Re: gEDA-user: Signal Source Setup for Electric Guitar?

[Andy Fierman <andyfierman@xxxxxxxxxxxxxxx>]

Quick tip on your schematic: it's a good idea to offset 4 wire
junctions (C1, R1, R2 & Q1b) so that they appear as two pairs of 3
wire junctions. That way even if the junction dot disappears in a .png
or .pdf or whatever, it is obvious that the 4 wires are all joined and
it's not just 2 wires crossing.

The reason you're seeing such distortion with your first attempt is
that the R1, R2 base bias potential divider feeds current into Q1b so
this node sits at between about 0.55V and 0.7V above ground (let's say
it's 0.6V and call this Q1vbe).

Your 1V ac source (V2) is driving 1V pk, i.e. 2V pk - pk directly into
C1. It is a pure (zero R) voltage source so R5 has no effect at all.
The signal swing at Q1b is then offset by Q1vbe so whenever V2 drops
below -0.6V, Q1 is cut off.

If you look closely at what happens to Voutput when V2 swings through
-0.6V, you'll also see that Voutput is increasingly distorted.

This is a long way of saying that you are right: you need to use a
much smaller input voltage swing.

And yes, you can probably get away with using some sort of high
impedance buffer in front of a sound card or even your on board sound
but for it to be of much use, the buffer has to be very linear, have a
suitably wide input voltage and frequency range and you'll need some
sort of reasonably calibrated signal source with which to calibrate
what you can see on any waveform display. If you don't have those
things then you can't relate what you see on the display with what is
actually happening at the point you are trying to measure.

By the time you've sorted all that out you may find it easier and
cheaper to get a reasonable quality PC scope card or even a 2nd hand
scope. Don't forget that you need to allow for the effect of the
connection you make into your circuit at the point you are trying to
measure it too. Connecting a 1Meg // 50pF scope input across a 10Meg
buffer input is not very helpful! Even x10 probes (notionally
10Meg//5pF) are not that much use.

(You can reconcile the scope and simulator displays if you model the
effect of the scope probe in the schematic too but that is getting a
bit esoteric.)

Whatever, Spice is a good tool for learning about electronics. With
gEDA/gaf the tools and all the components are free and you can't even
blow anything up!

That said you do need to sanity check simulations for things like
silly voltages or currents and power dissipations that would have
destroyed real devices.

: )

2009/5/13 Gareth Edwards <gareth@xxxxxxxxxxxxxxxxxxxx>:
> 2009/5/12 Joerg <joergsch@xxxxxxxxxxxxxxxxxxxxx>
>> >> For guitar pickup modelling, you might like to do a bit of Googling on
>> >> "spice model guitar pickup" or similar.
>> >
>> > Yes, that is a very interesting page. This is precisely the kind of
>> > thing I'm interested in. I had no idea the capacitance of the cable
>> > was so significant.
>> >
>> > Anyway it looks like their AC generator is using 2mV. So the 20mV
>> > value I used to get a good output SIN looks closer to reality which
>> > means my model is probably ok.
>> >
>> > I wish I had a real oscilloscope to find out what my guitar is really
>> > putting out but xoscope doesn't seem to produce voltages (but I'm not
>> > surprised since the sound card is probably oblivious to such things).
>> > Maybe I'll have to get one of those PC oscilloscopes.
>> >
>
> Unless you are building a pickup, don't get too hung up on modelling
> it - it's a complex and changing impedance across frequency (stray and
> interwinding capacitance, leakage inductance, DC resistance etc) and
> the unit-to-unit variation in impedance and output voltage is large.
> You've already spotted that cable capacitance plays a part - the tone
> and volume controls loading the pickup are probably just as important.
>
> The trick is to use a high impedance amp front end that isolates the
> pickup from the rest of the circuit (JFET or, my personal
> preference/prejudice, a valve triode stage); then you can simplify the
> pickup model to a simple generator without worrying too much. In my
> amp work, I use a 100mV pk nominal signal - this is recommended by
> Kevin O'Conner in his excellent Ultimate Tone series of books:
>
> http://www.londonpower.com/catalog/index.php?cPath=3
>
> However, modelling can only get you in the ballpark of gain staging
> and frequency response - eventually you have to evaluate with your
> ears, not with SPICE!
>
>>
>> If you want to display the waveforms or an FFT spectrum on the PC
>> cheaply you could use an emitter follower as a buffer so the voltage
>> gets transferred 1:1. Then a high input impedance is provided to the
>> guitar even if you connect a sound card.
>
> Agreed, guitar into soundcard is sonic disappointment but JFET source
> follower would be much better. Here's a good one (not mine, but
> derived from a Win Hill circuit he posted on Usenet):
>
> http://www.ciphersbyritter.com/RADELECT/PREJFET/JFETPRE.HTM
>
> Cheers
> Gareth
>
>
> _______________________________________________
> geda-user mailing list
> geda-user@xxxxxxxxxxxxxx
> http://www.seul.org/cgi-bin/mailman/listinfo/geda-user
>


         Andy.

http://signality.co.uk


_______________________________________________
geda-user mailing list
geda-user@xxxxxxxxxxxxxx
http://www.seul.org/cgi-bin/mailman/listinfo/geda-user