Guitar one octave Up

A phase-locked loop. The CD4046 plus a divide by N is the classic "PLL

101" approach. TI has a pretty good app note on this and there are tons of "phase locked loop" examples on the 'net.

--
Rich Webb     Norfolk, VA

Not one that can track changing music input.

But perhaps you could make an analog squaring circuit to double the input frequency. ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
      The only thing bipartisan in this country is hypocrisy

Whatever you do will, I think, give you tremendous distortion but may not make the thing sound a higher tone -- I suspect that at best you'll take out the fundamental, but leave the odd-order harmonics.

Nay-saying aside, a squaring circuit would be the first thing to try -- my knee jerk reaction would be to use an analog multiplier (goodness those things are expensive now!) with both inputs connected to your signal. Perhaps a _really fast_ AGC on _one_ channel, so that the output envelope is proportional to the input.

If that doesn't work, I'd try a phase shift network like the old-style "phasing" single-sideband receivers used -- this is a set of all-pass filters that shift phases in both channels such that the resultant phase is 90 degrees apart. You'd have a much better chance of getting rid of the DC bias on the output, and possibly of getting rid of the odd-order effects. But all-pass filters tend to have nasty effects on audio quality*, so even if it sounds better one way, it may not be worth it.

• I'm told. Personally, I have a tin ear. I think it's because they're not minimum phase, which is also the 'real' problem with much of the DSP that gets done.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com

True enough. I assumed he was looking for a tuning gizmo rather than a real-time doublers.

--
Rich Webb     Norfolk, VA

On a sunny day (Fri, 26 Mar 2010 09:11:57 -0700) it happened Tim Wescott wrote in :

I think 'double phase rectifier on small audio transformer' needs no battery at all... The distortion will sound like real good old analog hehe.

|\|

---- ----| |--------- )||( |/| | f )||(________ 2f ___| )||( | )||( |\| |

---- ----| |--------- |/| LOL

An aggressive low pass filter before the multiplier so that only the fundamental is used would help things a lot. A variant of a ring modulator circuit based on the good old MC1496 ought to do it.

One digital way that would work and preserve most of the guitar waveform characteristics is to sample it with a moderately fast ADC into a large circular buffer and play back the waveform between rising zero crossings stepping odd samples on pass 1 and even on pass 2. The devil is in the details of the implementation. There may be novelty voice doubler chips that will do something like this out of the box.

Regards, Martin Brown

AD633 has a bounceless doubler in the app notes in the data sheet.

Steve

A buffer, followed by a full wave rectifier, followed by a filter to smooth out the bottom 'points' ?

Charlie

This is not a phase shifter but simply a filtering problem. You must first determine the fundamental, which is a difficult process in and of it self, then use a sharp filter and or notch filter at the appropriate frequency.

With digital electronics one can do this relatively easy. Modern day digital electronics, if done right, can out-perform analog. The reason analog sounds better, in general, is because inaccuracies in analog are generally "smooth" and do not create unnatural frequencies in the original signal. IMO, if you are going to do this you should do it digitally first. The reason is that it is much much easier and may not have any of the drawbacks in quality you believe it will. You'll just need a modern 24-bit 96khz ADC with possibly oversampling and a processor(a few other bits and pieces). Then you can simply code everything in software. Using this technique will also allow you to easily extend and modify the algorithm along with doing other effects.

In any case, as far as doing it in analog, the first thing you'll need is a voltage controlled filter and a frequency to voltage converter.

The idea is that the frequency to voltage converter will output a voltage proportional to the fundamental which will program the voltage controlled filter. Together they represent a frequency controlled low pass filter. I.e., the frequency sets the cutoff point.

This is not necessarily hard to do but it might be hard to do well. It will not work well for more than one note and you may have to design a feedback system so that the frequency to voltage converter will lock onto the fundamental and not other frequencies.

There may be other tricks you can use such as squaring the signal which will emphasize the largest harmonics(which tend to be the first few harmonics) which can then be normalized again resulting in a special type of frequency controlled low pass filter.

A full wave rectifier is the simplest way as has been suggested. Make a phase inverter with equal resistances in both collector and emitter legs of a transistor amp or plate and cathode if you want to use vacuum tubes, or op-amps or whatever. Capacitivly couple both phases of the outputs to diodes being sure there is a DC path to ground, then combine the two diode cathodes, ala full a full wave rectifier, and filter with a low pass RC net. Capacitivly couple the output and amplify. It will probably sound like crap but it is simple. It will give a spray of harmonics with a dominant second.

Another way, much more complicated but probably a purer second harmonic is to use single sideband RF techniques. AM modulate a carrier with the original signal. Strip the carrier and other side band leaving only one side band. Re-insert a new carrier exactly 1/2 the frequency of the first carrier then demodulate the resultant signal. Filter as appropriate to eliminate spurious signals. The output should be the second harmonic of the original. This is an analog version of the digital signal processing methods but uses modulation instead of sampling

Yep. I was going to write that up and simulate, but I went to lunch and had a bottle of wine instead :-) ...Jim Thompson

--
| James E.Thompson, CTO                            |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
      The only thing bipartisan in this country is hypocrisy

Can a 12 string guitar be adjusted so string pairs are an octive apart?

look up the PDF file for the LM1496 chip.

Please note that the signal out of an electric guitar may contain up to six separate notes being played at the same time. Since the OP didn't say otherwise, I would say that advanced DSP is his only hope without having all sorts of weird spurious self-heterodyne signals popping up all over the place.

Run the guitar through autotune?

-- Paul Hovnanian mailto: snipped-for-privacy@Hovnanian.com

------------------------------------------------------------------ Yesterday upon the stair, I met a man who wasn't there. He wasn't there again today. I think he's with the CIA. -- apologies to Hughes Mearns

"Bob Eld"

** Absolute nonsense.

What is described above is the basic process of creating and receiving a SSB radio signal - but with a big error.

When receiving a SSB (speech) signal, one re-inserts a carrier on the SAME frequency as the missing one - this has to be done very precisely for the recovered speech to have a normal frequency range. Any difference in the two carrier frequencies cause a corresponding frequency offset in the speech.

This means the recovered speech is FREQUENCY SHIFTED compared to the original - a process entirely different from creating a multiplied version of the speech frequencies.

..... Phil

at all...

That right there is basically the heart of the "Octavia" octave fuzz pedal that Hendrix used.

is

SSB

version

OK, scratch that idea. So reinserting a different carrier is basically just heterodyning the side band up (or down) to a different frequency, translating it but not multiplying it. Thanks for the correction.