[sdiy] Homemade synth clone

Ben Bradley ben.pi.bradley at gmail.com
Sat Feb 16 04:28:56 CET 2019

"And anything is better than the MN5837."

Yes, but like many older instruments, sometimes one really wants is
"that particular cheesy sound." A modern microcontroller can recreate
the original bitstream and/or make a much longer pseudorandom pattern
before it repeats (yes, a two-second loop of noise is audible as a
two-second loop of noise).

On the other hand, many modern processors and controllers include a
"true" random noise generator. I programmed a STM32F4 Discovery board
to read the TRNG, and it would spit out close to a million 32-bit
random numbers per second, lots more than enough for audio use. ST now
has newer and even cheaper parts with TRNGs, and it's easy enough to
add a pink noise filter and similar functions in software.

I'm in favor of adding switches for "original, grungy and cruddy" and
"newfangled 'clean' noise" modes.

On Fri, Feb 15, 2019 at 6:36 PM KA4HJH <ka4hjh at gmail.com> wrote:
> > On Feb 15, 2019, at 12:14 PM, Tom Wiltshire <tom at electricdruid.net> wrote:
> >
> > Pete is dead right. NOISE 1B runs on 5V, so produces a analog noise level of 2-3V or so. MN5837 ran on 0V and -15V, so produced roughly three times the output. But fixing this usually involves just a simple gain resistor change.
> And anything is better than the MN5837. Someday I'll get around to cleaning up my MG-1 and then...
> Terry Bowman, KA4HJH
> "The Mac Doctor"
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at synth-diy.org
> http://synth-diy.org/mailman/listinfo/synth-diy

More information about the Synth-diy mailing list