[ARC5] Product detectors - the name.

Wed May 10 23:08:09 EDT 2017

Hi Bruce,

Your math argument is sound.  It describes the fundamental operation of any
frequency mixer, whether used for demodulation (i.e. detection),
modulation, or frequency generation.  The real tough part of mixer design
is keeping those undesired outputs suppressed, ports isolated and linearity
preserved.    Modern mixer design involves lots of transistors and
multi-phase osc injection.

Just so there isn't any confusion among the readers' one cannot have a
simple diode detector function as a product detector and an envelope
detector at the same time.  Leave out the BFO and it's just a half wave
rectifier whose filter cap smooths out the IF ripple but isn't large enough
to smooth out the audio "ripple."  When a healthy BFO signal is inserted on
the IF side of the diode it will chop the conduction of the diode at the
BFO rate.  In that situation it is a product detector.  Our ancestors were
forced to be minimalists and found ways to add product detection to AM
detectors a simple, inexpensive ways.

Dennis AE6C

On Wed, May 10, 2017 at 11:59 AM, Bruce Long <coolbrucelong at yahoo.com>
wrote:

> I agree with Les:
>
> As an RF design engineer I associate a product detector as a circuit block
> where the output voltage is the product of two input voltages hence the
> name product detector.
>
> It is entirely true that a diode amplitude detector will function as a
> kind of product detector if supplemented with BFO injection.  Here is how
> that works.
>
> Excuse the math
>
> The diode works because it has a non-linear Iout/Vin curve
>
> In circuit analysis you deal with the circuit non-linearity by
> representing the non-linearity as an infinate series expansion have the
> form:
>
> Vout = a0 +a1(Vin) + a2(Vin)^2 + a3(Vin)^3 + ............,.
>
> Where a0, a1 ... an  are coefficients
>
> Vin is the sum of two sinusoids,  one being the IF frequency signal, the
> other being the BFO injection signal.
>
> If you look up the formula for the square of the sum of two sinusoids in a
> math handbook you will find that expression has a Vif  times Vbfo  product
> term and it is that product term that provides SSB an or CW detection.
> All the other terms generate intermodulation and harmonic terms and are
> undesirable.   Some can be filtered out, others cannot be filtered and
> contribute distortion.
>
> If you have a circuit block that really and truly has an output voltage
> that is the product of the two input voltages and only the product of the
> two input voltages then all the all the unwanted, un-filterable
> non-product terms go away allowing clean non-distortion multiplicative
> detection.
>
> Communication theorists like product detectors because they are simple and
> easy to represent mathematically.
>
> So in short all SSB CW detectors are product detectors in as much such
> detection and down conversion requires the multiplication of two sinusoids
> ( looking at the trig section of any math handbook multiplication of two
> sinusoids creates both sum and difference frequencies) and any circuit
> non-linearity will provide a measure  of "product" detection becaue of the
> V, V^2, V^3 etc terms that appear when one attempts to define a circuit
> non-linearity in terms of a series expansion.
>
> A true product detector in my way of thinking is a circuit especially
> configured to provide an output that is the product of two input signals -
> and nothing- else to the degree possible.
>
>
> Sorry for the lecture but I once taught this stuff at the university level
> and they take away my professor membership card if I don't occasional make
> like a professor.]
>
>
> Bruce
>
>
>
>
>
>
> ------------------------------
> *From:* Leslie Smith <vk2bcu at operamail.com>
> *To:* Dennis Monticelli <dennis.monticelli at gmail.com>; ARC-5 List <
> arc5 at mailman.qth.net>
> *Sent:* Wednesday, May 10, 2017 2:59 AM
> *Subject:* [ARC5] Product detectors - the name.
>
> Hello Dennis,
> I always believe a product detector followed the trig function:
>
> sin(A) * cos(B) = 1/2[sin(A+B) + sin(A-B)].
>
> It's the PRODUCT of two trig functions to give the sum and difference.
> I have no idea why I believe this, but I suspect it's true (for a
> product detector, or product mixer)
> Put in two signal and get the sum and difference.  Isn't maths amazing!!
> Isn't electronics amazing!!  If anyone knows more than this - I'm
> interested!
>
>   73 de Les Smith
>   vk2bcu at operamail.com
>
> On Wed, May 10, 2017, at 03:43, Kenneth G. Gordon wrote:
> > On 9 May 2017 at 9:38, Dennis Monticelli wrote:
> >
> > >
> > > Guys,
> > >
> > > I feel compelled to contribute to this discussion on the impact of the
> detector upon the AGC.
> >
> > Good. Thank you.
> >
> > > First of all, the detectors we are discussing are just special purpose
> mixers and like all mixers they
> > > generate the mathematical products of their inputs.  Regardless of
> what we call them....detectors,
> > > product detectors, converters, etc.....the math is the same.
> >
> > Yes. I completely agree with you and I have been saying the exact same
> > thing for years.
> >
> > > Now, we don't want all of those math products because some of them can
> cause trouble.  If we
> > > put a strong BFO signal into a simple detector in order to extract a
> relatively low distortion CW or
> > > SSB demodulated output then the BFO energy must be substantially
> stronger than the incoming
> > > signal and that energy can easily overwhelm any downstream AGC pickoff
> point
> >
> > Correct.
> >
> > > When SSB came into being the linearity of the detector became more
> important than ever
> > > because even small amounts of voice distortion are irritating.   To
> overcome this problem the
> > > detector was made balanced with respect to the incoming BFO.  Properly
> executed this balance
> > > reduced the amount of BFO energy bleeding into the signal path by a
> lot.... about 40dB.  Now one
> > > could pump in lots of BFO, enjoy low distortion, and not mess up the
> AGC.
> >
> > Well, yes, but in addition, the AGC voltage was taken from some place in
> > the circuit OTHER
> > THAN the detector. Separating the two makes a huge difference.
> >
> > >  For some historical
> > > reason unknown to me this type of detector became widely known as the
> product detector.
> >
> > Supposedly because the output was the "product" of the two inputs, and
> > not a straight simple
> > addition. Supposedly also if the BFO signal is missing, output is zero,
> > so an AM signal
> > applied to the input of a true "product" detector would result in zero
> > output.
> >
> > However, I have never believed that.
> >
> > One of the simplest product detectors, first used by Collins in the
> > KWM-2, a simple triode,
> > demodulates AM just fine.
> >
> > Heathkit used that PD in almost all of their rigs.
> >
> > I also wonder about the term, "linear mixer": how can any "mixer" be
> > "linear" since it must be
> > "un-linear" in order to work?
> >
> > Even so, the so-called "product detector" is about the best detector
> > available for SSB and
> > CW.
> >
> > Ken
> >
> > ---
> > This email has been checked for viruses by Avast antivirus software.
> > https://www.avast.com/antivirus
> >
> > ______________________________________________________________
> > ARC5 mailing list
> > Home: http://mailman.qth.net/mailman/listinfo/arc5
> > Help: http://mailman.qth.net/mmfaq.htm
> > Post: mailto:ARC5 at mailman.qth.net
> >
> > This list hosted by: http://www.qsl.net
> > Please help support this email list: http://www.qsl.net/donate.html
> ______________________________________________________________
> ARC5 mailing list
> Home: http://mailman.qth.net/mailman/listinfo/arc5
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:ARC5 at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.qth.net/pipermail/arc5/attachments/20170510/14f1e3c9/attachment-0001.html>