Simple Opamp Triangle-Square wave LFO

[Sweetalk]

Hi everyone!, I want to build a simple LFO with triangle and square wave selectable by a switch, I looked at the Tremulus LFO but (that works really great) but I want to switch between waveforms easyly.

Looking in my computer I found a schematic that I took from this forum (I can't find the topic or the autor of the schematic, I hope it's OK with him to post it), very nice, opamp-based LFO... Oh!, just what I looking for, checked with my Opamp book and looks like it's OK, mounted on protoboard and NOTHING happened, the LED just stays on all the time,the depth pot works fine. Doube checked and all it's on it's place, there's voltage on the IC's (using a couple of UA741 v+ pin 7 and gnd pin 4), the transistor it's in the right position...

I have only one question about the schematic... the C1 cap, 47uF that goes from negative input to output it has to be non polarized? I'm using an electrolytic one, tried both ways and the result it's the same.

Well... here is the schematic, again, I took it from this forum but I can't find the topic or the autor, if the autor of the schematic doesn't want that I post the schematic I take it right away!!.




If anyone has another schem of a triangle-square wave LFO post it, I'll make a PCB layout for it.

Thanks!!!

PD: excuse me for my bad english!!

[Andre]

There is a dot before the 47uF, so I guess the value is 0.47 uF, in which case you wouldn't need a electrolytic cap at all.

André

[Sweetalk]

oohhh... you're right!.... I didn't notice that at all!!... I'll try changing the cap.

Thanks!!

[Sweetalk]

I tried the .47 uF cap and I got the same result... just a bright LED at the output. Checked all the circuit again (and rebuild it also) and it's just like the schem... no blinking at all... accidentaly took off the speed pot and the result was the same, I think that's a little bit odd.

Thank you all!

[Paul Perry (Frostwave)]

Incredibly long shot, but does it work if the - and + inputs on the first op amp are switched over?

[gez]

Incredibly long shot, but does it work if the - and + inputs on the first op amp are switched over?

The inputs look as though they're wired as they should be.  I think the problem lies in the type of op-amp you're using (what is it?).  Unless the amp has rail-to-rail output capabilities, it's never a good idea to make R1 = R2 as the trigger thresholds of the Schmidtt will never be reached and the thing just latches up.

Try reducing the value of R1 and compensate for the increase in frequency by making the .47u cap larger.

[gez]

PS  If you don't own a scope, you can still check if the LFO is working (at least you can rule out that part of the circuit) by subbing in a large value NP electrolytic for the integrator's cap then checking the output of the 2nd amp with a multimeter.  The LFO should now be slow enough to see the square wave output swing from rail-to-rail.

[gez]

PPS  The inputs of many amps don't like being pulled close to the rails, so unless you have a suitable op-amp for this application, you might never get the thing to work (the chip can be damaged).

[Sweetalk]

Incredibly long shot, but does it work if the - and + inputs on the first op amp are switched over?

I checked on an OpAmp book that I have and the basic structure it's alright, the first opamp it's wired just like in the schem.

The inputs look as though they're wired as they should be.  I think the problem lies in the type of op-amp you're using (what is it?).  Unless the amp has rail-to-rail output capabilities, it's never a good idea to make R1 = R2 as the trigger thresholds of the Schmidtt will never be reached and the thing just latches up.

First I tried a couple of 741's... then a TL072 and RC4558... I have a TLC2272, it's a dual low noise rail-to-rail opamp, could this work??.

Try reducing the value of R1 and compensate for the increase in frequency by making the .47u cap larger.

How much I reduce R1 and increase .47uF cap?? 47K and 1uF will be fine?

Thanks!!!

[gez]

How much I reduce R1 and increase .47uF cap?? 47K and 1uF will be fine?

Yes and No.  It will give you the same frequency (IIR - I'd need to check the formula), and there's less chance of latch-up or the inputs of the schmidtt being damaged, but it reduces the amplitude of the triangle output, so the associated circuitry around Q1 might need tweaking. 

An alternative is to take R1 down to the next rung and see if things work.  If not, go to the next resistor value down and so on until you get some joy.  A small decrease in this resistors value probably won't result in such a noticeable increase in frequency, so you might not need to increase the cap value, but if there is a slight increase you could just stick at small value cap in parallel with the one you've got in there, to compensate.

