When you need to implement voltage control into a circuit, what’s your preferred method? Any particular chips or circuits you like? Most effective or efficient all-purpose method? Is the LM13700 all it’s cracked up to be? Best practices or mistakes to avoid?
Depends on the circuit.
Vactrols are quick and dirty but man, dirty. Slow and imprecise. Particularly problematic if you try to drive one directly with a CV instead of an op amp driver. Non exponential unless you build an exponential converter, and then it’s not quick and dirty any more. But there are situations where they’re just the thing, when you don’t need fast response (it’s even a feature, not a bug, sometimes) or great precision.
LM13700 is very useful for this, but of course the DIP version is out of production and getting harder to find (genuine ones anyway).
For my Barton/AO 4046 VC Wave Shaper I needed 7 voltage controlled amplifiers — except the signals were pulse waves, so I could do something equivalent to a VCA using nothing but 1/4 of an analog switch and a resistor. And an op amp as it turned out, it needed buffering.
See also
1 Like
I was looking at the SOIC of the 13700 and considering throwing it onto a board with some other SMT bits to make a ready-to-use, just-add-potentiometer drop-in VCA “chip”. Not sure how widely applicable that would be though.
I tend to use AS3364 for VCAs, very few external components required, 4 in a can and easy to control.
2 Likes
Not my preferred method but one that I experimented with a while ago (successfully I might say) is this:
Personally, I do not so much like the OTA solution because of the very low voltages (about 50mV) you have to bring your input high amplitude signal to (normally this may be up to 5V which is 100 times more!) before feeding them to the OTA the result of which you have to crank up again afterwards. This is bound to introduce noise into the circuit.
If you are looking for alternatives to the 13700-and-such-like-approach you may be interested in taking a peak at the NE570 and TDA1524.
TDA1524 is listed as obsolete and not available at Mouser. It’s described as a “Stereo-tone/volume control circuit” while the NE570 is a compandor. If there’s a way to make these function as a voltage or current controlled resistor I’d be interested in seeing an article about that.
There are numerous voltage controlled amplifier chips available; another is the THAT 2180 which was used in the GMNC Pure VCA. But the OP’s query was about designing circuits to have voltage control, which is a more general question than about just VCAs.
What I’m trying to do is wrap my head around the different methods of creating a voltage control element in a circuit, so I can wrap my head around the concept. I know there are different ways to get the same result, but there are also different results that are required in different situations, so I’m trying understand that as well.
So, not only the different methods of designing CV into a circuit, but also the different applications of CV that require different solutions, ie: a voltage divider isn’t always the answer, so what kind of solution would you need instead?
If you are designing the circuit from scratch then you can design with voltage control in mind, thats the norm, If you are modding a circuit to be voltage controlled then that’s another story, much harder to accomplish. I’m struggling to understand exactly what your goals are here.
Fortunately Mouser is not the only source of electronic components, and there are many components (e.g. tubes, bucket brigade delay devices, vactrols (are or will to be banned in the EU) being used in audio electronics that are obsolete. So I wouldn’t care so much about that. And hasn’t the 13700 discussed above also gone extinct? That is not stopping most people who are building a well known VCF, is it? The NE570 is used in guitar pedals amongst others. Have a look at the ‘Stomp Box Cook Book’. It may not lead to a linear control but that could make it just the more interesting a device. The TDA1524 has CV controlled filters and a CV controlled volume / amplifier on board. Those can be used in conjunction or separate. So why not use these CV controlled capabilities (or parts thereof) for something else than ‘stereo-volume/control’?
1 Like
To quote analogoutput: quick and dirty - vactrols.
I have stocked up on 13600/13700 and have some LM3080, just in case. Also some MFC6040.
But if I am experimenting a vactrol is my first choice.
I guess I’m trying to understand all the options available and what kind of voltage control they provide. Like, you ask a question like this and you get answers like, “well, it depends on what you’re trying to control”, but I still don’t understand exactly what the different needs of various voltage controlled circuits are and how the CV part of the circuit changes based on those needs.
Like, I understand how most of the electronics in this hobby work, but I just can’t seem to get my head around control voltage.
