Problems making an LFO with a 7414 schmitt trigger

[JingleJoe]

The propblem:
Earlier this night I attempted to make a low frequency oscillator with a 7414 inverting schmitt trigger/buffer. I used the standard circuit for a CMOS schmitt trigger oscillator with one resistor between input and output and one cap from input to ground.
It didn't work.
Any resistance greater than about 2k didn't work and/or resulted in a distored square wave or triangle wave at the output.

The question:
Why?

I found out that this chip (the 7414) uses transistor logic (TTL) which requires more current than CMOS which is probably why the larger resistances didnt work. I've done that much of my own research but I want to know what you chaps say too and whether I am on the right track.

[Cynosure]

When i made my first synth circuit over a year ago i used a 7414 because that is what the person in the video tutorial used. It didnt work, but luckily he also mentioned 40106 in the video so I had bought some of those too. I switched the chip and never went back to any 7xxx ic's. I thought it was me being a newb, but it looks like maybe that wasnt the case.

The triangle wave doesnt sound like a bad thing though. What sort of frequency range do you get with the triangle? Can you take a shot of the waveform on your scope?

[JingleJoe]

[Kabzoer]

You can put it trough a divider to get an exactly square wave with 50-50 symmetry, found that trick in the CMOS cookbook

[JingleJoe]

[Kabzoer]

Yup, 'cause a 4040 divider is 12 cascaded T-flip-flops!
I heard, that in the 74xx series, the middle letters really make the difference, so for example, your circuit might work with a 74C14, but not with a 74HC14 or 74HCT14... But I don't really know what difference that makes....

[JingleJoe]

[Kabzoer]

If you want a 50-50 triangle, then I think the symmetry of a 7414/40106 is just too low (especially at low frequencies) so if you want something better, you can use the one in the atonement, essentially it works the same, but it uses Op-amps, and is optimized for symmetry at these frequencies... You can go with your wonky one too, to keep things simple....

http://bit.ly/uXRemx

---

synthkit_lfo_noise_power_001.gif
Description:
Filesize:	20.05 KB
Viewed:	461 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[JingleJoe]

Accuracy is unimportant, having a semblance of stability and acctually working is all that really matters
I was experimenting with an oscillator very similar to that one just last night, but I still prefer these nice simple CMOS schmitt oscillators.

[JingleJoe]

Just had an idea after posting that reply (see circuit in attachment for my idea) this circuit should still supply the current needed for the TTL schmitt trigger to work while giving the cap adjustable charge/discharge speed. A diode could be connected in parallel with R2 to make the output a pulse or saw wave. R2 should be a pot too to acheive variable frequency.
Values are guessed, my simulator didn't want to run this circuit so it is untested, experimentation time!

---

TTL osc untested.png
Description:
Filesize:	2.81 KB
Viewed:	13036 Time(s)

---

[JingleJoe]

experiment results are in:

The above diagram only sort of works, the wave from the top of R2 is really nice and odd

However simply using resistor values below about 3k and the original circuit configuration works okay! It has a nice big frequency range but the triangle wave is very lop-sided and the square wave is a pulse, at high frequencies it's a very sloppy pulse, but it works!

Additionally putting the control pot of about 5k in parallel with a 1 or 2 k resistor gives a perculiar frequency response and allows enough current for the chip's input when the pot is turned all the way up to highest resistance (too little current would make the oscillator work improperly, outputting a very high frequency and wonky wave as mentioned above).

[richardc64]

Symmetry is elusive, even with CMOS.

The attached recordings were made from an actual 40106 with actual resistors and capacitors. I will never completely trust simulations.

With a 1k feedback resistor the waveform looks 50-50. At 100k symmetry is beginning to slip, and gets more and more unequal as R is increased. This pattern held for different value caps. The ramp waveform available at the capacitor would likewise be asymmetric.

Although I have a nostalgic fondness for TTL, I wouldn't give myself headaches trying get a predicable or consistent squarewave from a 7414 as an oscillator.

---

asym_2.gif
Description:
Filesize:	11.79 KB
Viewed:	292 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[JingleJoe]