The national LF155 datasheet has a schematic for a 3 decade VCO made out
of an LF356 and an LM319. I don't know how linear it is, or how fast
it'll track the input frequency.
---
Nice one but, unfortunately, the OP's asking for from 10Hz to 30kHz,
and the one on the data sheet goes from 10Hz to 10kHz with a control
voltage of zero to 30VDC!
He only needs a period of 2 seconds for his tuning voltage, so that
shouldn't be too hard for the VCO to follow, and I'm thinking along
the lines of a two opamp triangle wave generator to generate the sweep
and a 7555 for the VCO.
Feeding the control voltage input with the sweep voltage and wiring
the chip up as an astable with a 50% duty cycle output would be kind
of interesting in that, leaving out the output buffer and the MOSFET,
the 7555 looks essentially like this:
Vcc>---------+
|
[R1] +--------[Rt]-------+
| | |
TH>----------------+--|+\ +------+ |
| | | >---|R Q|--+--->OUT
Vc>----------+-----|--|-/ | |
| | | |
[R2] | | |
| | | |
+-----|--|+\ | _|
__ | | | >---|S Q|
TR>----------------+--|-/ +------+
| |
[R3] [Ct]
| |
GND>---------+-----+
The input divider resistors, R1,R2,R3, are all the same value, so with
no external voltage on Vc the circuit will oscillate between 1/3 and
2/3 Vcc with an output period of about 1.4RtCt.
If we were now to connect Vc to a voltage source (the sweep voltage
generator) which varied between close to 0V and close to 5V, we would
be able to vary the output frequency over a very wide range, the
highest frequency occurring with Vc close to 0V and the lowest
frequency occurring with Vc close to +5V.
That's scheme 1, but I don't have any numbers yet.
Scheme 2 leaves the the input divider alone, but varies the timing
resistor resistance 3000:1 with the sweep voltage input. That's not
as far-fetched as it sounds, since with the saming timing cap, if we
could get 30kHz with 1000 ohms we ought to be able to get 10Hz with
3 megohms. Since
T = 1.4RC,
For 30kHz and 1000 ohms we'd need
T 3.3E-5
C = ------ = -------- ~ 23.8nF
1.4R 1400
and just to check 10Hz:
T 0.1s
R = ------ = -------------- = 3 megohms
1.4C 1.4 * 23.8nF
Implementation might be something as simple as this:
+5
|
[R1]
|
+----+
E |
Vc---[R]----B Q1 [R2]
C |
+----+
|
[R3]
|
+--->TO 7555 PINS 2 AND 6
|
[Ct]
|
GND
Where Q1 is an N-Channel FET or a PNP bipolar, R1 and R3 make up the
1000 ohm high-freq resistance and R1, R2, and R3 make up the 3M low
frequency resistance with the transistor cut off. Kind of iffy,
though. Maybe an LDR and an LED or an opto? Dunno yet.
Scheme three is a high side voltage controlled current source with a
3000:1 current range feeding the timing cap. 1µA low freq current out
and 3mA high freq current out? Or 10µA -> 30mA? Or 5 and 15? Sounds
pretty good and less klunky than scheme 1, and _lots_ less klunky than
scheme 2...
Time will tell.
