[sdiy] A query about the TR-909 noise generator

Brian Willoughby brianw at audiobanshee.com
Sat Dec 21 12:37:35 CET 2019


Look at the polynomials for maximal LFSRs in the wikipedia article.

https://en.wikipedia.org/wiki/Linear-feedback_shift_register#Some_polynomials_for_maximal_LFSRs

It only lists up to 24 bits, but has links to polynomials for larger shift-register lengths.


The wikipedia page also links to FPGA implementations - another option with less parts (potentially one little chip).


On Dec 21, 2019, at 2:37 AM, Tom Wiltshire <tom at electricdruid.net> wrote:
> The TR-909 uses a LFSR (shift register) based noise generator. It used two 4006 18-stage chips for the register, so there are potentially 36 stages available. However, the taps are taken at positions 13 and 31 (bits 12 and 30), so the effective length is shorter. The service manual claims 32 stages, which seems reasonable.
> 
> I can’t find any references to the two taps they’ve used though - could 13 and 31 be a non-maximal sequence for a 32-stage LFSR? Can anyone confirm this?
> 
> My interest in this is that the 4006 is now getting hard to find, so I thought it might be fun to do firmware for a TR-909 noise generator, since it could be identical to the original, but with less parts (one little chip instead of three bigger ones).





More information about the Synth-diy mailing list