Forum upgrade to SMF 2.1.2, still working on re-theming.

Main Menu
Welcome to Solid State Guitar Amp Forum | DIY Guitar Amplifiers. Please login or sign up.

May 19, 2022, 05:36:19 AM

Login with username, password and session length


Clip but limit highs

Started by skey, April 12, 2006, 01:52:43 AM

Previous topic - Next topic


Take a read at this:

Limiting can be used to prevent clipping, but this doesn't get the maximum drive out of the amp. Lows usually cause the clipping.  During bass clipping the high frequency of the input signal has not clipped and thus is still allow to gain DB's.   The lows compress far before the highs ever do.

I'm thinking the key to squeezing out the maximum from a lower wattage SS amp is to allow the lows to clip, but when they do start to compress the highs.  In other words instead of limiting the whole signal just limit the highs.  The amp will still clip and add "some" high frequency content, but as much as people believe.

This just might be the key to squeezing out the most from a small SS amp.


Good read...chip points your way! :)


Sounds like an interesting theory. Have you tested it? As far as i know the bass frequencies tend to dominate, and if they clip they will drown the highs. You know that farting sound coming out from the speakers, thats a clipping bass frequency. It's much more difficult to detect a clipping high frequency. This is why most distortion pedals cut down bass frequencies before clipping stages.

I did a quick test on your theory. Unfortunately, i was quite sure that i could predict the result. I took a .wav file and split it in low and high freq content. I clipped both of them in two ways: 1) I clipped more the low frequency content and less the high freq (like in your theory) and 2) I clipped more the high freq and less the low freq. Finally, i summed the signals back together. I did the test with both clean and overdriven guitar sample and in both cases the one with more clipped bass sounded much more horrible than the one where the amplification of low frequencies was limited. The overdriven sample especially sounded really horrible when it clipped on low freqs.

Anyway, i also ran some ltspice tests which showed me that your theory might actually have some point: A constant bass frequency seemed to have less harmonics after clipping when the summed highs were attenuated. You might be able to build something out of this so don't give up. I'm also ready to change my opinion when i'm proven to be wrong.

Rane's document had some good points. It is right about the detectors operating too slowly to detect a clipping bass frequency. Also, you are right about the fact that limiter's should operate on separate frequency bands. Best audio limiters do.


The more i have played with skey's idea the more sense it starts to make: The frequency band split frequency has to be chosen carefully, (so that the low freq will indeed mean low), and possibly the amount of overdrive on the low frequency side has to be limited a bit too. High frequencies do indeed seem to need more limiting than the low frequencies to sound clear but the most important thing is to attenuate both frequency bands before summing them up - otherwise the clipped low frequencies will dominate too much. If the signals are attenuated so that they will not overdrive the next stage the clipped bass actually seems to be harder to detect: The result is somewhat "unclear" tone which  has a "punchier bass" in comparison to the way of limiting the low frequency content, which seems to result into a "clear" low end but overdriven sound in general.  Both limiting tactics could benefit from a post low-pass filter that would attenuate the harsh high frequency noise that clipping the signal creates. Also, the idea seemed to work better on a clean signal, which naturally has a punchier low freq content than an overdriven signal. The results with highly overdriven palm-muted grind riffs weren't very impressive. In the later case limiting the signal equally on every frequency seemed to provide a better result.

I hope you will look further into this idea skey. At this point, the biggest downside i see in it is the added complexity of circuitry.


Thanks Teemuk - youve taken it 400x more than I can at my pace.
I'm not up on the spice tools, and manipulating wave files. 
I'm just very happy the whole idea wasn't shot down because I missed something.  :)
I was going to try some wire experiments.  Mainly at first doing the grain-of-wheat lamp trick on the output  to feedback to a ldr on treble control in the preamp.  This probably won't be fast enough, so maybe switch  to some LEDs.  Trying to keep it simple if possible.  Maybe not ideal but at least a little benefit.


If we're talking LM3886 chipamps, clipping those will sound horrible period. I would imagine low-end clipping would be worst. Those chips have a lot of headroom thank god. Limiting(but set only to limit at the headroom ceiling) is a good idea for chipamps.



Quote from: RDV on April 12, 2006, 12:07:55 PM
If we're talking LM3886 chipamps, clipping those will sound horrible period. ...

This is the generalization I'm trying to avoid here.  The article points out one reason why an amp may sound bad at clipping, and it's not due to the clipping.  It's due to the high frequency component of the input signal gaining unproportionate DBs.

Clipping harmonics are very small compared to the high frequency components of the input signal.


The reason a chipamp sounds bad @ clipping is because it's an opamp and they do not clip gracefully.



Having once gone to the trouble of splitting off highs in a (presumably) well behaved crossover, why wouldn't you keep the highs separate from there all the way to the speaker? Bi-amping has a long and well documented history. The limiting on the highs could be done by a limiter or just by using a correspondingly smaller amplifier for the mid/tweeter drivers.


Here's some good continuation for the Rane's article about tweeter damage due clipping presented by skey:


Wow - that link is a good find.  That whole site is great. Thanks!