Pribor Neofit - super simple CMOS synth

[priborpedals]

Hello! Here is my first post here. Sorry if something is not as well as it should as my poor english.

When building simple beating infrabass synth voice I found something different. I didn't found the exact project so I decided to share it and I think Lunettas forum is a best place to do that. It have very small amount of parts and I hope that it could be a nice entry level project.

Main idea was very simple - two square wave oscillators multiplied by MS20-style ring modulator. I started with 4049 inverter and 4093 NAND Schmitt trigger just because I had a bucket of both. Schematics was almost the same as if final circuit. But raw multiplied signal is pulse 9V peak-to-peak and I found that simplest way to make it more musical is to put an integrator at the output. This made sound more interesting than simple low pass filter and 100nF capacitor works there very nice but possible to experiment with different values to change triangle shape of output signal. 47nF filtering cap used to cut high frequency noise. Also I decided to make oscillators with different range and coarse/fine tune controls because in low frequencies sound is like amplitude modulated signal but when both oscillators tuned in close frequencies beating sound appears. And funniest thing I discovered than when power off sound didn't disappear simultaneously but frequency and shape is changing when power cam is discharging because of circuit requires very low current so I added voltage drop control also. I used Tesla MHB4049 and ST HCF4093BE and maybe voltage drop will act different with different manufacturers ICs so could be very interesting if anyone will experiment with.

I built that synth on stripboard - maybe layout is not perfect but well fits to 1590B.


Video with scope is here:
https://www.instagram.com/tv/CSb8GrZFk9U/

Hope you'll like it!
Thank you!

---

Pribor-Neofit-schematic.jpg
Description:	schematic
Filesize:	463.73 KB
Viewed:	267 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

---

Pribor Neofit layout.jpg
Description:	layout
Filesize:	285.19 KB
Viewed:	278 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.

---

[wakyct]

Cool, especially liked the sound around 7:20 in the video...nice plans too

[PHOBoS]

neat little box, nice idea with the integrator.
However, that inverter which has its output connected to ground might not be very happy. Especially if the input is also tied to ground
which means the output is trying to go high. It might have an interesting effect on the sound though and maybe it will survive. When
the current is limited it shouldn't be too bad but otherwise you might want to check if the chip is getting hot.

[priborpedals]

Reason to connect last inverter's both input or output to ground was to cancel a noise. Maybe with 100n bypass power capacitors and disconnected output is more traditional way to do it, but I like to make all as simple as possible and also i'm not sure that bypass caps will work as should with veroboard. Current is very small - circuit producing sound few seconds more when power off.

[Steveg]

You would connect the input to ground to prevent noise (otherwise the pin will act like an antenna picking up any stray signal and inverting it). But as PHOBoS points out connecting the output to ground is a really bad idea. Maybe you will get away with it with this chip in cold weather but I wouldn't bet on it in the general case.

[priborpedals]

I compared the sound when last inverter's output disconnected or connected to ground. When output is disconnected appering glitchy noises on both oscillators outputs, especially when voltage drops. When that output connected to ground waveform became more stable but i noticed a voltage drift between power pins with oscillators frequencies producing pseudo random clicks. No IC heatind was noticed all the time. I did'n measured a current consumption because I have no device to measure less than 10 mA.

I get some modern TI CD4049UBE and tested the circuit - sound became completely different and all the funny pseudo-random noise had lost. So I'm sure that that's all is because of previously used Tesla MHB4049 internal circuit. Unfoftunately I didn't found schematic of exact MHB4049 inverter unit but when compared some datasheets including TI 4049UBE (different datasheet revisions), RCA CD4049A, Fairchild CD4049UBC, Intersil CD4049UBMS and few more, all the internal circuits are different, mostly with diodes at input and positive terminals. Also MHB4049 could be a buffered inverter like Philips HEF4049B or Toshiba TC4049BP.

When I will check some more different 4049's I'll update the schematic.

[PHOBoS]

interesting that it becomes more glitchy with the output disconnected I'd expect the opposite. And yeah, manufacturer can make a difference
with some circuits, especially when you don't really use them as intended. Apparently it can even matter which gates of a chip you use.