Diode selection for clipping in ROG style input stage (Azabache, Britannia) ?

[ct_anthony]

Hello all,

I got a question related to the clipping diodes used by ROG in the stages of the Azabache and Britannia.

They are there to prevent the fet from hard clipping.

What  I currently don't understand is, that a typicall red LED will have a forward voltage of about 1,8v which is above the typicall Vp (about 1,5v) of a 2N5457.
So I guess the jfet will be able to do some hard clipping... but not full out... what is the deal here? Are there any rules how to calculate/design/estimate the diodes
or more specifically the forward voltage of the diodes against the Vp of the specific fet?

Regards,
anthony

[Frank_NH]

Interesting question.  I too have wondered about the efficacy of the diodes on the JFET inputs.  I can't imagine a guitar signal that large at the input (+/- 1.5V) unless you put a booster in front of the ROG amp sim circuit.  In any case, is it bad if the JFET itself clips?

[ct_anthony]

I´ve build a simple Fetzer Valve with a 2N5457 and a Vp of 1.26V on Breadboard...

I´ve put another boxed Fetzer with Gain of approx. 8 in front of it to "overdrive" the input using a SD JB Humbucker.

Adding different clipping diodes like in the first stage of the Azabache or the Britannia gave the following result:

• 1N914 with Fv of 0.6V: Sound was like the average Diode based Distortion into a Fetzer stage. You could here different stages doing different things.


• Red led with Fv of apron 1.7V: Clipping sound got a little bit "rounder" than without the clipping diodes. Harsh Pickattack. good decay.


• Double-1n914 in series with Fv of approx 1.2V: I was not able to tell which component is producing which part of the clipping, softer/rounder as without diodes, nice decay

So I think it is best to match Fv of the diodes and Vp of the JFet as close as possible?

[Frank_NH]

"So I think it is best to match Fv of the diodes and Vp of the JFet as close as possible?"

I suppose it depends on whether you want the diodes to clip a hot input signal or not.   In your example, with the LEDs, the input signal was limited to ~ +/- 1.6V by the LEDs.  But since the JFET has a Vp of 1.26V, the JFET will still clip the pre-clipped signal, which is perhaps why it sounded a little different.

If would be interesting to repeat your experiment with a less hot signal (say +/- 1V).  In this case, I would expect that the LEDs would not have any effect, whereas the 1N914 diodes would preclip the signal.

[PRR]

Those input LEDs do nothing. As you say, the JFET will be totally schmeared long before the LEDs wake up. There are no capacitors to saturate during overdrive.

They are not even protection. The 33K into the JFET is fine up to 100V steady and 300V momentary. If you have such levels on your stage, it isn't the JFET that is in danger.

[ct_anthony]

Small update:

I've played with different Jfets and different diodes during the last time...

The best results where given when I selected the diodes (or sereis diode combinationes) so that Vf is about 0.2V - 0.3V below Vp of the JFET.

Typical examples:

• 1n914 / 1n4148 for J201
• 2 x 1n914 in series for 2n5457
• red LED for 2n5458

With this setup the diodes and the fets are "melting" together and act as one unit, if that makes sense...

If would be interesting to repeat your experiment with a less hot signal (say +/- 1V).  In this case, I would expect that the LEDs would not have any effect, whereas the 1N914 diodes would preclip the signal.

When the Signal is below Vf of the Diode then there is no clipping.