Constant Current CMOS Fuzzer

[gez]

Had a little time spare this morning so breadboarded this:



It's just the bare bones of a circuit and I'm not recommending anyone build it.  The circuit is based on the Charles R Fischer idea (see the EM distortion article/schematics at GGG) of using a constant current source to power an inverter, biased as a linear amplifier.  OK, this is a poor man's version (I don't have the parts for the EM), but it has some advantages as will be seen in a moment.

Not much info in the supporting article at GGG, but my understanding is that in its quiescent state the 'composite' channel within the last inverter is as 'wide' as it's ever going to be - the p-channel MOSFET's effective resistance equaling that of the n-channel's - so max class A current flows (this is 'ideally', in reality it seems there's a bit of give and take). With an input signal present one device will always be turning off as the signal increases in amplitude.  This means the effective resistance of the composite channel will increase, so with a constant current source feeding the channel the voltage at pin 11 will rise.  In turn, the axis at which the output swings around - approx half supply - is being pulled up too, which should compress negative output swings and expand positive swings.  I say should as the amount is subtle and wasn't easily seen using the scope (but I didn't spend that much time on it…need to revisit this later in the week).  However,  on pin 11 a full wave rectified version of the input signal - albeit uneven - could clearly be seen, so I think my reasoning is right. 

Sound?  I wired up the chip as shown as it enabled me to connect pin 11 to either pin 13 - the constant current source - or V+, which meant I could A/B things.  With the constant current source there is  noticeably more compression/sustain, perhaps more than could be accounted for by the slight increase in gain/decrease in headroom as a result of dropping the supply to 7V?  It definitely sounded better, but the circuit is a little crude as shown, so nothing stellar.   Initially I had the thing set up for variable gain in the input stage but with the guitar signal divided down below a certain level the circuit sounded horrendous; it seems that the distortion caused is too much for small signals.

The constant current source is a little crude, but works well enough.  Moving the wiper upwards cuts off the p-channel if you take it too far and things sound really bad, as you'd expect.  Also, things get noisier as less current is applied to the inverter.  7V at pin 11 was a good compromise.

Perhaps this will inspire some of you to build the Fischer circuit (I doubt I'll have the time).  Be sure to report back if you do.

http://www.generalguitargadgets.com/index.php?option=content&task=view&id=174

[gez]

Er, it uses a 4007 chip (I'll alter the schematic to reflect this in a min).

[Dragonfly]

looks interesting...

also...Mark Hammer has a booster called "The Crank"...you might wanna change the name to avoid confusion...

AC

[gez]

Bugger, didn't realise (sorry Mark).

I'll alter it in a min (I'll leave it nameless otherwise this could go on all night!!).

[Dragonfly]

Bugger, didn't realise (sorry Mark).

I'll alter it in a min (I'll leave it nameless otherwise this could go on all night!!).

i know...ive run into the same problem several times !

AC