
circuits. If you use a separate twin diode tube for the
detector diode, run it at reduced voltage (4V vs 6.3V).
This makes for less "contact potential" around the cathode
that the incoming signal would have to overcome and maybe
be clipped a little. So weaker signals can be handled
better. See:

http://home.earthlink.net/~wa2ise/im...896lowhtxt.jpg
http://home.earthlink.net/~wa2ise/im...6lowhcurve.jpg

The 6AL5 and 6H6 are quite similar, and people have reported seeing
6H6s run at reduced heater voltage.

Also having independent diodes means that you can use them
where you'd really like to, instead of being stuck with cathodes
to ground all the time. Though be mindful that the heater will
look like a few picofarad capacitance to whatever the cathode is
connected to.

Here's a radio with the low heater detector diode, and with CF
buffer:
http://home.earthlink.net/~wa2ise/im...ompaa3v2-1.jpg

I've actually built this and tested the reduced heater voltage
on the diodes and it really did make a good difference. Shoulda
done that before suggesting that bad idea on making AM test signals
the other day... :-)

"Robert Casey" wrote in message
...
Something else to keep in mind when fooling with AM detector
circuits. If you use a separate twin diode tube for the
detector diode, run it at reduced voltage (4V vs 6.3V).
This makes for less "contact potential" around the cathode
that the incoming signal would have to overcome and maybe
be clipped a little. So weaker signals can be handled
better. See:

http://home.earthlink.net/~wa2ise/im...896lowhtxt.jpg
http://home.earthlink.net/~wa2ise/im...6lowhcurve.jpg

The 6AL5 and 6H6 are quite similar, and people have reported seeing
6H6s run at reduced heater voltage.

Also having independent diodes means that you can use them
where you'd really like to, instead of being stuck with cathodes
to ground all the time. Though be mindful that the heater will
look like a few picofarad capacitance to whatever the cathode is
connected to.

Here's a radio with the low heater detector diode, and with CF
buffer:
http://home.earthlink.net/~wa2ise/im...ompaa3v2-1.jpg

I've actually built this and tested the reduced heater voltage
on the diodes and it really did make a good difference. Shoulda
done that before suggesting that bad idea on making AM test signals
the other day... :-)

Yeah I'd heard that lower heater voltage thing somewhere before!
But looking at the second reference given (.... lowhcurve) above it doesn't
appear to make much difference to the plate current versus voltage curve, as
far as I can see. The slope is about the same on 6.3 or 4 volts?
If hum is the primary concern could one DC the diode tube heater off the 6.3
RMS volt heater supply using a diode, resistor and capacitor? Prob get about
4.3 volts based on roughly half the peak voltage of the 6.3 volt supply?
Come to think of ss.diodes wouldn't have a 'hum' problem, no heater!
Hmm! Wonder where this discussion is going?
Terry.

Yeah I'd heard that lower heater voltage thing somewhere before!
But looking at the second reference given (.... lowhcurve) above it doesn't
appear to make much difference to the plate current versus voltage curve, as
far as I can see. The slope is about the same on 6.3 or 4 volts?
If hum is the primary concern could one DC the diode tube heater off the 6.3
RMS volt heater supply using a diode, resistor and capacitor? Prob get about
4.3 volts based on roughly half the peak voltage of the 6.3 volt supply?
Come to think of ss.diodes wouldn't have a 'hum' problem, no heater!
Hmm! Wonder where this discussion is going?
Terry.



The curve at reduced heat seems slightly straighter, but I think the
main issue is the contact potential in high impedance detector
circuits goofing things up. The input signal would have to
provide enough energy to absorb the contact potential energy
before getting the diode to conduct. The contact potential
would look like a negative back bias on an ideal diode.

SS diodes have that 0.3 or 0.7V drop with a rather yucky knee that
makes "well, just bias it up by 0.7V" schemes not work that well.

In article , Robert Casey
wrote:

The curve at reduced heat seems slightly straighter, but I think the
main issue is the contact potential in high impedance detector
circuits goofing things up. The input signal would have to
provide enough energy to absorb the contact potential energy
before getting the diode to conduct. The contact potential
would look like a negative back bias on an ideal diode.

This always confuses me, so maybe I have it backwards, but I thought
contact potential tended to bias the diode on? If we connect an amp meter
between the cathode and plate of the diode, we will measure current flow,
so it sounds to me like the diode is already conducting because of the
contact potential, and we have to make extra effort to turn it off, not
on?


