[sdiy] Digital Noise, FYI

[Tom Wiltshire]

On 16 Feb 2011, at 14:58, Magnus Danielson wrote:

> On 02/16/2011 02:38 PM, Harry Bissell wrote:
>> hmmm...
>> 
>> When I got my ProphetV, it used the dreaded (and feared) MM5837 noise source.  I tried to make an exact
>> replacement with analog techniques, but could not get the 0>15V "all or nothing" voltage swing. I suppose
>> with my superior skills these days, a comparator would have done the trick...
>> 
>> A CD4006 based solution was exactly what the doctor ordered, it did not mess up the factory presets in the least, but
>> removing the MM5837 did eliminate the (your choice)<ticking time-bomb>  <roaring freight train>  <chinese water torture>
>> from the sound...
> 
> Oh, so you can't get those autentic sounds anymore? >:-)
> 
>> The digital techniques have their place. I don't really care about 'random'... a repeat every few hours or so doesn't
>> get too much of my attention :^)
> 
> The main problem they had was that they where too short. Bad implementation, but not necesserilly a bad solution as such.

They were too short, but also too slow. The MN5837 is a 17-bit LFSR, and runs at 24-56KHz according to the datasheet I found. That's 131,072 samples which will be over in 2.34 seconds at 56KHz. Richie B. suggested recently that 80KHz is a reasonable minimum clock rate for noise which is flat (white) in the audio band. That fits with my own experiments.

If you use a 32-bit LFSR, you get 2^32 samples = 4.3 billion, which takes 11.9 hours even at 100KHz. By the time you're up to Scott G's 64-bit LFSR (can you tell he uses FPGAs?!) you're getting into tens of millions of years.

So if you make it about twice as long and twice as fast as the MN5837, you'd actually have a very good source.

T.