Intermodulation distortion

[xtian]

I love this concept. When two frequencies are introduced into a non-linear amplifier, the distorted output will include frequencies that are highter AND LOWER than the input frequencies. Even Merlin uses an exclamation point in his description (p.17, Designing Valve Preamps...).

This is pretty easy to hear for yourself, especially under high-gain distortion. Just strum a pair of notes up around the 12th fret and bend one slowly up a step. If done right, you can hear a DESCENDING note as you bend upward.

Merlin points out that circuits that produce even-order harmonics (e.g., Fender tweed) are subject to intermodulation, and this contributes to perceived "fatness" and "warmth." On the other hand, stages with only even-order harmonics do not intermodulate, which is one reason high-gain "metal" amps are designed for more odd-order harmonics (less flab, tighter bass response).

I immediately thought of one remarkable example from rock music, where the guitarist demonstrates this phenomenon, while no other band members are playing--you couldn't ask for a more clear example. I'll give you a chance to guess, or to list other notable examples.

[John_P_WI]

I immediately thought of one remarkable example from rock music, where the guitarist demonstrates this phenomenon, while no other band members are playing--you couldn't ask for a more clear example. I'll give you a chance to guess, or to list other notable examples.

Rush - Working man solo....

[xtian]

John wins, and spoils the suspense.

http://youtu.be/h59mDlBSt7o?t=6m35s

[John_P_WI]

Sorry, to be a buzzkill Xtian, Alex's style (Working man and 2112) was very impressionable on me when I young and that solo is burned into my mind....

[Milkmansound]

this intermodulation is a nightmare for us steel guitar players whose only job is to bend strings

I find that by overbuilding the power supply, it can be reduced, but in a low power fender style amp it can cause a bad hair day

[Leo_Gnardo]

You can get sum & difference notes right off a guitar even with a very clean amp.

But you'll also get sum & difference with 120 Hz or 100 Hz (depending on the frequency of your power supply) in an extremely overdriven amp, or one with insufficient or failing filter caps. Usually I can get it by playing around the 12th fret on G or E strings & it sounds like some out-of-tune grunting along, almost like a badly tuned ring modulator.

[Aurora]

Harmonic and intermodulation distortion are the two sides of the same coin. They are both the result of the same non-lineariti(es) in an amplifier circuit. Somehow I fail to see how PSU deficiencies can influence this, except for making the amp circuit more or less linear by altering the working point(s) of the circuit.
Hum frequencies will of course enter the spectrum and having the same treatment, depending on where the hum enters the circuit.

[cxx]

I guess we are talking about two things. Modulation between notes played (with each other) and modulations between the notes played and the power supply.

I think modulation between notes is basic to how we hear. How we know they sound good together. I especially like it with a little help, like an old fashioned octave that has that ring modulator character.

Modulation with the 120 Hz coming of the power supply is double edged. Too much and you get irritating ghost notes. Too little and you miss it terribly. Without any it's all flat and sterile.

[Milkmansound]

Modulation with the 120 Hz coming of the power supply is double edged. Too much and you get irritating ghost notes. Too little and you miss it terribly. Without any it's all flat and sterile.

exactly what I tell steel players who order amps from me. When its not there (solid state amp) the sound is limp and sterile

too much and its impossible to play above the 12th fret

its all about the sweet spot

[JazzGuitarGimp]

I agree that IMD can increase perceived 'fatness', but only in strict moderation. Too much IMD, and the amp is completely unplayabe. I should state for the record, that I do not like ring modulation either...

[Randall]

I like to use that Alex Lifeson riff to show off my 5E3's.

[Reeltarded]

uhhg sustain killer

[Leo_Gnardo]

I should state for the record, that I do not like ring modulation either...

Well when Jeff Beck plays thru his ring mod . . . . the man can do no wrong.

Gonna have to revive my old Maestro ring ding, whip up some disturbing noises. We had a prime Bode ring mod in the SUNY Albany Moog lab, probably got heaved into the dumpster along with the rest of the rig. Drat!

[gingertube]

In a Push Pull amp there is another way to get rid of guitar signal + power supply ripple intermodulation products.

What you need is a common resistance in the cathodes which generates some common mode feedback to the 2 pentodes.

What works best is a resistance equal to about 12 to 15% the value you would use on each tube for cathode bias.
Example: EL84 will niormally run 270 Ohms on each tube for cathode (auto) bias. Use a fixed 33 to 39 Ohm (270 x 0.13 = 35.1) resistor.

1) For Fixed bias use a common resistor of that value between cathodes and 0v and reduce the fixed bias voltage slightly to get back to the same bias point.
2) For Common Cathode bias resistor with a bypass cap. Split the cathode bias resistor into 2 resistorss with bottom resistor value as above and top resistor value whats left to make up to the original value. Put the bypass cap across the top part ONLY.
3) For Separate Cathode resistors each with a bypass cap then lift the common 0V of the 2 bypass caps and put in a resistor of the value worked out above, between the bypass caps -ve and 0V.

In all cases what you are doing is generating some AC COMMON MODE feedback. This gets rid of most of the intermodulation with the power supply ripple products and tends to also reduce odd harmonic distortion somewhat.

I run this on my HiFi Amps all the time but it should also work on Guitar Amps. Sonically it removes some "Mud" from the sound, particularly on bass notes

Cheers,
Ian

[xtian]

That is very cool. Thanks, Ian!