Re the op-amps you're using.  Report back if you can't get the thing to work at all (either with your suggested values or by just reducing R1 step-by-step) and I'll take a look at the data sheets to see which might be causing the problem - I'd rather do this on a 'need to' basis. 

[Sweetalk]

I tried the TLC2272 with the original parts and nothing happened, the LED still on all the time. Changed R1 to 82K, nothing, 68K nothing, 47K, nothing, 33K nothing!!. All this keeping the 0.47uF cap, then I make all this with a 4558 and nothing really happened. I'll keep on trying but if someone has a opamp triangle-square wave LFO schem is welcome to post it!

Thanks!!

[ulysses]

the second half of roger mayers voodoo vibe is an extensive wave generator and is available if you search the board

cheers
ulysses

[slacker]

I'll keep on trying but if someone has a opamp triangle-square wave LFO schem is welcome to post it!

I've got a simple one I'll post it later.

[gez]

I tried the TLC2272 with the original parts and nothing happened, the LED still on all the time. Changed R1 to 82K, nothing, 68K nothing, 47K, nothing, 33K nothing!!. All this keeping the 0.47uF cap, then I make all this with a 4558 and nothing really happened. I'll keep on trying but if someone has a opamp triangle-square wave LFO schem is welcome to post it!

Thanks!!

Just because the LED isn't blinking doesn't mean to say the LFO isn't oscillating, it might be working only the circuitry that follows isn't functioning.  Did you try the test I mentioned (large value NP cap and check the output of the schmidtt for rail-to-rail output swings)?

I can't see anything wrong with the schematic linked to, except that with R1 being the same value as R2 without there being any recommendation as to what amp should be used is going to result in this circuit being a little hit or miss for some who try it.  Having said that, I've skimmed thru the data sheet of the TLC2272 and I don't see why it shouldn't work.  So, you've either wired things up wrongly, have incorrect part values in there (check all resistors), have damaged the chip (static) and it'll never work no matter what you do, or you've wired up Q1 etc wrongly.

[Joe Kramer]

Hey Sweettalk,

I've just been working with a similar LFO, and what you have there looks like it should work.  It's possible that even if the LFO is working, the LED will stay lit when there is simply too much voltage/current feeding the transistor driver.  I suggest a voltage divider pot between the output of the LFO and the base of the transistor, sort of like an "input volume control."

If that doesn't help, you might want to just borrow the LFO from the Boss CE-2,  schemo here:

http://www.godiksennet.com/images/sch/CE2.jpg

It's essentially the same thing as the one your using, with slightly different parts values.  I've used this LFO with a TL062 with great success many times. . . .

Regards,
Joe

[Sweetalk]

I'll try the pot and see what I get... Can you get triangle and square from the CE-2 LFO??

[Joe Kramer]

Yes, you can get a squarewave at Pin 7 of the CE-2 schemo LFO. . . .

[Sweetalk]

I'm checking with Tonepad's Schematic, your image it's a little bit small to see the numbers and I'm a little bit blind . The out of the LFO it's on the output of the second opamp, there you have square wave right?, and triangle?... I want to have both waveforms and select them by a switch.

[Joe Kramer]

Right, sorry.  The square wave is available at pin 7, or the output of the first op amp.  The triangle wave is at Pin 1, or the output of the second op amp.  Also, for a much wider range of speeds, try this:

--Change the .1 cap (C19) to .047

--Change the 1meg resistor (R32) to 470K

--Change the 10K resistor (R31) to 470 ohm

Hope that helps!

[Transmogrifox]

Try setting the 1M pot to about 1/2 way.  The schematic you have posted is very reasonable, so it seems very unlikely that it wouldn't work as shown, or at least with R1=82k.  If you have .47 uF (not .047 uF) and all other values shown, and it is correctly wired, then it should do something.  Maybe you put the transistor in backwards?  If you are using a 4.7uF, you'll be waiting quite a long time for it to snap over.  Make sure your battery has a good charge on it.  It may not work if your battery is nearly dead.