Invert LFO

[armdnrdy]

I just built a FL9 for a friend's son.
The sweep has a pretty good rise to the highest frequencies then it turns around and makes it's way back down. All of this is good, but when it reaches the bottom it seems to turn around too quickly.
It would be more pleasing to the ear if the LFO wave was inverted, spending more time at the bottom.

Is there a quick and dirty way to "flip" this LFO?

Here is an image of the existing LFO:



Here is the LFO section:

[samhay]

That might be, at least in part, because the 358 is single supply. Rather than inverting the LFO, which can be a bit tricky, you could try a couple of quick fixes. I'm just gessing here, but to make it more symetrical you could try clipping the LFO output with LED(s) and or, you could try playing with R58/59 to putmove the middle of the LFO around.

[Fender3D]

Where did you probe that signal?
If it's @ Widht lug 3, you may try adding an inverting op amp...

[armdnrdy]

Hi Federico,

I was testing the dual matched NPN that I made, so I probed it at Q8 collector that's connected to the clock. I was thinking about adding an inverting op amp but then I thought.....that is so obvious.....would that actually work?

@Sam
Thanks for the reply,
The schematic isn't exactly correct. I drew it up about a year ago and I replaced the original TL022 with another low current draw dual op amp (358 a bit easier to source) but when I built it....the 358 wouldn't oscillate in this circuit......so I changed it out to a TL022.

[samhay]

Ok - in that case, try a DC-coupled Inventing op-amp with gain <= 1 between LFO and width pot.

Edit - that should have been inverting, but inventing is growing on me.

[armdnrdy]

Ok - in that case, try a DC-coupled Inventing op-amp with gain <= 1 between LFO and width pot.

Okay,
Now this is where it gets a little deeper.
Do you think since we're inverting the LFO in that part of the circuit that it will affect the manual control?

[armdnrdy]

I remembered that there is the capability of switching between a normal LFO and inverted LFO in the Mutron Biphase design.



It looks like the LFO output switches in/out an inverting op amp.

Unfortunately, this board is already made, populated, wired, and I'm just waiting for the painted enclosure to dry so I can box it up before Saturday. (the going away party)

So, I would have to make a daughter board and add some jumpers.

[johngreene]

I don't know if you are aware but the type of oscillator your are using on the MN3201 already has a hypertriangular response if driven from a linear slope (i.e. triangle wave). Plot the voltage input vs frequency of the clock and you will see what I mean.
It is also why it is very important to properly center this kind of VCO.

--john

[Fender3D]

What if you add M117's Q2 between Q7 and half Q8?
Simulating that stuff and tinkering with parts, I remembered one point where clock was reversed (ie positive LFO = clock lowering), but I don't remember what I did anymore... 

[samhay]

Unfortunately, this board is already made, populated, wired, and I'm just waiting for the painted enclosure to dry so I can box it up before Saturday. (the going away party)

So, I would have to make a daughter board and add some jumpers.

Then I would try clipping the bottom of the LFO with an LED. You will lose a little depth, but I would imagine it might make for an acceptable trade off.

[armdnrdy]

I whipped up this drawing based off of the Biphase.

I have a question though.
the Biphase is powered by a bipolar supply and the FL9 powered by a single.

I made the appropriate change to the non inverting input, (added ref voltage instead of ground) but I'm not sure what to do with the 47K pull down in the "reverse" section. Should R6 be pulled down to ground or voltage reference?



@Sam
I'm going to breadboard this (may have to tweek a few things) and if I like the results I'll see if I can rush a daughter board. If not, I'll try messing with the existing LFO as you suggested.

I kept the Biphase switch in the circuit because I was thinking that the original sweep may sound better at faster speeds.

[armdnrdy]

Okay...
I drilled a few holes on the board, added a few wires, and bread boarded the above circuit and ended up with a reverse sawtooth wave when I flip the switch. The original wave remains unchanged in the "other" switch position.

Now I know some people pay extra for this but I was just aiming to invert the original sweep.

Any ideas?

[samhay]

That looks like a cool feature. Maybe your input impedance is too low? Try increasing R1 and R2 by 10 x or so.

[Fender3D]

Why didn't you attach positive input to LFO reference voltage (358 pin 5) ?

[samhay]

Why didn't you attach positive input to LFO reference voltage (358 pin 5) ?

Maybe to save drilling another hole in the PCB?

Another suggestion. Why not place the inverting stage between the LFO and width pot - it will buffer the LFO from the pot.

[armdnrdy]

Why didn't you attach positive input to LFO reference voltage (358 pin 5) ?

Hey Federico,

After I posted the above drawing, I looked at the FL9 circuit to see where I was going to connect this circuit when I noticed the LFO reference going to pin 5.
I drilled a hole, added a wire, and used the reference that's feeding pin 5 for the "inverting" section.

I tried using different higher and lower values (10K-100K) in the inverting section but saw no change in shape.

I inserted the inverted section exactly where it's implemented in the Bi Phase, connected to the depth pot wiper.
I used the same 5 volt supply that's feeding power for the TL022 in the LFO instead of 9 volt which powers the rest of the circuit.

I don't know....I'm running out of time to experiment!
Any other ideas?

[samhay]

If you scope the LFO with the inventing op-amp connected, does it look the same as without?

[armdnrdy]

If you scope the LFO with the inventing op-amp connected, does it look the same as without?

The last shot I posted is of the circuit with the "inverted switch" in the inverted position. When I change the switch to the normal position the scope image looks exactly like the first image of the inverted hyper triangular wave that I first posted in this thread.

[samhay]

OK. What is R6 doing - that will be loading down the LFO quite significantly? Oh - and unless you have AC coupled this, then R6 should be going to your 4.5V point, not ground. If you need it, put a big cap between it and ground.

[armdnrdy]

OK. What is R6 doing - that will be loading down the LFO quite significantly? Oh - and unless you have AC coupled this, then R6 should be going to your 4.5V point, not ground. If you need it, put a big cap between it and ground.

Since I was unsure, and since the Biphase's ground would be equal to the FL9's reference voltage, I tried R6 to Vref first. then I tried it to ground....no difference.

I wonder if it would make a difference to feed the op amp with 9 volts and use a 4.5 reference. I followed the 5 volt scheme that was going on in the FL9 LFO/clock section.