sup-simp lfo mods? A.K.A. I'll show ya mine ifa ya show me..

[ExpAnonColin]

LFO modulated LFOs...  You could also modulate the waveform if you were super careful.

-Colin

[guitarhacknoise]

Word!
So obvious, so simple! ( at least, idea wise)
I could just slap myself!

[ExpAnonColin]

OK, re-breadboarded with a 10k instead of 1k... I stand corrected.  On my 'scope, I can see a difference of about .2 volts in the amplitude over quite a wide range of frequencies (using a 10k resistor instead of 47k and a 500k pot instead of 100k).  Now I'll see what I can do about click-damping and sawtoothing.

-Colin

[ExpAnonColin]

Still clicking.  You can get a very nice sawtooth out of pin 1 by putting 2 1n4001 diodes in opposite directions from pin 2 to pin 7.  Changing which direction is first and second changes whether it is ramp up and down.  This causes pin 7 to output a pulse waveform.  The only stipulation is that with no resistance from the end of the second diode to pin 2 the sawtooth will be clipped.  This can be fixed by putting a ~4.7-6k resistor between the end of the diode and pin 2, however, this will change the frequency slightly.  So, either you get a bit of clipping on the sawtooth wave at the very tip top OR you get a slight change in frequency when you switch to sawtooth.

Either way, a dual OA circuit with 2 pots and one SPDT switch that can output a fairly amplitude stable triangle, square, ramp up, ramp down, and pulse waveform is quite nice

I will probably draw up a schematic and some notes later to replace the current schem that is on the site... I'll try to make it very obvious what you can/should tweak, and then puretube and gez can correct my mistakes  All I need now is a good way to soften the edges of the sawtooth + pulse waveforms without losing the stable amplitude.  I also think I will try to integrate an H11F3 or transistor of some sort to allow for a CV input for the frequency...  I could also probably get a wider frequency range with a transistor.

-Colin

P.S. GHN-how come it seems like in a lot of the threads you post it ends up with you and me bandying back and forth with our breadboard ideas?  I guess our thread choice matches!

[guitarhacknoise]

don't know..........
maybe someday I'll get more out of it than you.

I'll try your findings out later tonight, I think!
THANK YOU!
-matthias

[ExpAnonColin]

I've wired up an H11F3 as a variable resistor with a CV in for the LED.  The resistance range is HUGE...  I made it so that it couldn't quite get to full brightness, but it's something like 20meg->5k... super slow to super fast.  And it has a CV in, so I'm thinking this is a very good idea... now let's see if I can do it with a 2n3904 to make it useful to everyone.

Here's something I can't understand.

It's amplitude stable under DC... but not AC.  I'm sure there's an obvious explanation for this... right?

Also, the triangle wave's upslope is a fair amount shorter than the downslope... maybe 70% of the size.  Weird...

-Colin

[guitarhacknoise]

Also, the triangle wave's upslope is a fair amount shorter than the downslope... maybe 70% of the size. Weird...

does this have something to do with the Vb, not sure but I think that it needs to be exactly half of Vs for an even up and down?
-matthias

[ExpAnonColin]

Also, the triangle wave's upslope is a fair amount shorter than the downslope... maybe 70% of the size. Weird...

does this have something to do with the Vb, not sure but I think that it needs to be exactly half of Vs for an even up and down?
-matthias

Vs=8.82
Vb=4.42
It must have to be EXACT if that's the case... I can't imagine being .01v off would create that much of a difference... that's about .1%.

-Colin

[ExpAnonColin]

Late night thought... they should rename the ramp up waveform to the sisyphus waveform... it's trying to get up but it always falls back down.

-Colin

[gez]

Here's something I can't understand.

It's amplitude stable under DC... but not AC.  I'm sure there's an obvious explanation for this... right?

Also, the triangle wave's upslope is a fair amount shorter than the downslope... maybe 70% of the size.  Weird...

-Colin

Always use the DC input of a scope if you can.  The AC input is coupled through a cap which, at very low frequencies, has reactance (despite what the manuals say) so gives false readings.

PS The dual OTA VCOs are better/easier to control and aren't that much different in their operation (LM13700 data sheet).

[guitarhacknoise]

LOL!
Is that the corinthian dude who had to roll that bolder up the hill in hades?
ahh............I feel like that guy today!
spent the night messing with a lil 386 amp that just won't do its job!
I know it's me, cuz it worked right the first time and then I started swapin' components and a buffer section and a bigmuff tone stack and now it won't work right even when I stripped it back down to just the chip, but its not the chip cuz I switched it out.................it's just all gatey
and I know this has nothing to do with the thread, It's late and I just started rambling!
goodnight and merry xmas!
-matthias

[gez]

Also, the triangle wave's upslope is a fair amount shorter than the downslope... maybe 70% of the size.  Weird...

Disconnect the pot used for mixing, to see if it's still causing loading problems, and take a reading from the triangle output.  Amplitude should be rock solid.  Use the DC input of your scope.

The amplitude of both sides of a triangle has to be the same!    I take it you mean the times it takes to ramp up and down are different?  Could be the asymmetry of your op-amp's output.  Easily corrected with a trimpot and diodes (or subbing in a CMOS amp).

Regards

Scrooge

[ExpAnonColin]

LOL!
Is that the corinthian dude who had to roll that bolder up the hill in hades?
ahh............I feel like that guy today!
spent the night messing with a lil 386 amp that just won't do its job!
I know it's me, cuz it worked right the first time and then I started swapin' components and a buffer section and a bigmuff tone stack and now it won't work right even when I stripped it back down to just the chip, but its not the chip cuz I switched it out.................it's just all gatey
and I know this has nothing to do with the thread, It's late and I just started rambling!
goodnight and merry xmas!
-matthias

Sisyphus is not the kind of guy you want to be

I find with the LM386 it really helps to not omit any components at all from the datasheet schem...  even those little caps and resistors they put everywhere.

-Colin

[ExpAnonColin]

Some more findings from last night...  You only need one diode in either direction for the slope.  

Always use the DC input of a scope if you can. The AC input is coupled through a cap which, at very low frequencies, has reactance (despite what the manuals say) so gives false readings.

PS The dual OTA VCOs are better/easier to control and aren't that much different in their operation (LM13700 data sheet).

Disconnect the pot used for mixing, to see if it's still causing loading problems, and take a reading from the triangle output.  Amplitude should be rock solid.  Use the DC input of your scope.

The amplitude of both sides of a triangle has to be the same!    I take it you mean the times it takes to ramp up and down are different?  Could be the asymmetry of your op-amp's output.  Easily corrected with a trimpot and diodes (or subbing in a CMOS amp).

Regards

Scrooge

Concerning the scope... thanks for letting me know, that definitely makes sense.

I think the switch was loading the triangle wave (read, this is a TERRIBLE switch I'm using) because completely disconnecting it resolved it pretty much...  I think maybe I will wire up a pot instead of an SPDT switch, then I can blend in the ramp's and pulse waveform and make sure the symmetry is the same.

Thanks for the tip on the LM13700, but if you ask me, this H11F3 system I have wired up is amazing...  The range is like 1:1000.

-Colin