Two Tones, One Transistor

I seem to recall from long, long ago, that the DTMF phones that Ma Bell
put out used a single transistor to produce both tones. I can't find
any info on how this circuit would operate. How can you create a
circuit that will resonate at two frequencies at once? How did Ma Bell
do it?

Rick

 

Funny that the approach pops up right now here in SED. We're
right now in progress to use such a multitone oscillator to read out
an array of radiation detectors, a detector per frequency. Kind of
a multi-tone version of the HP200A.

Regards,
Mikko

 

This is fun. I threw together a simulation model based on John's hint,
and it looks like it works.

I got intrigued by multi-tone oscillators because of a trick mentioned
in the context of high-stability quartz oscillators, where Mike Monett
decribed how the difference between the 3rd harmonic and the 3rd
overtone could be used to stabilize the frequency to a few 1e-9.
I have yet to apply it though...

Anyway below is my LTspice dual-tone oscillator model.
Comments invited!

Jeroen Belleman

============================================
Version 4
SHEET 1 880 680
WIRE -272 -224 -336 -224
WIRE -128 -224 -208 -224
WIRE -48 -224 -128 -224
WIRE 48 -224 16 -224
WIRE -336 -144 -336 -224
WIRE -272 -144 -336 -144
WIRE -128 -144 -128 -224
WIRE -128 -144 -208 -144
WIRE -48 -144 -128 -144
WIRE 48 -144 48 -224
WIRE 48 -144 16 -144
WIRE -336 -96 -336 -144
WIRE 192 -80 192 -144
WIRE -336 -48 -336 -96
WIRE -272 -48 -336 -48
WIRE -128 -48 -128 -144
WIRE -128 -48 -208 -48
WIRE -48 -48 -128 -48
WIRE 48 -48 48 -144
WIRE 48 -48 16 -48
WIRE 48 0 48 -48
WIRE 80 0 48 0
WIRE -336 32 -336 -48
WIRE -288 32 -336 32
WIRE -128 32 -128 -48
WIRE -128 32 -208 32
WIRE -64 32 -128 32
WIRE 48 32 48 0
WIRE 48 32 16 32
WIRE 192 80 192 0
WIRE -288 128 -400 128
WIRE -64 128 -208 128
WIRE 64 128 16 128
WIRE 128 128 64 128
WIRE 64 208 64 128
WIRE 192 208 192 176
WIRE 192 208 128 208
WIRE 240 208 192 208
WIRE 272 208 240 208
WIRE 384 208 336 208
WIRE 192 240 192 208
WIRE 192 352 192 320
WIRE 192 480 192 432
FLAG -400 128 0
FLAG 384 208 0
FLAG 192 480 0
FLAG 192 -144 0
FLAG 240 208 e
FLAG 80 0 0
FLAG -336 -96 tank
SYMBOL res 176 224 R0
SYMATTR InstName R1
SYMATTR Value 1k
SYMBOL npn 128 80 R0
SYMATTR InstName Q1
SYMATTR Value 2N3904
SYMBOL cap 336 192 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C1
SYMATTR Value 100p
SYMBOL cap 128 192 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C2
SYMATTR Value 100p
SYMBOL ind2 -80 144 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 5 56 VBottom 2
SYMATTR InstName L1
SYMATTR Value 100u
SYMATTR Type ind
SYMBOL voltage 192 16 R180
WINDOW 0 24 96 Left 2
WINDOW 3 24 16 Left 2
SYMATTR InstName V1
SYMATTR Value 6
SYMBOL voltage 192 448 R180
WINDOW 0 24 96 Left 2
WINDOW 3 24 16 Left 2
SYMATTR InstName V2
SYMATTR Value 6
SYMBOL ind2 32 16 R90
WINDOW 0 5 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName L2
SYMATTR Value 25u
SYMATTR Type ind
SYMBOL cap 16 -64 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C3
SYMATTR Value 100p
SYMBOL ind2 -304 144 R270
WINDOW 0 32 56 VTop 2
WINDOW 3 5 56 VBottom 2
SYMATTR InstName L3
SYMATTR Value 100u
SYMATTR Type ind
SYMBOL ind2 -192 16 R90
WINDOW 0 5 56 VBottom 2
WINDOW 3 32 56 VTop 2
SYMATTR InstName L4
SYMATTR Value 25u
SYMATTR Type ind
SYMBOL cap -208 -64 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName C4
SYMATTR Value 120p
SYMBOL diode -208 -160 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName D1
SYMBOL diode 16 -160 R90
WINDOW 0 0 32 VBottom 2
WINDOW 3 32 32 VTop 2
SYMATTR InstName D2
SYMBOL diode -48 -208 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D3
SYMBOL diode -272 -208 R270
WINDOW 0 32 32 VTop 2
WINDOW 3 0 32 VBottom 2
SYMATTR InstName D4
TEXT 280 -32 Left 2 !.tran 10u
TEXT 280 16 Left 2 !.ic v(e)=1
TEXT -96 208 Left 2 !K12 L1 L2 1
TEXT -312 208 Left 2 !K34 L3 L4 1

