MM5437 white/pink noise generator

[armdnrdy]

I came across a MM5437N in a mixed bag of ICs that I purchased.
It's a digital noise generator on a chip that supposed to be good for white or pink noise generation.

I included the PDF below.
http://pdf1.alldatasheet.com/datasheet-pdf/view/103155/NSC/MM5437.html

Save for the limited example circuit in the PDF, I can't find any other application circuits. I know that synths used this IC but no examples.

Is anyone familiar with this IC or know where it was implemented in a circuit?

I had the thought to mix a small amount of white or pink noise with guitar in a phaser or flanger circuit to see how it sounds.

[oskar]

The only example of it in use I've seen was a schematic by Craig Anderton in an old (~1980?) Keyboard Magazine.

[armdnrdy]

This is the only thing I could find. This was posted on a site from a drawing in National 1984 Logic Data book vol. 1. 
Can anyone decipher these hieroglyphics into a schematic?               



                          Top View


                        -----------------
                         | o     U       |
         White noise out |1             8| V+
                         |               |
                 N/C     |2             7| N/C
                         |               |
                 V+      |3             6|  V+
                         |               |
                 GND     |4             5|  N/C
                         |               |
                         -----------------


The output impedance should be 3kohms or greater. Use an op-amp
buffer if impedances are less. The white-to-pink noise converter is


IN >------ 3K----------------------------------------10uf(opt)------>Out
                 |          |         |          |               |
               3Kohm       1Kohm    300 ohm    .082uf           100Kohm(opt)
                 |          |         |          |               |
                 1.0uf    .27uf      .15uf       |               |
                 |          |         |          |               |
GND >--------------------------------------------------------------->Gnd


It's output load should be 10K or greater if accuracy is to be maintained.
Note that without the 10 uf capacitor, the output will be DC-coupled
with DC being one-half (average) the supply voltage.

[armdnrdy]

Okay,

I figured out the hieroglyphics.

The above drawing is the pink noise generator straight out of the data sheet.

I wonder if to use this IC as a white noise generator you just come straight out of the output pin.

Now I have to breadboard this and figure out what I need to do to implement this in a circuit. Buffer? Attenuator?

Any suggestions?

[mph]

I came across a MM5437N in a mixed bag of ICs that I purchased.
It's a digital noise generator on a chip that supposed to be good for white or pink noise generation.

I included the PDF below.
http://pdf1.alldatasheet.com/datasheet-pdf/view/103155/NSC/MM5437.html

Save for the limited example circuit in the PDF, I can't find any other application circuits. I know that synths used this IC but no examples.

Is anyone familiar with this IC or know where it was implemented in a circuit?

I had the thought to mix a small amount of white or pink noise with guitar in a phaser or flanger circuit to see how it sounds.

Hi

I know it's used in the Oberheim OBX, Moog Memorymoog, and maybe also in the Buchla 266e module... it's not so interesting compared to a common transistor noise circuit as its reputation is to be not really reliable.

The best advice would be to keep it if you own some synth that uses it, or sell it as spare part to someone who need it... 

Just my 2cents.

[armdnrdy]

There's another IC the MM5837 though not a pin for pin replacement seems to be the "follow up" to the MM5437.

Both ICs are supposed to provide frequencies based within the range of human hearing as well as provide uniform noise and amplitude.

I'll have to check out some transistor circuits and compare results.

[Mark Hammer]

I have a 5837, bought from Radio Shack years ago.  The word on the 5837 (and I assume the 5437 as well) is that it is not the very best you can do for a noise source, when using it as a control voltage source, even though as a noise source for producing audible noise, to be filtered or whatnot, it is about as simple a solution as you can find.  The reasoning is that the 5837 is essentially a counter, and that it is not quite as random a voltage source as one might like for something like a S&H.

[armdnrdy]

Thanks for the info Mark.

I have three options:

Let this IC sit and do nothing with it for the rest of my life.

Smash it with a hammer. 

Try to make something useful out of it.

I'm choosing option three! 

As I stated earlier, I came across this IC kind of unintentionally. My thought was to mix some white/pink noise in with a flanger or phaser dialed in to the heavy jet plane mode.

