New person needs help with noise toaster

Allison · Joined: Apr 23, 2013 Posts: 6 Location: Washington, Dc

Hello,
I'm just getting into DIY synth... I have a technical background, but most of the work I do these days is on a computer, and I haven't built anything since grad school. Here's my problem.

My end goal is to build a drone synthesizer, with a few VCOs, filters, and a simple mixer. I've been working with the noise toaster design for its relative simplicity. Thus far, I've built the expo converter and the ramp core of the VCO. I've left out the inputs from the LFO and ARG, as well as the sync (C6 and R24). For ease of testing, I've used two different lm324s just so I can see what I'm doing. The only substitution I've made is an MPF102 instead of the 2N5457. I also used a 1N4148 diode.

My problem is that the frequency of the output sawtooth is way too high, and I only get a change in frequency over about 1/3 of the pot. The frequency ranges between 17-44kHz, which obviously isn't terribly useful. If I replace C2 with a .01uF cap, it shifts to 2kHz-16kHz. I can increase my range a bit more by increasing R2, but I feel like I'm missing a fundamental problem here. The other issue is that the peak-to-peak voltage for the output varies with the frequency, which is also problematic.

Does anyone have any troubleshooting hints? I figured this might be an obvious problem for someone with more experience.

Thanks!!

Allison · Joined: Apr 23, 2013 Posts: 6 Location: Washington, Dc

I should note that I will be happy to post any schematics or measurements... If anyone would be kind enough to help....

elmegil · Posted: Thu May 02, 2013 9:25 pm Post subject:

I haven't built a Noise Toaster, but I've built a Soundlab Mini Synth and troubleshot another. They not identical VCOs but somewhat similar, particularly the FET for the frequency control.

Have you looked at the waveshape? That might give you some clues.

Have you built this on a breadboard? If not I'd recommend re-checking for shorts/solder bridges.

Are your transistors matched?

Triple check R3 to be sure it's what's supposed to be there. I had frequency issues with the SLMS when I had incorrect resistors in the same sort of place (R56), and initially I tried to solve that with a cap the same way you have.

diablojoy · Posted: Thu May 02, 2013 9:46 pm Post subject:

hi and welcome
try removing the mpf102 from the circuit and rechecking the oscillator cores frequency with out it.
while the MPF102 shares the same pin out as 2n5457 there are slight differences also because you havent completed the rest of the sync circuitry it may be effecting things

Allison · Joined: Apr 23, 2013 Posts: 6 Location: Washington, Dc

Hi - and thank you!

How vital is transistor matching for this? Mine are not. I thought it was just important from a temperature stability perspective... But I don't know what I'm talking about.

I think the problem is with my exponential converter. I've tried putting the portion of the circuit in a simulator, just so I can figure out how it works, and it doesn't seem to be behaving like I expect - I'll post a link later. I really appreciate the advice!

elmegil · Posted: Fri May 03, 2013 8:38 pm Post subject:

The matching is involved in how the expo converter works. I have not yet grokked expo converters in fullness, so I can't explain, but you might give it a shot.

Ian Fritz has a great article about a pretty simple method that you can throw up on a breadboard to do matching:

http://home.comcast.net/~ijfritz/projects/transmat001.pdf

bubzy · Posted: Fri May 03, 2013 10:39 pm Post subject:

I don't recommend throwing up on breadboard Very Happy

elmegil · Posted: Sat May 04, 2013 6:10 am Post subject:

Zodiak · Joined: May 20, 2007 Posts: 226 Location: Kent UK

Hello, looking at the schematic I doubt that the transistor matching will cause your problem, indeed unless you are going to use it to play melodies I don't think you need to bother. Smile

The lack of sync circuit link shouldn't make a difference either, most of the time you will probably have it switched out anyway.

The missing feed from the ARG and LFO shouldn't make much odds either.

(1) Check that U1 pins 3 12 and 10 are all at half the battery voltage.

(2) The try measuring the voltage between UC1 pin 8 and pin 10, assuming a 9V battery it should vary from about 90mV to -150mV. If not then check out everything to the left of U1 on the schematic.

I haven't really worked out how the exp convertor works, so next I would remove U2 and U3 to see if that helps. After that its down to checking the components one by one Crying or Very sad

Allison · Joined: Apr 23, 2013 Posts: 6 Location: Washington, Dc

Thank you!! I have a lot of things to work with now... as soon as my 3 kids under 5 leave me alone long enough.

Working this out on the simulator, I think I do need the feed from the ARG (or, rather, a resistor tied to the virtual ground as input into the summer...)

Have you guys ever used these online simulators? They are seriously awesome. Here's a link to the expo converter... that is working as predicted now.

https://www.circuitlab.com/circuit/z6u3d9/noisetoaster-expo-converter/

elmegil · Posted: Sun May 05, 2013 7:16 pm Post subject:

I have tried circuit lab, but I'm not terribly clear on how the simulations are intended to work.

I use this one a lot: http://www.falstad.com/circuit/

It simulates pretty much real time, but doesn't let you save your work, so it's mixed.

Zodiak · Joined: May 20, 2007 Posts: 226 Location: Kent UK

jukingeo · Posted: Sat Sep 07, 2013 5:56 am Post subject:

autowitch · Joined: Nov 19, 2011 Posts: 10 Location: Denver, CO