low frequency micropower vco needed

I need a low frequency vco (to operate between 55 and 65 Khz) that
draws very little power (microamp supply current if possible). Goggle
turns up uhf and microwave vco's that draw 10's of ma. It needs to be
a square wave, but can live with a triangle or sine wave output.

A single chip solution would be preferred, but can build it from
discrete components if necessary.

I'd like to have it sweep between 55 and 65 Khz with the varying
control voltage.

A cmos r/c oscillator might work just fine, but I can't figure out how
to vary the frequency with a dc input voltage:>:

Any suggestions?

Thanks,

A

 

Vary VDD

...Jim Thompson

 

Thanks Jim,

It sounds too good to be true though, is it really that easy?

A

 

Use a cmos 555
Typically draws 60uA


Peter

 

I know it works for ring oscillators used as VCO's, but as soon as you
add R/C's I'm not sure if it still works.

(The ring oscillator is really varying gm.)

...Jim Thompson

 

Again, thanks Jim.

After I read your first message, I googled with different keywords. As
a result, I did find a schematic which showed a 4584 cmos gate
oscillator with the input (Vcc) varied to control the output
frequency. With an R of 2.2 meg and a C of 220 pf, the author claims
that 2.5v in produces 1 Khz out and 9v in produces 20.8Khz out.

I'd post it, but I think binary attachments are forbidden here.

Thank you so much for taking to the time to reply, I appreciate it.

A

 

Frequency shouldn't change more than a few percent with Vdd over the
normal operating range of a CMOS gate.

You could make a VCO with an op-amp and a comparator, but how about
the VCO portion of a 4046?



Best regards,
Spehro Pefhany

 

[snip]

Post it to...


...Jim Thompson

 

That works because the 4584 has hysteresis. I think an ordinary
inverter will vary only a small amount.

I have made custom chips where the inverters had current mirrors in
the rails.

...Jim Thompson

 

That's a ST hex inverter.

just post a link, like this one:

http://www.discovercircuits.com/PDF-FILES/4584vco.pdf


Best regards,
Spehro Pefhany

 

I looked at the 74HC and 74HCT variants on the Phillips www site. They
make some similar chips with slightly varying part numbers, but I
checked them all.

These chips are intended to run as high as 19 Mhz, and they all draw
way to much quiescent power too. There might be a chance of one
working if there was a slower speed part, but I don't see anything
like that. I'm not sure why they draw so much current when they aren't
switching, but they do.

Thanks,

A

 

The On MC74HC4046A has maximum Iq of 4uA -55 to 25°C, 40 uA < 85°C.
That's not so bad.



Best regards,
Spehro Pefhany

 

OK, th e file is at:

http://www.discovercircuits.com/PDF-FILES/4584vco.pdf

It appears to met my needs very nicely, although I don't need nearly
as much frequency variability as this circuit provides.

If a standard 4000 series gate can provide a 10 percent variance, I
could probably use it.

I think the 4584 is easily available and cheap enough, I'll probably
use it.

Thanks to all who helped out.

A

 

Barry hat on.

If you cared to take some time and consider the data provided by the OP then
I am sure you will notice that he requested that the supply current was in
the order of a microamp. Obviously your solution is unsuitable by a factor
of four and, perhaps, an order of magnitude.

Pick one of the following

1) I have personally designed such circuits which worked perfectly.
2) I have personally been involved with people who design such circuits and
had many interesting discussions in order to steer them in the right
direction.
3) I have personally been involved with people who design such circuits and
had many interesting discussions with them so I know what I'm on about.
4) I was personally involved in a project where such a circuit was designed.
No-one took my advice and it didn't work.
5) I was personally involved in a project where such a circuit was designed.
My advice was taken, it didn't work and they were subsequently proved to be
wrong. If they would have taken my advice properly then it would have
worked.

Now, I think I have asserted my credentials and thus proved that you are
wrong and I am right.

Would you care to respond and give me an opportunity to............

Oh Perlease!, I'm gagging for it

BNA

 

Perhaps, but his "microamp level" statement is sufficiently hazy that
10uA or 50uA might also be okay... he mentions 10mA as being
unsuitable.

Well, 1) except for all the differences, but hey, you've convinced me.

Okay, go for it. 1uA (maximum? typical?) at 70kHz operating current
VCO, with realistic parasitic capacitances.


Best regards,
Spehro Pefhany

 

[Derf Transform Applied]

I see no reason to *go for it*. If you know the answer then I look forward
to your explanation. I will not waste my time with needless arguments.

I see that you understand. Perhaps we should wait for the OP to respond so
we can further discern his actual requirements.

Yes, but if you read OPs question again you will see he was asking for
65kHz-70kHz operation. This is a very narrow range to achieve given your
original solution of a 74HC4046 which operates to 19MHz.

You are quite right. I should have posted to ask the OP to further define
his actual requirements. This is probably a battery application and we
should be more concerned about the gradual draw down due to internal
impedances.

Welcome

BNA

 

If it was possible to do it with picoamps, I'd say go for it! It's not
likely to happen for awhile though.

I said 'microamps'. The statement was meant to discourage those
replies from people that had milliamp solutions-after all, some people
consider milliamps to be low power.

Thanks again to all who had suggestions.

A

k

 

Best regards,
Spehro Pefhany

 

Ah, okay, perhaps I misunderstooood what you meant when you said you
were "gagging for it". Perhaps that's just as well.

3.0/11 MHz min at 3/4.5V Vdd for the one I mentioned, but that's more
like 19MHz when you convert to GBP. Anyway you don't want CMOS sitting
inbetween the rails too much of the total period or it will blow away
the uA limit. Or get rid of the CMOS bit. IIRC there were some really
low power oscillators that put high-value resistors in the drains of
the MOSFETs in a 4007.

Best regards,
Spehro Pefhany