I think augmenting the guitar sweep with a "wash" of harmonics might make for an interesting effect.
After all look what Steve Stevens did with a toy ray gun!

[R.G.]

These chips predate microprocessors. They are a hard-logic implementation of a linear feedback shift register noise generator. Their limitations and abilities are all from that.

They have a limited length, determined by the number of shift register stages inside. The limited length makes for their chief limitation, the short time before the sequence repeats. This was about every two seconds, and they do have an every two second "hey, that's repeating" note.

This chip is one of the instigators of my working on a  noise generator in an 8-pin PIC in another thread. I did some programming to generate noise that repeats every several millenia.  

[armdnrdy]

I did notice "typical cycle time is one minute" and "The first, OUT 1 is sequence-limited to reduce "thumps." in the data sheet.

Maybe I should choose option 1 or option 2 from above!

[PRR]

What's to know?

Apply 0V to Vss (pin4), 5V or 10V to Vdd (3). Take white noise from Out1 (1).

You do not need to deal with the other pins. (Look like BS to me. The thing of making quasi-random numbers for a CPU, any good CPU can do that internally.)

If you like it Pink, Fig 3 is indeed a passable Pinking filter. Part values are very non-critical; 330 instead of 300 is almost undetectable.

Either way, you want a blocking capacitor on the output 'cuz there's +4V of DC on it.

Minimum load for White hiss is under 1K. Min load for their pink filter is probably 10K for low bass error, 1K if you just want "hiss" and are not too fussy about frequency distribution flatness.

1 minute repeat is not an awful thing. I got tons of use out of the older chip with a 7-second(??) repeat. In instrumentation, there's a bobble; but measuring noise is all aboout averaging. Gating for a bit of "chuff" on note-attack, you don't hear the repetition. However not a good choice for a sleep-aid device (you'll count thumps all night long).

[Mark Hammer]

All all of the above is precisely why I described it as "not the best" one could do for a noise source.  "Not the best" does NOT mean "the worst".

Gotta wonder exactly how much randomness is required for people to perceive randomness, when it comes to sound effects.  For example, if one is using a noise source to dictate "random" shifts in a filter's center-frequency movement, as in a Maestro S&H unit,  how non-repeating does it need to be to be perceived as non-repeating?  If one was looking for a random-number generator for a video gambling terminal, or some other computed outcome, that would be one thing, but when all something has to do is be unpredictably different than whatever happened during the preceding few seconds, that is quite another.

[artifus]

Gotta wonder exactly how much randomness is required for people to perceive randomness, when it comes to sound effects....

maybe as with most other senses that the brain doesn't find it so boring as to not perceive it? like a persistent smell that is no longer noticed. the random variation required to retain interest no matter how subtle. should that be the desired outcome. just thinking out loud here...

[R.G.]

Gotta wonder exactly how much randomness is required for people to perceive randomness, when it comes to sound effects.  For example, if one is using a noise source to dictate "random" shifts in a filter's center-frequency movement, as in a Maestro S&H unit,  how non-repeating does it need to be to be perceived as non-repeating? 

You're perhaps the best qualified of all of us to answer that.    

I would guess that when you're listening to the output of the noise generator itself is the worst condition for perceiving repeats. We know that repeats every few seconds are discerned as thumps or beats. The loudness envelope is memorable over longer time periods than the hash itself. Just as raw speculation, something like a minute ought to be good for forgetting enough - again, just guessing.

For shifts along the lines of an LFO, I think the criteria changes. An LFO with random-over-regular is fairly pleasing. Think of the sound of an ocean surf - a regular-ish repeat, itself randomized a bit from beat to beat, and with a much longer-term fade in/out or cyclic variation.  Wind does something like this too.

Leaving aside the rhythmic-but-variable thing like surf and wind, you could get really random-within-bounds stuff by feeding the binary-random stuff into a filter of some kind. A simple R-C works, but a lowpass filter or integrator makes this work really well, converting the raw 1/0 binary into analog variations in control voltage. This of course makes the thump aspect worse, but any modern implementation of a noise generator will be on a PIC or similar, and will easily have repeats of centuries to millenia, because there is no scarcity of bits in registers inside for shifting, as there was in the old hard-logic stuff.

So -Mark, you wanna try out a random generator?  

