Discrete TubeLess...

[puretube]

The "SANSVALVE" :



Actually, RP & RL can be omitted, and be replaced by a bare wire (link)

block schematic here...

inspired by a post from gez in brett`s other thread.

good as is as a standalone "grinder" for humbucker type guitars.

Modifications coming up...

[gez]

 

Did you match the transistors?  If not, whereabouts do the input/output bias up at?

[gez]

Ton, almost forgot to mention.  You can take the signals at the sources and feed them to diodes to create an octave up effect (works the same as a phase splitter).  You might need to increase the value of the source resistors to do this though: I did this with 4007s a few years ago and I think I had to use quite high value resistors to get a decent signal voltage across them.  The beauty of this is that you can still take the signal from the 'normal' output. There should be some info in the archives.

[puretube]

ahh - there you are 

no matching - just grabbed the first ones out of 2 bags...

automatic exactly 4,5V (4V5)/(4.5V) biasing!

and that`s what stumped me, too:
no datasheet or schoolbook description has been able to explain to me why this actually happens...
(is it, like in this case, that I apparently happened to grab 2 matching P / N devices with nearly identical U_GS/I_D,
and that they just by accident are balancing?).

[puretube]

Ton, almost forgot to mention.  You can take the signals ...

HUSH! 

[puretube]

Modifications:

1.) "stone"-mod: shortcircuit (= insert a link instead of) R_S.
2.) "metal"-mod: shortcircuit (= insert a link instead of) R_D.
3.) "wood"-mod: shortcircuit (= insert a link instead of) R_S & R_D.

1.) & 2.) will produce "dirtier", louder, less compressed distortion (asymmetric),
while 1.) is treblier and less aggressive than 2.).
3.) will be softer (almost symmetric) and more dynamic than the basic version and than 1.) & 2.)...

2.) is noisier than 1.)... ("hiss")

(gez: 1.) & 2.) do severely unbalance the DC offset at the output: 6.5V / 2.5V respectively, instead of 4.5V...)

haven`t measured the current consumption of mod 3.)...

ps:

" R_P " & " R_L " are the protective and/or current limiting resistors,
inspired by cmos-inverter datasheets,
but of no influence, soundwise...
" R_M " can be shortcircuited (linked), but otherwise defines a certain minimum "drive" (gain-factor),
which can go from <1 to ~80 times... (the latter for small input volumes up to ~25mV)*

*anything louder than that, will be squashed/distorted at higher drive/gain settings**

** signals lower than ~250mV can be boosted clearly by a factor of <1 to >10.

Max. output signal voltage (@ 100k load) is ~ 4Vpp; (~6Vpp for mods 1.) & 2.)...

[magikker]

Thanks for posting this, I'll have to break out the bread board later and take a look.

[The Tone God]

I played with something almost similar for the Mosfet FX-X. I didn't get it done in time though so I ended up with what I entered. Noticed my entry had BS170s and BS250s ? Left overs from playing a similar idea. I'm glad someone got something in that vain out.

Thanks!

Andrew

[puretube]

ooops - hadn`t noticed...
came up with this coz I didn`t have non-SMD inverters at hand to quickly view what happens with brett`s "deadshort"
in the other thread... (thanx, gez!),
and I really wanted to know (but still don`t...).


(the "deadshort" secret probably lies in the fact, that it is being (DC-coupled...) biased to half-supply by the previous stage...)

http://www.diystompboxes.com/smfforum/index.php?topic=59106.0

please note, that Brett`s R8  is a " R<sub>D</sub> " for all stages simultaneously,
coz it`s not de-coupled at pins 1/8... 

[The Tone God]

ooops - hadn`t noticed...

I did try to create something similar to enter but ran out of time. Instead I turned the Mosfet pair into an one shot timer for a noise maker called Forsaken.

Andrew

[jaytee]

Looks good that. I've been thinking about this since the comments on the other thread. I'm wondering did you try plugging some random samples in to see if it always biases half way? I don't have any P mosfets to try it out. I have a feeling a similar thing can be done with jfets with different biasing. Does it sound good with a guitar through it?

[markm]

I forget where I read about this but, one of the "older" electronics manuals referred to this as a "totem-pole".
Wish I could remember where I saw it though. Now I feel older than I did a minute ago.....

[Dragonfly]

Geez...now I've gotta build this one....        

Looks cool PT...as usual !

[jaytee]

I tried simulating it. It has a gain of 600 before it clips. The input impedance looks about 1.5k. Is that right or what? That's without the 100k gain control.

[gez]

haven`t measured the current consumption of mod 3.)...

What happens (sound wise) if you use source bypass caps?

[gez]

PS  Have you seen the new 'members only' area? 

[puretube]

gez: initially, I did the Rd/Rs bridging with caps...
but when I found those mosfets so perfectly matching,
I wanted to save parts...

This morning I did some random mixing of several BS170 (from 3 different manufacturers (dunno, which)),
with several BS250 (from one batch):
they all stayed within 150mV of Ub/2!

However the 3 bs170s from one manufacturer sounded harsher with either bs250,
2 from an other sounded "different",
while 10 other bs170s from one batch all were "cool" (=warm) sounding and at the exact same bias.

So: manufacturer does matter in sound!
(btw: I have not tried other P- or N-MosFets than the 170/250s!)

Since I noticed the "harsher" ones got quite hot with Rd/Rs shorted by wires,
I now again prefer the bridging with 10µ caps, especially with unknown transistor types.
(although this doesn`t sound as saturated and smooth, when driven hard...)
When done with caps, you probably  can omit the safety resistor Rl safely.

Jaytee: simulations are one thing... 
I measured ~80mV out for 1mV in, @ open as well as with 4M7 feedbackresistor,
with the "gain"-pot at zero.

With the sound I hear coming out of this, I don`t care about the input impedance 


ps: oh yes - the circuit accepts/welcomes integrating ("de-hissing") caps from out-to-in like all those other integrated
multilegged tempered halfused hexcircuits...

[gez]

This morning I did some random mixing of several BS170 (from 3 different manufacturers (dunno, which)),
with several BS250 (from one batch):
they all stayed within 150mV of Ub/2!

Wow, that's really interesting.  I wonder why (seems counter-intuitive)?

Nice to know that if the 4000 series ever goes the way of the dinosaurs there's an alternative.  We'll just need to power these circuits from the mains in the future...

[puretube]

from mains to headphones? 


btw:
noticed another thing:
similar to not liking zero Ohm between out & in,
they don`t like to be DC-coupled in cascade: so better use caps between consequent stages... 


oh, yes: "piggybacking": forget it... (zero improvement)

[jaytee]

I tried random pairs of fets (simulation) and they all balanced up. It would be interesting to see what happens with a buffer in front of it. If the input impedance is so low you might not be getting full gain. The standing current might be quite high because theres 4.5v bias on the gate. A lower supply voltage or bigger source resistors would reduce it.