4046 frequency multiplying

[Uncle Krunkus]

I'm just wondering whether it is feasible to put an 8bit counter across a 4046 in order to get a output which is 8octaves up.
I've seen on the datasheet that it only goes to about 1Mhz, but, apart from this limitation, would it work if the incoming signal stayed below 3.5 khz?
Anyone have any ideas or experience with this?

[dugernaut]

Interesting. I built a CD4046 based PLL using a CD4059 with some thumbwheel switches. I tried to make all the parameters such as loop damping, locking, etc (I forget right now) available. I figured I could get a quick freqency multiplier.

I only played with it a couple hours and lost interest due to more pressing experiments in frustration. I could sort of get a lock for an octave or two, but it was touchy and wobbly - not that wobbly is bad!

I ordered some 74HC4046's that are supposed to go way higher in freq, but haven't tried them yet.

If your looking to multiply (frequencies, that is) I've got a very rough idea on paper that could take a regular VCO square wave and I hope, multiply it up at least 5 octaves - maybe more - and there's NO PLL involved. Like I say, it's only a brain fart on paper, but if you'd like to work on it together (ie, build it up since I can't find the time right now) I'd be interested in working with you.

[Uncle Krunkus]

Yeah, I had an idea for doing it a different way as well, (kind of a self righting cmos multiplier thing) which was untested, had a multitude of chips, and probably slower to lock on.
I decided I'd be better off trying to get the 4046 to do what I wanted.

[dugernaut]

Actually I mispoke. My idea uses a ramp or saw and a counter. There's essentially zero lock time, wide frequency response and biggest problem is keeping the ramp at the same amplitude.

BTW, I have a couple extra of those PLL boards if you'd like one. Nothing special but you might have some fun with it.

[forbin]

I have been building the Thomas Henry X-4046 VCO and I changed it a bit to make use of the phase discriminator as a sort of sync variant. It seems to track really quite well -- it is being driven from a log converter so that should help a bit... maybe? I am going to try and do some measurements of its "pull-in" range to see how easy it is to control. It currently seems to jump to x3 the input frequency. The hysteresis is quite bit though...

Also... watch out for the 74HC4046 -- i got a nasty result when I forgot it was a 5V part....

[JovianPyx]

I designed and built this:

http://home1.gte.net/res0658s/fatman/4046pll.html

as a FatMan add-on. It has 7 frequency multiplier outputs, the highest of which multiplies the input frequency by 12. It works quite well and can settle quickly enough for audio use. I believe that settling time worsens as the multiplication is increased, but I don't know how to calculate that.

As dugernaut points out, 74HC4046 has a VCO which goes much higher than the CD4046 part. CD4046 goes to 1.3 MHz while 74HC4046 goes to 17 MHz. Just remember that 74HC4046 is a 5 volt part, but that's not really an issue since you can run CD4xxx parts at 5 volts too.

[Uncle Krunkus]

Hey Scott!
It was "Gravenhorst" which stuck in my head when I was trying to find that design of yours.
I actually googled "4046 gravenhorst" to find it!
Yeah, I'm hoping to have a go at it soon and see how far up it can go. Maybe I should try the 74HC4046.

[JovianPyx]

I just calculated what 8 octaves works out to with 1.3 MHz: 5078 Hz.

An input of 5 KHz will push the highest output (really, the PLL VCO output itself) of an 8 octave divider to 1.3 MHz - a bit into the ultrasonic.

So even the lowly CD4046 can produce inaudible high frequency signals with a large divider. But I'd still use the HC part in this case so that it won't freak out with what is otherwise a rather normal audio frequency.

If the desire is to purposefully have longer lock times, this can still be accomplished with smaller dividers and a loop filter with adjustable Fc that goes sufficiently low.

The reason I built my unit was to get tones with perfect fifth relationships to the fundamental. In my case, I wanted to eliminate as much of the lock time as possible, hence the adjustable loop filter. The key to the perfect fifth relationship is the inclusion of a divide by 3 element in the counter system.

[Uncle Krunkus]

I think I'd like to have the shortest lock times possible.
The idea is to fix my "Re-Animator" with some ideas I've been mulling over.
First is to make the waveform 256 steps long, no matter what the incoming frequency. Hence the 4046 with an 8bit counter in the feedback loop.
It would more than likely be used at input frequencies of 1K or lower, but covering up to 3-5K would be nice.

The next step is to replace my home made ADCs with ADC0804s which will give CV control of the start position for the output counter, (CV controlled phase) and will also control the curve of the wave, (CV controlled waveshape)

[Sebo]

Some time ago, I was looking for a freq multiplier to make a kind of
wavetable/waveshaper and use the high frequency square as a clock.
I'm working on it now, but my approach of the freq multiplier (for the clock)
is a little different, I keeping it analog, starting with a sawtooth I'm going
to use 8 saw doublers (there are a few schematics floating around) and
then a comparator to square the waveform. In this way there is no lock time.
I didn't finished it, so I don't know if it work.

[dugernaut]

I don't know if any of you would find this helpful, but I recently bumbled across this part. Didn't even know it existed! 74HC7046:
http://www.alldatasheet.com/datasheet-pdf/pdf/27074/TI/CD74HC7046A.html