Dream about a 4047!

[StephenGiles]

The 4047 is used with perhaps a 4007 or some other means of voltage control in a number of flangers to provide clock signals for the BBD. What happens to pin 13??? The answer - absolutely nothing. But why, when there is a perfectly good pulse output, the width of which is continuously being varied in sympathy with the change in frequency of the clock outputs at pins 10 & 11 - something to think about if you wake up at 3am and can't get back to sleep!!! The possibilities must be endless.

[puretube]

continuously variable?
(as for PWM purposes?)
to me it only says for pin13:
"50% dutycycle not guaranteed",
in the datasheet...

[redeffect]

On the A/DA flanger schematics (1024 & 3010) pin 13 is the clock test point. 34.8 kHz - 1300 kHz. The clock outputs are 17.4 kHz- 650 kHz .Maybe the pulses can be tapped off pin 13 and fed into a 4041. This will maintain the 34.8-1300 kHz frequencies, while giving you the 2 phase clock to drive BBDs. (see Craig Anderton Hyperflange + Chorus for more details about this clocking setup) Choice between clock frequencies for one BBD (4047 or 4041 output)?  Maybe run 2 BBDs; one off each  clock output (4047 and 4041 output)? These are just some thoughts. Who knows ?  Happy brainstorming to all!
red

[puretube]

If I interprete the datasheet of the 4047 correctly,
it already puts out 2 symmetrical phases @ pins 10 & 11...
any wellknown buffer or inverter could be subsequenced to each of these outputs,
to increase "driveability"

[StephenGiles]

Could the pin 13 output be "normalised" perhaps, then processed to produce a fixed frequency clock, whipped through a 4013 for 2 further symmetrical phases  - fixed delay - through zero flanging with one clock???

[redeffect]

The clock outputs of the 4047 at pins 10 & 11 are half the frequency of pin 13. Using pin 13 to feed a 4041 will give you a 2 phase clock with the same frequency as the output of pin 13 (2x the f of pins 10 & 11, 4047). The 4041 does not halve the input f like the the 4013. Yes, you are not changing the ratio. You are however changing the delay range. Maybe this can be used to create  a flanger with 2 modes. "Standard" & "High Band". Like I said, this is all just the speculation of a bonzo like me. Maybe Mark H will read this and shed some light/thoughts/clarifacation on this matter. Damn Stephen! Now look what you started Making people actually think "outside the box"
red

[StephenGiles]

Well redeffect, if you are as bright as our 3 legged dog Bonzo, this is all good stuff. Anoter thought, put a few PWM phase stages in the flanger feedback path, controlled by pin 13.

[puretube]

 

[redeffect]

Well redeffect, if you are as bright as our 3 legged dog Bonzo, this is all good stuff. Anoter thought, put a few PWM phase stages in the flanger feedback path, controlled by pin 13.

I would easily give the nod to 'ol 3 legs... 
red

[puretube]

Stephen: finally played around with scope & 4047...

pin13 is the buffered output of the oscillator you put up
by inserting a resistor & a capacitor on pins 1/2/3;
this R/C combination with the built-in gate-loop can only oscillate,
when the chip is allowed to do so by enabling circulation
by means of pins 4/5/6/8/12.

From this oscillator, the signal also goes to another buffer,
and from there to a flipflop (similar to half a 4013),
which makes a 50% duty-cycled half frequency signal out of it,
whose one output is again being buffered/inverted twice to feed pin11,
and is being buffered/inverted 3 times to feed output10.

(if the logic is programmed to output those signals, again with pins 4/5/6/8/12).

Those 3 last buffers to pins 10/11/13 are each comparable to one of the
inverters of the well-known 4049/4069 hex-types.

So pins 10/11 are always symmetrical half frequency of pin13
(when in astable mode).
However: the oscillator itself will only output a more or less symmetrical square,
(well-squared by the buffer)
when used with a "normal" (=linear) resistor.

As soon as you do the diode-trick to vary the frequency,
the wave (pin13) starts to deviate from 50% dutycycle at high frequencies
towards ~1% or less at low frequencies;
never more than 50%, though...
the variation from 50 to 1 doesn`t seem to take place linearly with the decrease
in frequency, and looks uselessly for "our" purposes, IMHO.

Interestingly, I did some testing with caps (resembling a BBD clock-input)
at pins 10 & 11.
Their loaded output looks the same like a single 4069`s inverter loaded output.
The more (2/3/4/5) inverters you wire up parallel, the steeper the slopes get
at higher frequencies (as expected...).

The MOST interesting part of playing with this chippy was however
(besides getting annoying heterodyning in my FM-radio above 800kHz clocks
when powering the breadboard with a 12V switchmode power-supply
instead of battery!!!),

was a new way to control the frequency:
by a variable resistor to ground!

in combination with a slightly modified "known" diode-trick voltage-control
(e.g. from an LFO), (or singularly...)
it is now possible to hook up a shielded mono cable to a passive external rocker-pedal
with a pot wired as a variable resistor,
on/off switchable by a SPST stompswitch,
to vary the clock-frequency of any device
between ~2Hz to ~1.7MHz. (without switching caps!)
One resistor can be added, of course, to limit F_min,
and another one for F_max...

In combination with the usual diode-trick voltage-control,
F_min goes up to ~60Hz.

the circuit consists of a 4047, a 220p, and two 1N4148...

[oldschoolanalog]

The idea of controlling the frequency w/a variable resistor to gnd sounds like it has some real potential. Could you please post a diagram showing how to go about this? Folks less informed (me) would find this very helpful. Thanks!

[puretube]

time is not ripe yet, for this... 

[puretube]

DIP16 is a bit hefty,
with all the unneccessary logic on board (for this case, i.e.).

today`s gonna search for s.th. smaller... 

[puretube]

hmm - the resistor to ground works fine, as long as it is a (linear) resistor (or LED)...
however: the "hot end" carries the HF signal, which would be loaded down
by hooking up a long shielded wire.
So: no remote-control possible! 

[lovekraft0]

Maybe use a fast Vactrol for the resistor, and use a pedal to sweep the LED current?

[puretube]

hmm - the resistor to ground works fine, as long as it is a (linear) resistor (or LED)...

make that: or LDR, of course 


the usual (floating) vactrol remote thing can be done with almost every inverter-based oscillator...

it was the "one side grounded" thing, that intrigued me.

[puretube]

OK: 2 more days of playing around with PWM, flipflops, astables and inverters...

got the HF signal which was riding on the "tip"-wire to an external "resistor-to-ground"
cable from 10V pp down to ~10mV,
and the cable may as well contribute over 10µF on capacitive load,
while the external resistor to ground still can span 160Hz to >1.6MHz...

removing the resistor, the standalone voltage control (via a 100k pot)
spans from 280Hz to 2.1MHz.

there indeed seem to be more than 3 ways to hookup a couple of inverters,
a 270p, and 2 Diodes
to a pot...

of course, the output of the oscillator loves to be divided,
to get proper 50% duty-cycled pulses for clocking a BBD.


This got me quite excited!





Ok-Ok: I could`ve used a 4046 instead...