For instance, this is a simple single stage RC filter:
The cutoff frequency is 1/(2πRC) and it’s fixed in this case. If you replaced the fixed resistor with a potentiometer you could vary the cutoff manually.
If you replaced it with a vactrol you could control it with a control voltage.
This is sort of a voltage controlled filter. The vactrol’s a pretty bad way to make a general purpose VCF but if you only send 0 V or +5 V to change the filter between a cutoff below the audio frequency range and a cutoff above that range then you can use it to turn an audio signal on and off, and that is a passive low pass gate. It works because the CV changes the brightness of the LED in the vactrol, which changes the resistance of the LDR, which is like changing the pot to change the cutoff frequency.
This is a 2-stage filter:
There are some op amps for buffering and a feedback path for resonance, but at the core of it are two RC filters with a stereo pot to control the cutoff frequency of both.
If you replace the resistors with two OTAs (CA3080, or the two halves of an LM13700) and associated resistors, and you can remove a buffer, you get something about like this:
A resistor connected to a voltage source gives you a current proportional to the voltage. So does an OTA, but the proportionality constant depends on a control current (coming in here from the top through the uppermost 10k resistors). So it can act like a current controlled variable resistor. (Or voltage controlled, because in the section above this there’s circuitry that converts a linear voltage to an exponential current.) Again it works because the “resistor value” is changed by the control voltage, and that changes the filter cutoff frequency.
It’s a better solution than the vactrol for a VCF because it’s more precisely controllable, and it responds faster. On the other hand it’s much more complicated, and if all you need is something to turn a signal on and off a vactrol, as in a low pass gate, may be a better choice.
Here’s a simple oscillator:
The output voltage of the 40106 produces a current through the pot and fixed resistor which charges or discharges the capacitor. But the cap can instead be charged by a current from a transistor connected to a control voltage:
It’s a simplistic sort of voltage control and does not make an outstanding voltage controlled oscillator, but apparently it’s pretty good considering how simple it is. Once again it replaces a fixed or manually controlled current (via a resistor) with a variable current source based on a control voltage.
6 Likes
Thanks for the this! I understood most of this, at least the concept (ie: an applied voltage changes the state of a component, which acts as a variable resistor or kind of valve, depending, and that replaces the pot in the circuit. The examples were extremely helpful, and I think that was what I am looking for.
One thing I’ve struggled to understand since I started with audio electronics recently is how a component like a resistor “creates” a current in a circuit. All the lessons I remember tell you that voltage and current are both simultaneously in a circuit. I mean, they’re proportional–if you don’t have one, the other one also can’t exist in the circuit. But people constantly talk about voltage being converted to current, and I think I just don’t get understand the concept. Like, I = V/R, so with a wire connecting a positive voltage source to ground, if you give the wire a nominal resistance of 1ohm, there no resistor there to “convert” the voltage to current but you still have 5A of current on the “circuit”. Now, I understand that this is why short circuits are bad and you always want resistance between the circuit and ground, and 5A is way more than you’d want an IC to receive, but for the purposes of the math, if you have voltage and no resistance, you still get current–lots of it. So how does resistance (or anything else) “convert” voltage to current when, if you have current in a section of a circuit, then legally (Ohm’s Law) speaking, there’s voltage there as well?
I’m beginning to think that the answer to this questions is going to be “it’s complicated”, because you have to consider the equivalent resistance of the rest of the circuit, if it’s ever more complicated than just one wire and one resistor. But if the resistance of an op-amp is at least conceptually infinite (open circuit), wouldn’t the voltage there be zero?
In case you’re wondering, though it’s probably obvious, the reason I’m an IT guy now and not an embedded systems engineer boils down to Math. Circuit approximations and complex equations and multivariable calculus were all just too much for my brain to handle.
If you have a voltage source not connected to a load then you have a voltage with no current. Put a resistor to ground and now there is a current. How much current depends on the resistance.
If you have a current source connected to a current sink you have a current with no voltage. Put a resistor between and now there is voltage. How much voltage depends on the resistance.
Granted voltage and current sources are idealizations but they are useful ways to think of things.
4 Likes
