Looking for simple peak limiting ideas

[Bill Mountain]

What would you use for peak limiting huge spikes (like those than would come from a violent bassist )

I don't like the way compressors dip and swell.  I've thought about making a super-mega-ultra-light-overdrive.  Maybe even a simple boost with pleasant clipping.

I'm talking 1-5 volt signals on a passive bass.

I've made heavy use of dirt pedals but I'm moving towards a cleaner sound as of late.

I could try to fix my technique but I'm a wild player at heart.

I have tried a few sansamp and mxr bass preamps that I like.  I just don't like the way they clip.  If I could control large transients before I get to the preamp I think I'll have the best of both worlds.

So I guess I'm looking for ideas for light overdrive or compression for high output passive instruments.

My main circuits ideas are a tube sound fuzz or a bluesbreaker modded for low gain and bass guitar.

What would you all do?

[Mark Hammer]

Get thyself to Small Bear and score yourself an SSM2166.  Then look at the datasheet/appnotes for the 2166 and figure out to nail what you're looking for.

The 2166 not only lets you pick the threshold at which gain reduction kicks in, it also provides for some pleasingly selective downward expansion to clean up noise during quiet parts.  Just a great great chip, that I wish was more commonly available and cheap so that more folks could benefit from it.

I also can't emphasize enough that this is a full-menu dynamics-conditioning chip and not just a compressor.  When it comes to bass, one wants a limiter, rather than a compressor.  You want to keep the dynamics as much as possible, but forego the disruptive transient peaks.

[amptramp]

There are all kinds of diode noise limiters that rely on having a large time constant on one side of the diode and a reduced amplitude signal from a voltage divider on the other end.  Once the instantaneous level goes above the average/divider ratio, the diode clips.  These have been used in communications receivers back to the 1940's.

There is one idea I ran into that is extremely simple and elegant: take the signal through an op amp stage that uses external compensation like an LM101 or LM108.  Use an outrageously large compensation capacitor - enough to pass the high end (which is not that high on a bass guitar) but not much higher.  When you hammer a string, the compensation capacitor goes into slew-rate limiting mode.  Unlike a diode clipper, it is bidirectional.  The cool part?  Switch off the large capacitor leaving a small one in place and you have a usable linear buffer.

[Bill Mountain]

Get thyself to Small Bear and score yourself an SSM2166.  Then look at the datasheet/appnotes for the 2166 and figure out to nail what you're looking for.

The 2166 not only lets you pick the threshold at which gain reduction kicks in, it also provides for some pleasingly selective downward expansion to clean up noise during quiet parts.  Just a great great chip, that I wish was more commonly available and cheap so that more folks could benefit from it.

I also can't emphasize enough that this is a full-menu dynamics-conditioning chip and not just a compressor.  When it comes to bass, one wants a limiter, rather than a compressor.  You want to keep the dynamics as much as possible, but forego the disruptive transient peaks.

I wonder if that's the same limiter that's old Sunn Coliseum 300's???  I was my favorite built in limiter.  Maybe I could look into the SWR or Hartke limiters.

[Mark Hammer]

I doubt it's the same.  History does not have those two things overlapping in time.  Although the chip is not cheap, in comparison to 25-cent dual op-amps, it is a ridiculously cheap solution to dynamics-control challenges.  http://www.analog.com/static/imported-files/data_sheets/SSM2166.pdf

[teemuk]

There are all kinds of diode noise limiters that rely on having a large time constant on one side of the diode and a reduced amplitude signal from a voltage divider on the other end.  Once the instantaneous level goes above the average/divider ratio, the diode clips.  These have been used in communications receivers back to the 1940's.-rate limiting mode. 

Clever scheme. I understand how that works by principle but I don't see how it could work properly as simply as what you describe.

Can you point out a design (or pehaps a few) that uses this scheme for reference?

[Bill Mountain]

I doubt it's the same.  History does not have those two things overlapping in time.  Although the chip is not cheap, in comparison to 25-cent dual op-amps, it is a ridiculously cheap solution to dynamics-control challenges.  http://www.analog.com/static/imported-files/data_sheets/SSM2166.pdf

Thanks for the link.  They really explain how it works.  I wish more datasheets didn't make so many assumptions for you.  We're not all engineers!

[R.G.]

Back to back IR or old, standard red LEDs clip about 1.2V. Two pairs of silicon diodes will clip at about 1.4V. A bridge rectifier IC can be set up as two sets of silicon diode clippers in series.