New at runoffgroove.com: BRITANNIA

[stm]

Hi!

It's been more than a year since our last project was released.  We are finally ready to share our new circuit, BRITANNIA, which is an overdrive inspired by the Vox AC-30 Top Boost amp, and a major revision of our previous work, the English Channel.

Hope you like it!

link: www.runoffgroove.com/britannia.html

[nate77]

Awesome. Thanks man, I love your projects, certainly some of the best in the effect community, and injected into the diy community as readily as you do is priceless. Thank you!

[sajy_ho]

+1 to runoffgroove.com

[samhay]

Looks interesting, and nice demo from Jon.

[commathe]

This is the first time I've seen diodes before fets like this. Very interesting! I look forward to breadboarding this.

[samhay]

This is the first time I've seen diodes before fets like this. Very interesting! I look forward to breadboarding this.

http://www.runoffgroove.com/azabache.html

[mac]

This is the first time I've seen diodes before fets like this. Very interesting! I look forward to breadboarding this.

Take a look at the links below,

http://www.sugardas.lt/~igoramps/article68/article.htm

http://www.pat2pdf.org/patents/pat5619578.pdf

mac

[commathe]

Thanks guys!  Those also look great

[Jopn]

Well there goes my weekend.

Very cool guys, I'll be building this one asap.

[Davelectro]

This is the first time I've seen diodes before fets like this. Very interesting! I look forward to breadboarding this.

Take a look at the links below,

http://www.sugardas.lt/~igoramps/article68/article.htm

http://www.pat2pdf.org/patents/pat5619578.pdf

mac

Those are good examples of diodes + transistors, but the ROG approach is (obviously) different.

ROG circuits sound beautiful. No doubt about it. Now, I must admit I don't quite get the logic behind using diodes to ground right before a supposedly triode-like FET stage. If their clipping is hard and symmetrical, why are FETs still required? Why not cheap and easy to find BJTs + asymmetrical diode clipping?

[Bill Mountain]

This is the first time I've seen diodes before fets like this. Very interesting! I look forward to breadboarding this.

Take a look at the links below,

http://www.sugardas.lt/~igoramps/article68/article.htm

http://www.pat2pdf.org/patents/pat5619578.pdf

mac

Those are good examples of diodes + transistors, but the ROG approach is (obviously) different.

ROG circuits sound beautiful. No doubt about it. Now, I must admit I don't quite get the logic behind using diodes to ground right before a supposedly triode-like FET stage. If their clipping is hard and symmetrical, why are FETs still required? Why not cheap and easy to find BJTs + asymmetrical diode clipping?

The diodes are there to keep the FET's from hard clipping.  The diodes set the max input voltage and then the FET boosts that and possibly clips.  If the diodes clip it's no major crime because they clipping will be masked by all the eq and gain stages going on.

[Davelectro]

Yes, I can see the LEDs are there to keep the fets from hard clipping. But my point is LEDs and regular diodes will clip hard too if signal is large enough. So we are sorta preventing hard clipping by means of (potentially) hard clipping.

The question is why in this case is hard clipping from the diodes more desirable than hard clipping from the FETs? I can only think the cumulative effect is key; the whole thing working like some sort of diode ladder under serious clipping.

[Lurco]

Diodes to ground through resistors like 33k, 100k, 18k will hardly clip "hard".

[blackcorvo]

I'm seriously thinking of building this circuit and using it as a preamp for a TDA7240. It would make a lovely combo with the 10" speaker I have here!
I wonder if there's a way to add a tremolo to this circuit...? Hmm.

Oh well, I guess it's layout drawing time!

[Davelectro]

Diodes to ground through resistors like 33k, 100k, 18k will hardly clip "hard".

But those Rs are not in series with the diodes. I see no voltage divider providing any extra softness.

[ggedamed]

Diodes to ground through resistors like 33k, 100k, 18k will hardly clip "hard".

But those Rs are not in series with the diodes. I see no voltage divider providing any extra softness.

If you're talking about source resistors, they're not really involved in the diode clipping. Lurco was talking  about the "grid stoppers", the resistors before the diodes pairs and in series with JFETs gates or the non-inverting input of U1a. The JFETs gates and the non-inverting inputs of JFET/MOSFET op amps have impedances so high you can say they're not there. So the only path to the ground is through those resistors and then the diodes (there is  the 1M bias resistor on the op amp input but it's large enough to not count).

[Davelectro]

No, by Rs I meant "resistors", not source resistors. Sorry if I wasn't clear.

As far as I know, diodes to ground need some resistance in series with them to provide a softer clipping (since this resistance forms a voltage divider in conjunction with the source impedance + any resistance in series with the signal path).

And I'm not really sure about this, but at least from what I've seen in software simulation, large source resistances like 33K, 100K, 18K, etc. by themselves (NO divider) seem to make clipping start sooner, not softer.

[Lurco]

Let`s take the first pair of 1N4148 diodes http://www.runoffgroove.com/britannia.png. There`s a voltage at the wiper of the gainpot, which will be roughly 3V maximum on its positive excursion. The 100k resistor and the left diode form a potentiometer for these positive 3Volts, where the "wiper" goes to the gate of Q2. According to the dynamic forward resistance http://datasheet.eeworld.com.cn/pdf/CHENYI/56576_1N4148.pdf fig.2, the lower half of the "potentiometer" will vary its value during the excursion of that max. 3V wave. At the max. point assume 3V divided by the 100k+dynamic forward resistance Rd of the diode (~2k @ 30µA), multiplied by  Rd gives ~60mV at the gate of Q2. At lower values of the positive wave, the current through 100k+Rd will be reduced, and Rd will increase so that the "wiper" will go upward, thus reducing the "potentiometer"`s attenuation for lower input volumes. IMHO this describes an automatc gain control based on the curve of mentioned fig.2. This curve doesn`t show a sharp bend, and at larger inputs to the "pot" there`s but a low voltage across the diode.
For the negative half of the wave, the right diode takes over.
Where did I go wrong?
http://www.kennethkuhn.com/students/ee351/diode_characteristics.pdf

[Davelectro]

So hard clipping was not that hard after all. 

Interesting.

[scintillation]

Thinking about building this for recording guitar jams/headphone amp. The sounds from clip 1 on the website sold me!

I see some shaping in the final stage, so would a ROG condor not be necessary on the output of this?

My hope is it would fit into my set-up like this:

guitar -> pedals -> britannia -> soundcard -> pc.

Or headphones instead of soundcard.