Everyone in the history of synthesizers starts with an oscillator so of course you start with a noise source just to be different. 😏 I approve of this course of action.
Try a reverse biased transistor base emitter junction as well, as that also will break down around 7V, and give similar noise. You can even use TTL gates for this, abusing the input structures and output structures, with only 2 connections, to get a noise source. For higher more impulsive noise use a RF transistor as amplifier, and connect diode direct to base, without breadboard capacitance, which will reduce that initial falling time a lot.
@@paulcohen1555 Either way will work, though shorting to base can reduce external pickup of noise, especially if you use a TO can device where you can ground the can and collector.
I just built the white noise generator and it works great. The only changes are that I used 9v supply, a 2.2k resistor instead of the 18k resistor, and an 8.2v zener instead of a 8.6v zener. It took a little bit of trial and error to find out a good replacement resistor value, but it was still pretty easy. Thanks for the circuit idea
You're awesome, this is exactly my cup of tea The video hasn't started yet but besides this channel being great this is exactly what I've been doing/inventing over the past years. Great fit to my world, exactly what I love and live.
Also awesome to see the noise shape, almost like the breakdown curve of the diode before it resets, awesome. It's like seeing the breakdown curve real life
Awesome order as well I've been building oscillators and effects for years and everytime I have a massive struggle hooking it up to something to test the sound😂😂 And then I wanna build a test speaker but before I get started think of another crazy effect and end up with wires everywhere again stretching from breadboards to random amps😂
I have a feeling this is going to be a fantastic series of videos. I've been working on my analogue synth build since Christmas, and with an arduino based midi to CV converter it's now at a point where I can get something vaguely resembling music out of it. Really looking forward to seeing what you come up with....
Yes, I've been into it for a year now. My second project was a MIDI to CV converter as well. Just to save you some frustration. Forget to use the inbuilt DAC. Even though my microcontroller was not an Arduino but a PIC, I still can tell you, that you need a more linear DAC.
@@PermireFabrica yes I have 2x MCP4822, that gives 4 voltage control channels that have 12 bit resolution. I have voltage outputs for note CV, pitch bend, mod wheel and velocity.
@@brianharper9798 okay nice. Quite a powerful module you've built! Mine is only dual channel and only gate and CV output. But one channel can be set up as a 16-step sequencer.
Calculating is cheaper than soldering. If you want to have 19 different oscillators for DIY sounds take a look to the Synthesizer Workstation: th-cam.com/video/oI_mTEu3WqE/w-d-xo.html
Peter Blasser creator and owner of Ciat Lonbarde uses this technique in his: Tocante Zenert device, Peter Blasser was the Student of Don Buchla, the co inventor of the VCO. Cool for showing how to create noise with a Zener, once i did a search on that, but nothing found. I got "alot" of zeners, but not that one.
I can relate to you like everyday I come up with new designs and creative ways to build amplifiers. And sometimes I have dreams about how I would like to build and design amplifier. It is a passion for sure. 🤗💯💖 Thank you for sharing yours 😁 I just love creativity and unique designs think outside the box. Once you understand how amplifiers are built and how solid state chip amps and ,op-amps ,transistors and so forth work then you can get creative and build your own schematics for building amplifiers different from what's already out there. I tend to keep my schematics to myself though for personal use only. 🤗 But not to say I won't share them and maybe sell some one day but for now it is just for fun.
I love your circuits, say could do a video using a diode OP-AMP Log Amplifier and how to plot the input and output Voltages and how the op-amp generates a log output thanks
"Moog" pronounced like "vogue". You and I should chat since I just started on a similar trajectory, also op-amps, also those light jumper wires (but an analog computer-like learning-PCB, not a synth). My casual interest in music synths suggests pink noise is preferable? I suppose there is an easy filter you can add to get it pink.....
Sweet series. I have a new scope on the way that is somewhat respectable and I have thr lm386 circuit already breadboarded from something I was messing either last week. Is it better if I just build the output on some veroboard?
Does that lm386 have a 1w output or a 700mw output? Just wondering which one you are using 😁 of course the 500mw lm386 would be better suitable for a preamplifier stage. Almost all the preamplifier op-amps that I have looked at on a data sheet usually range from 135mw 250mw, 350mw, 500mw and 625mw. But the 700mw and the 1watt lm386 are great for pushing a 2 watt or 1.5 watt 8 ohm speaker. But I have used the 1watt lm386 op-amp as a preamp amplifier but you have to used a trim pot or a resistor on the output of the 1watt lm386 before connecting it to a amplifier stage to get great sound quality control. I have also used the 1watt lm386 op-amp to make a distortion pedal at full tilt or cranked all the way up it'll get really distorted. No need for germanium diodes or LED lights to create distortion😂 You can use a 10K potentiometer along with a 10uf capacitor across pin 1 and 8 of the lm386 to control the gain.
Check out the 'teenage engineering pocket operator modular 400' Not great for learning the design, but very educational with how to interact with each block
Wouldn't the open loop LM386 work like a comparator and completely saturate into his 12V rail with any voltage over 0V? I can imagine it would look like a square wave at the output? (There is maybe a low pass filter on the output, maybe not a super square wave) And which 47µF capacitors do you recommend using for audio circuits? There are a lot of rumours on the internet about "audio capacitors" :/
I love the system building bock ... And to see what sounds you will make... May a Theremin...can be inserted...that you make......some kind of VCO....to increase and decrease the frequency...of IMSAI Guy voice... At first when I looked at LM386....I was expecting some negative feedback... But it looks like their are some gain set pins...so maybe this circuit has a fixed gain...of 20?...default.. I will go look...at your circuit Nice white noise source...simple ... Good job...
I believe the LM386 has internal feedback as it's meant specifically for audio amplification and thus the huge internal gain of a typical opamp isn't needed here. The datasheet shows an extremely simplified version of the internal schematic which has gain control circuitry. By default the gain is 20, and with a capacitor between pins 1 and 8 (as you can see here, disconnected) and a cap on pin 7 you can change the gain
Lm386 ist not a opamp. It is a audio amplifier for small speaker. For sure it can be used for some other things. On the other Side this ist not true for all speaker amps The classic tda2030 ist more like a Power opamp😉
I'm having trouble replicating your noise with your circuit, using a 2N3904. It looks like you have about 200mv spikes. I see nothing (just ambient noise), connected to the collector of the transistor. Does it have to be hooked up to a load? Maybe the scope has to be set up a certain way? I would love to see more detail on that.
@@IMSAIGuy Thanks, that helps -- I'm seeing the ramp shapes now, but it doesn't look like "hair" like yours -- there's no "floor" to it, it seems to be riding on other random AC noise of 30-70mV, even if I use a 12V battery for the supply. Maybe it's just the cheap zener I'm using, or just way too much ambient room noise.
@@IMSAIGuy I was using a 1GHz scope, tried limiting BW and also tried a 100MHz scope. I think I figured out that most of the noise was due to using long leads to the power supply (or battery) picking up RF, looks like mostly around 100MHz (FM stations?). Thanks for the help!
Does it matter what type of NPN you use in the noise generator? Using a 2N2222 I can't get noise out until the Zenner value (+ NPN Voltage Drop) is less than 0.3v lower than VCC (ie 11.7v -12v). An 11.1v Zenner worked but anything less generated no noise at all. Tried at lower voltages (VCC) and got the same result using lower voltage Zenners.
you need avalanche, lower voltage zener diodes zener, higher voltage zeners actually avalanche . the avalanche voltages of NPNs will vary. so best to use the 3904
@@IMSAIGuy Hmmm. Put in a 3904 and no change. With your circuit, except using a 8.2v Zenner - 1N4738A (since I don't have an 8.6v Zenner) - No noise - no output at all. OK, reading further down the comments I see connecting the 18k to the Collector rather than 12v is a thing and now I get 1v PP white noise.
@@davidh1187 it is sometime picky about the layout, you try the circuit in: th-cam.com/video/Ow9zQXPYm3g/w-d-xo.html it is almost the same but on resistor is tied to the collector instead of +12
Depends a lot on the transistor amplifier and the layout, low capacitance diode and RF transistor you get up to GHz noise out with common components. RF noise source design is hard, you have to use really low capacitance and make large air gaps to reduce stray capacitance, and short leads as well to lower inductance.
@@paulcohen1555 You would probably be better off using a RF transistor in reverse bias there, smaller junction and lower inductance will make the wideband noise figure better. Plus breakdown better defined as between 3 and 10V, as opposed to the 1N34 being anything from 20 to 60V.
LM386's are unaccountably popular. They hiss. They oscillate, particularly if you're at all careless with layout. They sound like crap at best. I actually get better results with the venerable LM380. You can put a couple of big copper foil heatsinks on it to get more power, it's more stable and quieter, you can wire a pair in push-pull for even more power out and drive a fair-sized speaker if it's reasonably efficient. If it's an actual audio circuit rather than a synth, NE5532 pairs well with it as a preamp. Yeah, they're old-school but they're still available in multi-thousand quantities overnight.
I understand what you are saying but carrying your logic further suggests that calling white light white or a white sandy tropical beach white leaves the same aftertaste. Rather than redefine what the word white means in the physical sense perhaps a more appropriate solution would be to refer to "white" people as pale or faded or bleached or even better yet since in the human population there is a whole range of skin colors let's just make skin color a non-issue.
Everyone in the history of synthesizers starts with an oscillator so of course you start with a noise source just to be different. 😏 I approve of this course of action.
The fact about zener diodes generating white noise by the electron flow....well it's magic!
Try a reverse biased transistor base emitter junction as well, as that also will break down around 7V, and give similar noise. You can even use TTL gates for this, abusing the input structures and output structures, with only 2 connections, to get a noise source. For higher more impulsive noise use a RF transistor as amplifier, and connect diode direct to base, without breadboard capacitance, which will reduce that initial falling time a lot.
What about the Collector?
Leave it open, floating or short to Base?
@@paulcohen1555 Either way will work, though shorting to base can reduce external pickup of noise, especially if you use a TO can device where you can ground the can and collector.
I just built the white noise generator and it works great. The only changes are that I used 9v supply, a 2.2k resistor instead of the 18k resistor, and an 8.2v zener instead of a 8.6v zener. It took a little bit of trial and error to find out a good replacement resistor value, but it was still pretty easy. Thanks for the circuit idea
You're awesome, this is exactly my cup of tea
The video hasn't started yet but besides this channel being great this is exactly what I've been doing/inventing over the past years. Great fit to my world, exactly what I love and live.
Also awesome to see the noise shape, almost like the breakdown curve of the diode before it resets, awesome.
It's like seeing the breakdown curve real life
Interesting video series idea, looking forward to what’s to come.
Awesome order as well
I've been building oscillators and effects for years and everytime I have a massive struggle hooking it up to something to test the sound😂😂
And then I wanna build a test speaker but before I get started think of another crazy effect and end up with wires everywhere again stretching from breadboards to random amps😂
I have a feeling this is going to be a fantastic series of videos.
I've been working on my analogue synth build since Christmas, and with an arduino based midi to CV converter it's now at a point where I can get something vaguely resembling music out of it.
Really looking forward to seeing what you come up with....
Yes, I've been into it for a year now. My second project was a MIDI to CV converter as well. Just to save you some frustration. Forget to use the inbuilt DAC. Even though my microcontroller was not an Arduino but a PIC, I still can tell you, that you need a more linear DAC.
@@PermireFabrica yes I have 2x MCP4822, that gives 4 voltage control channels that have 12 bit resolution.
I have voltage outputs for note CV, pitch bend, mod wheel and velocity.
@@brianharper9798 okay nice. Quite a powerful module you've built! Mine is only dual channel and only gate and CV output. But one channel can be set up as a 16-step sequencer.
Just started hobby of diy guitar effect pedal builds using veroboard. Really like the simplicity of your videos! Learning alot about opamps!
This is really an amazing idea for a video series! 😍♥
The amount of money you can make selling these has a boutique product is absolutely incredible😮
Awesome and exciting! Can't wait to see how you filter and change the noise source!
This is going to be very useful for me so thanks a lot for sharing this series of videos, you're the best.
Interesting. A very cool project. Thank you. I love audio projects because they are so much fun to listen to.
Many thanks for sharing this.
Calculating is cheaper than soldering. If you want to have 19 different oscillators for DIY sounds take a look to the Synthesizer Workstation: th-cam.com/video/oI_mTEu3WqE/w-d-xo.html
Peter Blasser creator and owner of Ciat Lonbarde uses this technique in his: Tocante Zenert device, Peter Blasser was the Student of Don Buchla, the co inventor of the VCO.
Cool for showing how to create noise with a Zener, once i did a search on that, but nothing found.
I got "alot" of zeners, but not that one.
Nice. I will be following along on a breadboard
Great video series. Really going to enjoy you doing this! Thanks.
waiting for part 11 where we learn to control the sounds with our brain..
This could be a start to a digital noise generator I wanted to build.
-Thanks for posting!
www.tautec-electronics.de/Datenblaetter/Schaltkreise/MM5437N.pdf
This will be a great series. :)
I can relate to you like everyday I come up with new designs and creative ways to build amplifiers. And sometimes I have dreams about how I would like to build and design amplifier. It is a passion for sure. 🤗💯💖 Thank you for sharing yours 😁 I just love creativity and unique designs think outside the box. Once you understand how amplifiers are built and how solid state chip amps and ,op-amps ,transistors and so forth work then you can get creative and build your own schematics for building amplifiers different from what's already out there. I tend to keep my schematics to myself though for personal use only. 🤗 But not to say I won't share them and maybe sell some one day but for now it is just for fun.
very excited to see this.
RIP Ray Wilson (Music From Outer Space)
As far as I know all the kits and PCBs for the Music From Outer Space stuff is only available from Synth Cube now.
I love your circuits, say could do a video using a diode OP-AMP Log Amplifier and how to plot the input and output
Voltages and how the op-amp generates a log output thanks
So what you just showed me is that if I use a zener diode as a voltage reference it is going to be noisy and I might want to include some filtering.
Yes and 78xx 79xxseries linear regulators make noise too alot of ppl have no idea about this either
Zener noise is usually pink noise, isn't it? I remember when I built my HF noise bridge, it needed a filter network to flatten it.
How do you get a negative 12 volts and a positive 12? will you show how to build a duel power supply?
I'm thinking pink noise as a sleep aid, I think this is great and I'm glad its pat 1 of ? ...cheers.
you can have pink noise by adding a filter to this
@@IMSAIGuy yes I think this is something I will play with...cheers.
"Moog" pronounced like "vogue".
You and I should chat since I just started on a similar trajectory, also op-amps, also those light jumper wires (but an analog computer-like learning-PCB, not a synth).
My casual interest in music synths suggests pink noise is preferable? I suppose there is an easy filter you can add to get it pink.....
Unfortunately not as easy as you think. Check out this video: th-cam.com/video/0yB_h_wFkh4/w-d-xo.html
Sweet series. I have a new scope on the way that is somewhat respectable and I have thr lm386 circuit already breadboarded from something I was messing either last week. Is it better if I just build the output on some veroboard?
soldering is more reliable
@@IMSAIGuy that's what I thought on both. Thanks man. A decent multi output psu is next on my list.
Does that lm386 have a 1w output or a 700mw output?
Just wondering which one you are using 😁 of course the 500mw lm386 would be better suitable for a preamplifier stage. Almost all the preamplifier op-amps that I have looked at on a data sheet usually range from 135mw 250mw, 350mw, 500mw and 625mw. But the 700mw and the 1watt lm386 are great for pushing a 2 watt or 1.5 watt 8 ohm speaker. But I have used the 1watt lm386 op-amp as a preamp amplifier but you have to used a trim pot or a resistor on the output of the 1watt lm386 before connecting it to a amplifier stage to get great sound quality control. I have also used the 1watt lm386 op-amp to make a distortion pedal at full tilt or cranked all the way up it'll get really distorted. No need for germanium diodes or LED lights to create distortion😂 You can use a 10K potentiometer along with a 10uf capacitor across pin 1 and 8 of the lm386 to control the gain.
Check out the 'teenage engineering pocket operator modular 400' Not great for learning the design, but very educational with how to interact with each block
I have one of those, the factory one I think
Wouldn't the open loop LM386 work like a comparator and completely saturate into his 12V rail with any voltage over 0V? I can imagine it would look like a square wave at the output? (There is maybe a low pass filter on the output, maybe not a super square wave)
And which 47µF capacitors do you recommend using for audio circuits? There are a lot of rumours on the internet about "audio capacitors" :/
the LM386 is a fixed gain part.
this is not audiophile, any 47uf is fine
@@IMSAIGuy Thank you 🤗
According to LM386 data sheet the capacitor should be 0.05 µF (i.e.50nF)
@@petrituominen8709 yes, I just didn't have one. I usually see 47nf
I love the system building bock ...
And to see what sounds you will make...
May a Theremin...can be inserted...that you make......some kind of VCO....to increase and decrease the frequency...of IMSAI Guy voice...
At first when I looked at LM386....I was expecting some negative feedback...
But it looks like their are some gain set pins...so maybe this circuit has a fixed gain...of 20?...default..
I will go look...at your circuit
Nice white noise source...simple ...
Good job...
is the opamp in open-loop? I dont see any feedback element
I had this same question.
I believe the LM386 has internal feedback as it's meant specifically for audio amplification and thus the huge internal gain of a typical opamp isn't needed here. The datasheet shows an extremely simplified version of the internal schematic which has gain control circuitry. By default the gain is 20, and with a capacitor between pins 1 and 8 (as you can see here, disconnected) and a cap on pin 7 you can change the gain
Lm386 ist not a opamp. It is a audio amplifier for small speaker. For sure it can be used for some other things. On the other Side this ist not true for all speaker amps
The classic tda2030 ist more like a Power opamp😉
the LM386 is a fixed gain device.
Ray Wilson (MFOS) died in 2017 or so.
Dam …. What was the name of that company that sold synthesizer module kits that always ran ads in Popular Electronics???
PAIA perhaps.
I'm having trouble replicating your noise with your circuit, using a 2N3904. It looks like you have about 200mv spikes. I see nothing (just ambient noise), connected to the collector of the transistor. Does it have to be hooked up to a load? Maybe the scope has to be set up a certain way? I would love to see more detail on that.
instead of hooking the 18k to +12v, connect it to the collector
@@IMSAIGuy Thanks, that helps -- I'm seeing the ramp shapes now, but it doesn't look like "hair" like yours -- there's no "floor" to it, it seems to be riding on other random AC noise of 30-70mV, even if I use a 12V battery for the supply. Maybe it's just the cheap zener I'm using, or just way too much ambient room noise.
what is the speed of your oscilloscope?
also what voltage is your zener?
@@IMSAIGuy I was using a 1GHz scope, tried limiting BW and also tried a 100MHz scope. I think I figured out that most of the noise was due to using long leads to the power supply (or battery) picking up RF, looks like mostly around 100MHz (FM stations?). Thanks for the help!
Does it matter what type of NPN you use in the noise generator? Using a 2N2222 I can't get noise out until the Zenner value (+ NPN Voltage Drop) is less than 0.3v lower than VCC (ie 11.7v -12v). An 11.1v Zenner worked but anything less generated no noise at all. Tried at lower voltages (VCC) and got the same result using lower voltage Zenners.
you need avalanche, lower voltage zener diodes zener, higher voltage zeners actually avalanche . the avalanche voltages of NPNs will vary. so best to use the 3904
@@IMSAIGuy Hmmm. Put in a 3904 and no change. With your circuit, except using a 8.2v Zenner - 1N4738A (since I don't have an 8.6v Zenner) - No noise - no output at all. OK, reading further down the comments I see connecting the 18k to the Collector rather than 12v is a thing and now I get 1v PP white noise.
@@davidh1187 it is sometime picky about the layout, you try the circuit in: th-cam.com/video/Ow9zQXPYm3g/w-d-xo.html
it is almost the same but on resistor is tied to the collector instead of +12
Raise your hand if you've experimented with the original 40-pin 0.6"-wide DIP SN76477N?
I made a really shit analogue synth for my electronics gcse and now the whole internet keeps reminding me of it I don’t know how make it end
Interesting.
Synth ❤️ love
oooo do a 4-pole cascaded OTA VCF !! :D
Yes, I will do that
What is the upper frequency limit of the noise in your circuit?
Please show the spectrum.
Depends a lot on the transistor amplifier and the layout, low capacitance diode and RF transistor you get up to GHz noise out with common components. RF noise source design is hard, you have to use really low capacitance and make large air gaps to reduce stray capacitance, and short leads as well to lower inductance.
@@SeanBZA And of the Zener itself? (Assuming low power 0.4W Zener)
@@paulcohen1555 Pretty much just needs low capacitance, so often for the high end it is a point contact diode.
@@SeanBZA Will 1N34 do the job if operated at breakdown voltage (with a proper current limit, of course)?
@@paulcohen1555 You would probably be better off using a RF transistor in reverse bias there, smaller junction and lower inductance will make the wideband noise figure better. Plus breakdown better defined as between 3 and 10V, as opposed to the 1N34 being anything from 20 to 60V.
LM386's are unaccountably popular. They hiss. They oscillate, particularly if you're at all careless with layout. They sound like crap at best.
I actually get better results with the venerable LM380. You can put a couple of big copper foil heatsinks on it to get more power, it's more stable and quieter, you can wire a pair in push-pull for even more power out and drive a fair-sized speaker if it's reasonably efficient. If it's an actual audio circuit rather than a synth, NE5532 pairs well with it as a preamp. Yeah, they're old-school but they're still available in multi-thousand quantities overnight.
Analog computer.
Thx for letting us listen to a zener diode. But to call it white noise in these days has some aftertaste. 😁
I understand what you are saying but carrying your logic further suggests that calling white light white or a white sandy tropical beach white leaves the same aftertaste. Rather than redefine what the word white means in the physical sense perhaps a more appropriate solution would be to refer to "white" people as pale or faded or bleached or even better yet since in the human population there is a whole range of skin colors let's just make skin color a non-issue.
Audio noise spectra are named by analogy to the visible light spectrum. The Wikipedia article on "Colors of Noise" explains it well.
Oh man are you off into a deep hole now. Ran out of things to spend money on? HA.
oh you are so right, this a very deep deep hole