[Mark Hammer]

I work for government.  I LIVE in a random generator.

I have to see if I can find this old article again, from JAES, by Bernie Hutchins on "fat sound".  It was from around 1980-1982 or so.  Bernie was attempting to identify how many independent sources of modulation it would take to make a synthesized sound more interesting to the human ear.  Keep in mind that this was occurring at a historical point where we were making the transition from "home organs" to polyphonic synths (i.e., pre-MIDI, but not by much).  The big complaint at the time was that keyboards that played more than one note at a time tended to sound pretty lifeless.  Bernie went off in search of some guidelines for incorporating enough modulation to avoid obvious periodicity, without verging on the annoyingly chaotic.  And IIRC, more than 4 sources tended to mush-ify things.  But again, I'd need to find the article.

P.S.:  A-Ha!  A quick search reveals the reference. https://secure.aes.org/forum/pubs/journal/?elib=3873  This is why you want me on your Trivial Pursuit team.

[R.G.]

Hmmm. I can't read the article, because I'm too cheap to buy it or join AES.  But the description gives me enough to extrapolate.

This ticked one thing I have been intending to do, based on long ago info on sitars and such. I dimly remember you being involved in that, as well.

Part of the sitar sound is the selective nulling of specific and changing harmonics of the string by the string physically hitting the very carefully shaped bridge. This amounted, according to the article I read on sitar bridges to a randomly moving notch in the frequency response.

I speculate that a two-stage phaser is a notch generator that moves in response to an LFO, and that one might generate a number of controlled notches by breaking a four to eight stage phaser into two-stage notch generators and feeding a different LFO, perhaps unrelated random LFOs or a mix of random and cyclic LFOs to the phase stages.

Of course, for more "animation" one could modulate amplitude, delay, filter response, etc., and I suspect that this is a part of Hutchin's paper.

[Mark Hammer]

I'm too cheap to buy it too, but now that you have me thinking about it, I think I remember what binder it's in at home.  I'll try and take a look for it later tonight.

For now, I have m yhands full with "random" appearances by a squirrel that has entered our home, can't seem to find its way out (despite finding plenty of places to hide), and occasionally makes appearances atop the living-room curtain rods, staring at me like Chris Griffin's "evil monkey".

[armdnrdy]

Mark,

As a former squirrel wrangler and one of the top men in my field, let me give you some advise.


First.......dress like a nut. There are two methods in which this can be achieved:


Method number one:



And method number two:



Next, when the little bugger comes to crack you open......grab him and hold on to the frisky little devil with all of your might!
If you are in need of a squirrel relocation service, I can offer reputable referrals.

[armdnrdy]

Seriously though,

@ R.G.

I like your ideas for a sitar effect. Could your Pseudo Random LFO be incorporated into a circuit such as this?

My observation about sitar sounds is one of a hollow resonance, not unlike certain settings of a flanger. Another noticeable sitar quality is different harmony related notes ramping up at different intervals when the player hits those particular strings.

[Mark Hammer]

Hmmm. I can't read the article, because I'm too cheap to buy it or join AES.  But the description gives me enough to extrapolate.

This ticked one thing I have been intending to do, based on long ago info on sitars and such. I dimly remember you being involved in that, as well.

Part of the sitar sound is the selective nulling of specific and changing harmonics of the string by the string physically hitting the very carefully shaped bridge. This amounted, according to the article I read on sitar bridges to a randomly moving notch in the frequency response.

I speculate that a two-stage phaser is a notch generator that moves in response to an LFO, and that one might generate a number of controlled notches by breaking a four to eight stage phaser into two-stage notch generators and feeding a different LFO, perhaps unrelated random LFOs or a mix of random and cyclic LFOs to the phase stages.

Of course, for more "animation" one could modulate amplitude, delay, filter response, etc., and I suspect that this is a part of Hutchin's paper.

Just so that folks are clear about this, before attempting anything, these ought to be a collection of parallel 2-stage paths.  Running 8 stages in series, with each pair of successive stages controlled by a different LFO just gets you the same 4 notches moving in ganged fashion, in unpredictable patterns.  Remember that it is the sum of all phase-shift that counts to produce the notches.  What you want is 4 different notches produced, and moving around independently...at least if I've understood RG correctly.