40106 Osc question

[Jaicen_solo]

I've got a square wave oscillator made from a single inverter of a 40106 Hex Schmitt. Basically it's a 1M pot between pins 1&2, output from pin2 and a 0.1uF cap to ground from pin 1. The problem I have is two-fold really.
Firstly, to get the range of frequencies I need using the 0.1uF cap, I need a 1M pot. Is there a formula I can use to work out what value of cap will give the same range of frequencies using a 10K pot? This is because I've scavenged some sliders from an old dano eq to use and they have a value of 10K. Basically, the oscillator needs to be at 32Khz when the sliders are at 5K, and give a 10-20Khz range either side. Ideally it would go from 15-45Khz, is that too much to hope for with 10k?

Secondly, I need to have the control value for the oscillators be removable. When nothing's plugged into them, then they will give a steady 32kHz or so. I can't use break jacks either, there's no room! I'm just using a D-Sub connector.
I've no idea how to achieve this though!

[MR COFFEE]

10 K is a bit low for a 40106 oscillator. I think that's why nobody responded to your post.

[Jaicen_solo]

Yeah, that's pretty much what I thought. I think I'm going to go with a voltage controlled 555 oscillator, maybe that will do what I need.
I still can't figure out how to get the frequency up and down, and still have it at 33kHz when there's no control voltage/resistance. Any ideas?

[gez]

Are you trying to modulate it?  It would help if you can elaborate on what it is you're trying to do.

[Jaicen_solo]

Basically, I just want an oscillator that will give more or less a stable 33kHz square wave output. I then want to be able to feed it a control voltage which will raise or lower the pitch by around 10-15kHz either way. When no control voltage is present, I need the oscillator to be stable at 33kHz.

[gez]

Yeah, but what is the oscillator for?  Presumably you're clocking a BBD or something? 

What is the control voltage?  Envelope?  LFO? 

The fragment you've provided isn't enough to answer the question really.

The 40106 probably isn't up to what you want it to do.  Maybe a look at some dedicated VC oscillators would be a better idea?

[slacker]

You can make the frequency vary by adjusting the supply voltage to the 40106. You could do this with a LDR driven by an LED in series between V+ and pin 14, don't know if that would give you the accuracy you need though. There's probably other ways to control the supply voltage.
I think the formula for frequency is f = 1/RC.

[Jaicen_solo]

Surely the intended purpose is irrelevant, provided that the specs are given?? It's just a clock circuit for a digital ROM.

I thought about using an LED/LDR combo, but i'd prefer not to use them. I was thinking more along the lines of transistor CV. I can't modulate the supply voltage, as i'm using all six inverters on the 40106 seperately. I don't think it's possible to do what I'd like with simple electronics really, but I have the beginnings of an idea.

[gez]

Surely the intended purpose is irrelevant, provided that the specs are given?? 

Not at all, it gives us an idea as to whether the circuit you've chosen is suitable. 

I still can't say, but if I were you I'd scan over some analogue chorus/delay circuits and use the clock part of those.  You haven't said how you intend to use your control voltage, or what it is, so it's difficult to give further advice, but you could set the clock for a quiescent oscillating frequency then switch in the control leading to the clock as desired.  Are you using switching though?  Again, not enough info; had you given a little more you'd get a more precise answer.

With the simple oscillator you've chosen if you want to make it voltage controlled you're going to have to find a way of making the resistance variable.  LDR is one way to go, but I still think you're better off with a more complicated circuit.

[Jaicen_solo]

The problem I have is space, so i'm limited as to how complex I can go with the circuit.
I'm replacing the clock signals inside my TR-505 with oscillators made from a 40106. I know it sounds too simple to work, but when I was testing it worked fine so I decided to stick with it, rather that go with 6 555's as originally intended.
The thing is circuit bent, with the bends all on a D-sub 25 way connector at the rear of the unit, so it looks more or less stock. I don't want to drill the case to fit knobs, so I want the oscillator controls to connect to the socket. Unfortunately, if i put the osc's in an external box, it won't work normally without the 'bend console' as it does now.
I'm pretty certain that there's no way to toggle the control resistors in and out simply. I considered using a transistor as a switch, ie:
plugging in the bend console grounds the base which turns off a transistor which is connected to a fixed resistor to give 33kHz. The bend console can then connect a variable resistor in parallel with the fixed resistor.