 

Ma Bell had Fairchild make an IC to spec for DTMF; I otta know, as i
did a lot of testing of the first fab ICs, and helped devise a way to
test them in a decent time (that d*mn inductor takes a looong time to
charge from zero).

It has been a rather long time (since around 1969 more or less), and
do not remember the designation given; i believe it was in the uA700 series.

Digging thru the datasheets i kept from those daze, i found the uA745
which looks like that may be the specific IC for that purpose.
ONE: It is a DUAL, and uses 10 transistors total; it is in a 14 lead
flat cerpak (which matches my memory of the package) and the datasheet
shows two test circuits.

#1 circuit uses 27 ohms in series with 0.75Hy(32 ohm) inductor and
that network is paralleled with 340 ohms; this is the load to 12.6V; and
used to test gain, Zout, bias current tests and supply current test.
#2 circuit uses 68 ohms in series with 2.5Hy (68 ohms) inductor and
that network is paralleled with a 660 ohm resistor in series with a 4uF
capacitor; this is the load to 6.7V; and used for the THD test.
The load was common to the two amplifiers.

Ma Bell had a rather tight series of test limits including THD,which
obviously cannot be tested in a reasonable time for production purposes.

We (mostly me) characterized a thousand devices over temperature (5
temps) for all given specifications and scatter-plotted THD against
everything else.

There was absolutely ZERO correlation; the BeeHive plots did not even
allow limits on any given parameter to guarantee THD, much less in the
spec temp range.

The engineer involved did some analysis and made some slight changes
in resistor values so that biasing of the one transistor causing poor
THD at -40C (worst case) would not have a forward Vcb greater than its
Vbe and thus would not be in actual or practical saturation (only in
theoretical saturation). In other words, it still acted as an amplifier
and not a polarity-reversing rectifier.

Very dicey as Ma Bell demanded the design and values within some range.

 

Well, the Baby Bird (Goo Gull) presented only a few.
In any case it seems impossible to find the ONE (of hundreds) that
mentions DTMF much less gives a circuit.

 

Isn't this what the early touchtone generators did?

--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email:

Please visit my GURU's LAIR web site at http://www.tinaja.com

 

I remember learning about phone phreaking when I was in college. I went
directly to the library to find that copy and of course, it was gone,
most likely taken by a student rather than removed by the school I
expect. It had the exact frequencies published although at that point
the horse was out of the barn.

Rick

 

The secret insider phone phreaking document was -- The Bell System
Technical Journal!

It was originally much easier. All you needed was a safety pin and a
bent coat hanger.

--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email:

Please visit my GURU's LAIR web site at http://www.tinaja.com

 

Ma Bell used LC tanks - multi tapped too! There was one transistor on
the DTMF pad, acted as an amplifier.

 

I was wondering if there was a way to tune a single circuit for two
frequencies. I get that there are two tank circuits on one transistor,
but what keeps the tank circuits from interacting with each other?

 

I appreciate the response, but that wasn't exactly what I was asking.

I was wondering why the two tank circuits don't interfere with the tuned
frequencies. Are the two circuits totally isolated? If so, how
exactly? I don't get that from the schematic, although I don't have
that in front of me at the moment.

 

If the tank circuits and the amplifier are
linear, the superposition principle keeps
the tanks from knowing about each other.
Because stable oscillation requires a
nonlinearity, to keep the amplitude
finite, the trick (as John pointed out)
is to have a *separate* nonlinearity for
each tank.

We`re planning to use bolometers
as the nonlinearities. We also expect
to open a can of worms there, for instance
the harmonics created by the nonlinear
element *will* be seen by other tanks. So,
one has to arrange the frequencies
suitably.

Regards, Mikko