Distortion Prototyping in LT Spice

Started by doitle, July 08, 2009, 09:11:53 PM

Previous topic - Next topic

doitle

I was wondering if anyone else does this with much success? I don't know how accurate a result it is but if the models are true to life I'd expect it to be fairly similar. I recently tried it with a single stage of triode tube amplification and the results sound pretty cool I think. The only problem is these 10 second clips took a few hours to run! :P

Here's the circuit I was running them through:


Here's what I ran through it:

Fly By Night: Input - Output

Say It Ain't So: Input - Output

Livin' On The Edge: Input - Output

I was just interested to know if anyone else does this or has some cool example circuits of this kind of thing.

brett

Hi
LTspice certainly is a LOT of fun.

I was surprised, however, that you built a valve amp and then used diode clipping to ground.  That is likely to give fairly similar result to a single BJT stage followed by diode to ground (except maybe for a higher Zin, and if that was the aim, a JFET would do the trick).

To get real magic out of a triode from an unamplified guitar signal, you should probably bias it "off-centre", to get some valve-based clipping.  In my opinion, biasing high for some grid conduction is always good (especially if driven from a low Zout source like a Tubescreamer), but biasing near cutoff works fine, too.  Some circuits, like the Real McTube, use 2 mis-biased triodes in series to take big "chunks" out of sides of the waveform.

cheers
Brett Robinson
Let a hundred flowers bloom, let a hundred schools of thought contend. (Mao Zedong)

doitle

I was mostly just goofing around after I finally got tube simulation to work. I was actually using the diode clipping as an attenuator of sorts. Left to it's own devices this would have generated voltages well above where the "cutoff" is for the wave output. Without the diodes it was a completely square signal as the output signal was several volts in amplitude. I have not figured out yet if there is any way to scale the node when it is converted to the wave file or what the best way to prevent this was but the diodes did work and I thought it sounded kinda neat.

I will attempt to lookup off-centre biasing though to see if I can understand what you mean. I will also see if I can find any info about the Real McTube to learn more.

valdiorn

Idea: Use the transformer from a 6V wall wart instead of the voltage-multiplier. Then, get yourself a 12V wallwart to power the pedal. Connect the 12V to the 6V input on the other transformer, get 240v out on the other side, but with complete electrical isolation.