Unfortunately, I can't see a way to get this to work, as the transistor will need to be grounded at the emitter, which will mess with the oscillator, if that makes sense?? I think i'll just have to go with external oscillators and have the bend console permanently connected.

[gez]

I'm replacing the clock signals inside my TR-505 with oscillators made from a 40106.

The original clock might be the thing to look at.  It'll probably have a simple RC combo to set the clock rate.  Replacing the resistor with a pot would give you your variance.  Use a third hand and you have foot control.  Might be one way to get round the problem of space?

PS  my post should have read chorus/flange (not delay).  Not that it matters! 

[slacker]

what about using a transistor as a variable resistor like in the Music from outer space WSG http://www.musicfromouterspace.com/analogsynth/YOUR_FIRST_SYNTH/your_first_synth.gif
I guess you could bias the transistor via an internal trimpot to get your fixed frequency then apply a control voltage to the base to vary it, again I don't if that would give you the precision you're after.

[Jaicen_solo]

Whilst that's a good idea in principle since it's a simple inverter osc itself, the 505 has a single clock fed to all the oscillators, so modifying that would pitch all the instruments. My design gives variable pitch to all of them. Also, it doesn't get around the problem of the removable console.
With regards to the WSG, I believe that the transistor is just mixing in the output of a second inverter. The variable resistance is still just a pot (R5 & R17).

[slacker]

you could be right but to me it looks like U1-A, R11, C5 and the 2N3904 make up an oscillator.

[Jaicen_solo]

To tell the truth, i'm not entirely sure what's going on really! Having given it more than a cursory glance I can see that my initial reaction is completely wrong.
It doesn't look like it's set up as an amplifier or oscillator, as there's no feedback resistor between collector and base, and there doesn't appear to be any DC bias on the emitter.
My guess is that U1-C is set up as a square wave LFO, which feeds the base of the transistor, this turns it on and off. I imagine that when the transistor is on then R11 is shorted, which is pretty much what i'd described in an earlier post. I just couldn't get my head around how it works.
If it does indeed act as a variable resistor, then it could be just what i've been looking for. I'll easily be able to squeeze 6x transistors and a 40106 into the 505.

[slacker]

yeah it could just be a switch turning on and off at the zany frequency, but the zaniness amount control looks like a "depth" control which infers some sort of variabilty to the effect.

[Jaicen_solo]

Yeah, the amount is a crude sort of depth control. U1-C will be oscillating from 0V to 2/3rds of the supply voltage, in this case 0-6v.
The depth control is a voltage divider which bleeds off some of the voltage to ground, resulting in higher resistance from the transistor.
I've got a design for something that just might work actually, i'm just hosting it so it'll be online in a minute or two.

[Jaicen_solo]

Here's the schematic i've worked up.
It might need a small limiting resistor somewhere but it's just a rough schem.
The idea is that when nothing is connected to the EXT input, the pull-down resistor will hold the transistor at roughly the right resistance to keep the osc at 33kHz. The external input will be a 100k pot connected from gnd to +9v, so the pull-down resistor will need to be at least 200K.
I'm not sure how it will work with the transistor, but that should allow me to get a range of frequencies higher and lower than 33kHz.
What do you guys think??
EDIT: Pulldown will be a trim pot most likely, so I can get as near as possible to the original sounds. I've worked real hard to keep my 505 looking and sounding as stock as possible when not bent!

[slacker]

like you said, it might need some tweaking but it's worth a try.

[Jaicen_solo]

Aye it is. I'll try it as soon as I get back home next week.