Regards,

John Byrns


Surf my web pages at, http://users.rcn.com/jbyrns/

John Byrns wrote:



This always confuses me, so maybe I have it backwards, but I thought
contact potential tended to bias the diode on? If we connect an amp meter
between the cathode and plate of the diode, we will measure current flow,
so it sounds to me like the diode is already conducting because of the
contact potential, and we have to make extra effort to turn it off, not
on?


I think you're right, in that the diode is biased on a bit. Which
makes for problems on weak AM signals, in that in some time segments
the diode is always conducting and thus not detecting.
Reduced heat would reduce the contact potential, thus allowing
proper detection. Or at least something resembling proper detection

I've actually built this and tested the reduced heater voltage
on the diodes and it really did make a good difference. Shoulda


This trick usually helps, but not always. I always try it, but once or
twice I've had to revert to normal heater voltage. I can't recall the
specific circuits that didn't benefit. I'm not sure if they were audio
detectors or ANL diode clamps.

It's fun waiting for the tube to slowly reheat and the sound to change
when shorting out the heater resistor during testing.

Brian

Brian wrote:


This trick usually helps, but not always. I always try it, but once or
twice I've had to revert to normal heater voltage. I can't recall the
specific circuits that didn't benefit. I'm not sure if they were audio
detectors or ANL diode clamps.

Probably just the audio diode. The AVC and ANL diodes may want more heat.


It's fun waiting for the tube to slowly reheat and the sound to change
when shorting out the heater resistor during testing.

That's what I did. Looked for a sweet spot. Also checked that strong
signals didn't get munged when weak signals were good.

Robert Casey wrote:

Brian wrote:


This trick usually helps, but not always. I always try it, but once or
twice I've had to revert to normal heater voltage. I can't recall the
specific circuits that didn't benefit. I'm not sure if they were audio
detectors or ANL diode clamps.

Probably just the audio diode. The AVC and ANL diodes may want more heat.


It's fun waiting for the tube to slowly reheat and the sound to change
when shorting out the heater resistor during testing.

That's what I did. Looked for a sweet spot. Also checked that strong
signals didn't get munged when weak signals were good.
early DeForest audion detector boxes had variable settings for
the filament and plate voltages to find a sweet spot to operate the
tube. Back then this was necessary because the tubes weren't pumped
to a high vacuum and no two bulbs were alike. The trace of gas left
in the tube would ionize if the plate or filament voltages were set
too high, the sensitive spot was just below the ionization breakover
point.

Before he added the grid to the valve, DeForest played around with
a diode type audion. He used a B battery in his hookup to operate
the detector near saturation with detection taking place as the signal
drove the tube into and out of saturation. Fleming operated the
diode across the cutoff point, between positive and negative plate
voltages. With the early low vacuum tubes it is possible that
DeForest's circuit might have been more sensitive than Fleming's.
With a high vacuum diode, the tube probably works better as a detector
without the plate voltage set near saturation. Of course, what Deforest
was trying to do was to amplifiy the signal by adding the B battery
potential to the signal. He finally got that idea to work by adding
the grid and creating the triode.

Robert Casey wrote:




Yeah I'd heard that lower heater voltage thing somewhere before!
But looking at the second reference given (.... lowhcurve) above it
doesn't
appear to make much difference to the plate current versus voltage
curve, as
far as I can see. The slope is about the same on 6.3 or 4 volts?
If hum is the primary concern could one DC the diode tube heater off
the 6.3
RMS volt heater supply using a diode, resistor and capacitor? Prob
get about
4.3 volts based on roughly half the peak voltage of the 6.3 volt supply?
Come to think of ss.diodes wouldn't have a 'hum' problem, no heater!
Hmm! Wonder where this discussion is going?
Terry.



The curve at reduced heat seems slightly straighter, but I think the
main issue is the contact potential in high impedance detector
circuits goofing things up. The input signal would have to
provide enough energy to absorb the contact potential energy
before getting the diode to conduct. The contact potential
would look like a negative back bias on an ideal diode.

SS diodes have that 0.3 or 0.7V drop with a rather yucky knee that
makes "well, just bias it up by 0.7V" schemes not work that well.
Just a note ---- National HRO 60 does this on this on the 2nd det/avc
and noise limiter both 6H6. They use a 4.3 ohm series resistor.
John K3OPC