Variable Square Wave Generator

I use a 20 old Philips function generator, that fulfills your needs. Besides that it also can generated a sinus, a triangle and has adjustable slope and duty cycle. The deltas, you call it has a rough step switch and a fine tune.

I would check for "function generator" under instruments. D

Something like this maybe....

http://www.ebay.com/itm/10MHz-DDS-Function-Signal-Generator-Module-Sine-Triangle-Square-Wave-TTL-SWEEP/321157000855?rt=nc&_trksid=p2047675.m1851&_trkparms=aid%3D222002%26algo%3DSIC.FIT%26ao%3D1%26asc%3D261%26meid%3D4048477143611782464%26pid%3D100005%26prg%3D1088%26rk%3D2%26rkt%3D5%26sd%3D281148493311%26

I don't know if this device meets the OP's requirements, but it sure meets mine. Thanks for posting this Solar Eagle.

:)

Since it says 'TTL Sweep', it's probably only 0-5V (or less), while OP specified 0-10V.

As you indicate, it is an interesting device!

Probably draws too much current, as well. The OP is looking for something CMOS-ish - microwatt stuff - in terms of current draw. Battery operation. I'm sure there are low-power digital freq. synthesis chips Out There (like on GlobalSpec, for info, and Digi-Key, maybe, for parts?) Methinks the OP wants something compact, as well - minimal parts count and small size.

To the OP: Would a Raspberry Pi or an Arduino card meet your needs? You'll have to do a bit of programming...

Something whose output is in the low MHz range but with freq steps of 50 Hz is unlikely to be programmable to such fine resolution using an RC network. I'm speaking here of something like a (CMOS) 555, etc., so you're pretty much stuck with digital freq. synthesis.

The OP did say 50kHz deltas, not 50Hz, but gave no indication of the precision required.

Some years back, I built a crystal-controlled function generator with a phase-locked-loop and 4 digit decimal control. I thought I knew right where it was, but it wasn't there. I couldn't find any reference to it in my computer. That means I must have made it before the year 2000, when I first got access to a digital camera. I think it was CMOS...

yes it is 50KHz (+- 5%), and it will have to be a CMOS device. I don't mind programming. Reuven.

Thanks for clarifying.

You still haven't indicated the precision required of the steps or output frequency. Does the frequency need to be correct within ± 25kHz, ±0.1kHz, or ...

If the precision needs to be high, then you need a crystal-controlled original source, but since you indicated using inductance or capacitance for tuning, I suspect you don't need that high a precision. I don't understand your aversion to resistance; It's a whole lot easier to use a variable resistor for tuning than variable inductance or capacitance.

I don't understand the objection to resistors, either. Finding precision resistors to set freq (or a simple pot) is a *hellava* lot easier and cheaper than precision capacitors or inductors (even if you wind them yourself, the latter also typically feature rock-bottom-dismal tempcos when using ferrites).

±5KHz will be more likely. I am not enough experienced but I tend to assume that resistors are more of energy wasters than caps or coils.

I built a CMOS-555-based oscillator which ran continously for about five months on two AA cells. That circuit used resistors.

Why the bold face?

I frequently find my assumptions wrong too...

If you were trying to use a resistor to control the speed of a motor, then yes, it would waste considerable energy. In this case you have to have at least one resistor to control the charging rate of a capacitor (assuming [again] an R/C oscillator). With an appropriately chosen capacitor, that resistor will be in the 100s of kΩ, so the charging current will be tiny; virtually no energy wasted. If you need discrete steps, you will need a series of resistors, but again the current will be tiny.

The easiest way would be to use one fixed resistor to set the maximum frequency, and one variable resistor to drop to the currently desired frequency. Look up CMOS 555 circuits.

Yup.

To the OP: And since a CMOS 555 has very-high-impedance comparator inputs, you have lots of latitude in how you scale the resistor values in the RxC combo. A large cap with small R values or a small cap and large R values (up to around 10 megohms I recall), either will work for a given RC product, so opt for for the latter to conserve battery life.

if I have no other choice I'll use it like that. Thanks.

That is more or less my direction. Thanks.

Blame this accursed autocorrect ('automangle,' more like). I caught it trying to substitute 'GHz' in place of 'kHz' but then it did an end-run and plopped in 'Hz' instead.

(I think the technical term for it is 'Progress' - with a capital 'P')

:(

Your PLL-based synth used a 4046 (or 4044 or something close. Call it 404x out of sheer laziness)? I recall seeing something along those lines in an old RCA or Nat'l Semi data book way back when Christ was a Colonel. :)

yeah. Remmber him from that tme

If there is no choice, I will use 5v p to p

Reuben, what is your application? If your freq stability and accuracy requirements are fairly lax, you could save yourself a great deal of trouble and expense by using a CMOS 555.

the frequency is accurate to the extent of plus minus 25KHz. I need it to build up 300-400 v on a cap. If it is 10-20 v p to p it will be quicker and if it is 5 p to p it will take a long time but I can take it. I only need the entire unit to consume as less energy as possible

Reuven, why is the freq important for this app? Are you using the oscillator to drive a high-Q, LC resonant circuit? You know you can also use your oscillator to drive a diode/capacitor voltage multiplier without concern for frequency, yes?

Another thought: use a CMOS 555 oscillator with a freq somewhere in *audio range* - 15 kHz or so, and connect the output pin (#3 I think) to the low-impedance (8-ohm) side of a cheap, miniature audio transformer. Connect the transformer high-impedance (1kohm) side to the inputs of a diode/capacitor voltage multiplier. Experiment a bit. I got 80-90 volts just from the transformer alone when rigged this way.

If you live in the U.S., you can buy all the parts at Radio Shack.

You echo my thoughts exactly! 2 GAs!

Hey, thanks! :))

Was your question answered satisfactorily?

Yes guys, Thanks a million

What are you building that you need to charge a cap to three or four hundred volts? Just curious (as always :)

If you need something quick & dirty (and cheap) to charge a 300-400 V capacitor (in the 200+ uF range, typically), buy one of those cheap, so-called 'disposable cameras' at your local Walgreens and remove its photoflash unit. The cameras cost what, ten bucks?

It is only a back up power supply, and your "camera" idea is interesting. Thanks.

A 'backup power supply' for something requiring 300-400 volts, to be powered by a 9-volt battery? (???)

Okaaaay...