Posted: Sun Jan 31, 2016 9:07 am Post subject:
Quad triangle VCO with gate Subject description: a versatile building block
Here is a simple but versatile little building block that I am very happy with. It is a single-supply symmetrical triangle waveform VCO with only a few parts, but nice characteristics.
The CV input range goes from Vdd down to 0V. The frequency range goes from 2.2KHz down to about 40Hz. It goes even lower, but then Vpp starts to decrease to about half value.
When a quad Opamp like the LM324 is used, four VCOs can be built with only three chips. I plan to use this circuit in a micro Lunetta board that I am currently soldering.
Triangle VCO with gate.JPG
Description:
Filesize:
62.53 KB
Viewed:
1885 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
The circuit can be improved a bit when operated at Vdd=5V. If a 74HC132 is used instead of the 4093, the hysteresis of the Schmitt trigger and the triangle wave amplitude will be bigger.
The output is then a triangle wave with Vmin=1.88V and Vpp=1.32V. However, the 4093 behaves a bit better. The 74HC132 shows frequent overshots, but they are not audible. Last edited by synaesthesia on Tue Feb 09, 2016 7:25 am; edited 1 time in total
Several of the above triangle VCOs (one as LFO driving another VCO, one for PWM) sent into a resonant low pass filter.
Sorry for the bad sound quality. This was recorded using my smart phone and an inadequate small speaker.
Here is a major improvement at the cost of three additional resistors per oscillator: To increase the amplitude of the triangle wave output, the input to the Schmitt Trigger is shifted and scaled using a voltage divider (idea courtesy of elektrouwe). That generates a much bigger amplitude swing of Vpp=3V when operated at 5V power. The wide CV input range is not affected. The ratio of R4/R5,R6 is important here. R4 must be no bigger than 1.5 times the value of R5 and R6. I am using a 4093 for this version again.
EDIT: Just tried another 4093 chip and found that the resistor network at the input depends on the IC manufacturer and age of the 4093 . If the 33K/22K/22K combination doesn't work, try 47K/100K/100K and experiment a bit to find the optimum for the highest amplitude. I found good values for all chips I tested that will produce a triangle wave amplitude between 2.8 and 3.0 Volts.
VCO2.JPG
Description:
Filesize:
52.51 KB
Viewed:
665 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
Last edited by synaesthesia on Wed Mar 16, 2016 2:09 pm; edited 2 times in total
Joined: Jan 02, 2016 Posts: 217 Location: Scotland
Posted: Mon Feb 15, 2016 3:41 pm Post subject:
At first I built my circuit with a 4093, but was likewise disappointed with the small hysteresis of the 4093. That was the rationale for building a Schmitt trigger with a buffer ( the XOR is just acting as a buffer ) with external resistors to set the size of the hysteresis, with the values in the schematic that I posted the sawtooth was around 5 volts peak to peak. The second XOR gate is wired as an inverter. I may try synaesthesias's idea of using the three resistors at the input of the 4093 to widen the hysteresis.
I also reckon that you could get away with replacing the 4066 gate with a small N-Mosfet such as a 2n7000 or a bs170. That being said a 4066 is probably cheaper than 4 MOSFETs!
They all show CV in (VCOin is a misnomer) as in control voltage. So apply a voltage here from the wiper (2) of your pot. Put the 3 up to supply and 1 to ground. Gate will be tied to supply for a signal output.
(Or research mixing CV and a pot with an opamp virtual mixer and an opamp inverter. Further research will reveal an opamp comparator being useful on your gate input.)
Ideal candidates for you to breadboard and observe working before you commit to solder.
You can also mix a control voltage with a resistor network (e.g. 10K, 22K, 47K, 100K, plus 10K pull-up) driven by some logic outputs (e.g. from a counter) for a simple melody generator.
Joined: Jan 02, 2016 Posts: 217 Location: Scotland
Posted: Tue Jun 07, 2016 2:39 am Post subject:
synaesthesia wrote:
You can also mix a control voltage with a resistor network (e.g. 10K, 22K, 47K, 100K, plus 10K pull-up) driven by some logic outputs (e.g. from a counter) for a simple melody generator.
This will not result in notes on the equal tempered scale, for that you'd need an expo converter, but then again that probably wouldn't be in the spirit of Lunetta.
Joined: Sep 26, 2016 Posts: 134 Location: norwich, uk
Posted: Mon Nov 07, 2016 8:32 am Post subject:
This looks like a good optionfor the oscillators on the synth im building. To get the triangles full range i suppose it would just be a case of another op amp or transitor to give it a boost. Has anyone done this yet?
Joined: Jan 02, 2016 Posts: 217 Location: Scotland
Posted: Tue Feb 07, 2017 4:44 pm Post subject:
VCO behaviour
A couple of other comments on these VCO's they will work with CV all the way down to zero, but not reliably below around 200mV ( 0.2 V ) in the other direction it is Ok for the CV to exceed the +Ve supply rail by 2*Vs-3 ( such that the opamp inputs don't exceed Vs - 1V5 )
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum