I have just shared a new project: “Analogue Quad Oscillator Drone Synth”
I have a peer-reviewed study on my desk and it confidently tells me that all the capacitors on planet Earth want to be synthesisers when they grow up. To make the world a better place, let’s realise the dream of four fortunate electrolytic c…
Read more
Hi Pix
There are some largish resistors on your breadboard near the LEDs which appear to have the coloured rings Brown, Black, Yellow. The last time I looked this was 100kΩ. where do these fit in or the colours coming our wrong in the pics and the rings are actually Brown, Black, Orange (10kΩ).
Cheers Bob
Hey Pix,
Awesome project and what a great write-up as well! I love music / audio-related electronics but its something I’ve never ventured into, glad I can live vicariously through things like this. 
-Jaryd
Well - now I know a lot more about synthesisers! thanks for the deep dive -
It’s a Super cute concept that all capacitors want to grow up to be synthesisers when they grow up. I wonder now what other electronic components dream of becoming participating in.
is there a resister in that drawer that wants to be part of a cosplay? So many questions, including what does monkey paw soup taste like and where can I try it ?
Hi Bob 
Yep that’s a 100k resistor.
Because there is 12-18v running around the oscillator getting loud volumes from the toy-synth isn’t really a problem. This is especially true because I used an amplifier.
The maker-verse amp has a screw on top that set’s the volume. That screw is sensitive to variation at softer volumes but not sensitive at higher volumes. That logarithmic response is typical of audio.
100k ohms is just a nice big value to put in between the oscillator and the amplifier.
It meant I could get nice soft volumes out of my amplifier and fine control. The cost is I probably lost a bit of dynamic range, and spectrum.
Nothing special about 100k, someone might find and recommend superior values if they re-create it.
Pix 
Hi Pix
Thought it might be.
The bit I was getting at is the lack of accuracy.
100k does not appear in your parts list or your schematic. 10k does and that is a factor of 10. If this is the sort of tolerance you work with it is a wonder you get anything done.
I think the lesson here is if you can’t post an accurate schematic or whatever don’t post anything. It only can cause confusion. In this case it doesn’t matter much but it can have a very adverse response. From me anyway.
It is the human ear that has the logarithmic response. That pot sounds as if it has a linear curve. For Audio it should be a log curve or these days called “Audio Log”. Don’t quite know what the difference is exactly except that I think some of the cheaper units actually have 2 or 3 linear sections sort of making something resembling a log taper where a better quality device will be a “proper” log taper.
As a bit of trivia, back in the day there used to be available a “reverse log” pot. Required in things like the frequency control of a Wein Bridge audio oscillator. The use of this prevented the dial scale cramping up the settings at one end.
Cheers Bob
I found this graph to show those curves. It also shows off how the linear sections are used to estimate the curve.
Yeaaah Jonny! The demo video is a 
vibe
Amazing what you can get done with a handful of jellybean components 
Thanks for this 5-star writeup. I’m really impressed by the quality and care that went into this project.
Love ya work 
Harsh. Good attention to detail, but a bit heavy-handed.
If you’re not prepared to make a mistake you’ll never do anything new or fun or original 
I think the stakes are pretty low with a fun, creative, educational maker project.
How many 100+ page datasheets from giant manufacturers have an errata section at the front?
This is an analogue project, the value will change the characteristic and I suspect Pix selected 100k to taste .
It looks like the exact value is not critical as shown in the BOM
Hi Michael
From this end I don’t think so.
There was a case not so long ago where a schematic and photos were posted. After spending some considerable time researching and trying to guess what this person (I forget who now) was trying to do it emerged that the published material was not what actually existed and was incorrect. In fact the publisher stated this and admitted publishing the info KNOWING it was incorrect.
No don’t you tell me THAT is good practise. Hobbyist or not. I don’t see anything wrong in learning to be accurate. Will get better results in the long run. And the only way you are going to learn is to be told.
Where does the BOM say that. I think that the 2 Watts for this application might be the heavy handed bit.
Here is the relevant section of the BOM.
The number provided (CE09886) is for a 250mW resistor anyway.
BUT, this is quite an ingenious method of getting electronic “brassy” sounds in a simple manner.
A couple of comments I would like to make however which could be part of the learning curve.
If the electrolytic cap values need to be close to their nominal value the higher voltage (63V) ones will not form to their full value when rated so far above their actual working voltage.
You are driving the transistors past the reverse breakdown voltage repeatedly. I don’t know how long they would stand that (I have never had the reason to try). You MIGHT get a similar result with a Zener diode and do it a lot safer and longer.
All up a neat little gizmo.
Cheers Bob
A list of reasonable values for a single oscillator is a little higher up in the article.
I agree, 0.25 watts is absolutely fine in this scenario, I should have removed that requirement but I missed it, even on when checking the final draft. 
. Small mistake; no drama.
I agree precision can be really important. 
Someone who is debugging might find it helpful to have a tested schematic to fall back on as a sanity check. 
I wanted to encourage some experimentation but one day I’ll return here with a schematic that precisely describes out a single quality oscillator, values and everything. 
It will be a nice little appendix to the write up that balances encouraging exploration with easy implementation.
Hi Pix
Firstly well done. I for one would have never have thought to generate a triangleish (new word) waveform using this method.You can really get some weird sounds with this little beast.
But
This is exactly what I was getting at. “a little higher up in the article” is NOT in the BOM. When provisioning to build this device the constructor is going to use the BOM for is shopping list. This is the sort of thing which can cause the confusion I was getting at.
If I may make a suggestion I think things would be much clearer if you marked this sort of variable component (asterisk or something) then noted at the bottom any variation that could apply. That way any such variation is associated directly with the BOM.
Also annotate your schematic like R1, R2, C1 etc and reference these in the BOM so there is no confusion regarding which component is affected. Annotating also simplifies things when it comes to describing circuit operation. Much easier to say R? than “the resistor connected to the third transistor from the top base” as an example.
All in in a process of learning
Cheers Bob
Thanks Robert.
Good thoughts for next time.
And you can get our latest projects and tips straight away by following us on:
