Fetzer valve as buffer?

[PBE6]

Does anyone think or know if the Fetzer would be better used at the end of a distortion circuit? Or will that virtually guarantee that the Fetzer will clip?

[pappasmurfsharem]

Davelectro I thought the MPF102 would be the best choice because it has the highest Vp of the JFETs I have right now. I want to keep the JFET as close to the "tube" operating conditions as I can which means it shouldn't be clipping at all. I think if I go lower with a 2N5457 or 2N5458 I might run into trouble with humbuckers.

Again my Fetzer valve is squeeky clean with a j201 and humbuckers.

[pappasmurfsharem]

Does anyone think or know if the Fetzer would be better used at the end of a distortion circuit? Or will that virtually guarantee that the Fetzer will clip?

Its really designed as an input stage. You could use it after but if the incoming signal is above normal guitar levels it will probably clip.

[PBE6]

Sounds like the Fetzer isn't going to add much triode character at either end if distortion gets thrown into the mix.

I should breadboard it anyway, all speculation and no experience makes PBE6 a dull boy.

[sajy_ho]

This circuit was designed based on Child's law, to add some HARMONIC distortion to the sound, not CLIPPING distortion! These two are not the same.
The term overdrive/distortion guitarists use is basically clipping distortion, but tube like harmonic distortion will add some even-order harmonics to the sound that we call it as warm or fat sound. Those even order harmonics are the same note as the original note, but one or two octaves higher.
So as long as you do not overdrive this circuit you only will get clean sound. I think for 9V supply the best choice would be a 2n5457 or low Vp MPF102, Just buy some and try measuring Vp and find a part with Vp between 1-2 volts and design the circuit using its calculator.

[Davelectro]

my Fetzer valve is squeeky clean with a j201 and humbuckers.

Yes, i'm not saying that every single j201 will clip, but some of them might get into clipping under certain circumstances. It's definitely harder to overdrive a normal fetzer stage based on a lower gain jfet like 2n5458 or MPF102.

On the other hand, some mpf102 won't be able to amplify a given signal in a fetzer-type stage running on 9V.

That's why both the jfet tester and the fetzer calculator are important.

[pappasmurfsharem]

This circuit was designed based on Child's law, to add some HARMONIC distortion to the sound, not CLIPPING distortion! These two are not the same.
The term overdrive/distortion guitarists use is basically clipping distortion, but tube like harmonic distortion will add some even-order harmonics to the sound that we call it as warm or fat sound. Those even order harmonics are the same note as the original note, but one or two octaves higher.
So as long as you do not overdrive this circuit you only will get clean sound. I think for 9V supply the best choice would be a 2n5457 or low Vp MPF102, Just buy some and try measuring Vp and find a part with Vp between 1-2 volts and design the circuit using its calculator.

I can confirm there is definitely a pleasant sound when running the fetzer at unity volume vs dry guitar signal.

[PBE6]

sajy_ho - yes! This is exactly what I'm trying to get at. I don't want the Fetzer to clip so as to keep it in the "tube" range.

One question, why is -1 to -2 V better for Vp than -2.34? Wouldn't an MPF102 give me maximum headroom? I can make up additional gain later in the circuit.

OK, two questions: it sounds like distortion from a later gain stage will likely negate the intended effect of a Fetzer input stage. But what is the output voltage of gain stage with a pair of silicon diodes in the feedback loop? It is limited to about +/- 0.7 V, is it not? If that's the case, and the Vp of a JFET used in a Fetzer stage placed after that is -1 V or more, the Fetzer stage won't clip! Is that correct? If that's true, then maybe a Fetzer output stage can impart triode-like harmonics on a signal that's been previously clipped.

[PBE6]

To clarify my earlier post:

Input --> JRC4558 with pair of silicon clipping diodes in feedback path --> Fetzer output

Q1: Will the silicon diode pair clamp the signal to about +/- 0.7 V?

Q2: If Vp(Fetzer) is in the range -1 to -2 V, will the Fetzer output stage clip?

[sajy_ho]

Q1: Will the silicon diode pair clamp the signal to about +/- 0.7 V?

Yes they will.

Q2: If Vp(Fetzer) is in the range -1 to -2 V, will the Fetzer output stage clip?

No, that will not. But the thing is Fetzer was designed to simulate Fender's input stage and better to use it before clipping stages. If you place it right after guitar, the sound will warm up a bit (of course not like a tube, but more than a bipolar gain stage!), and then this warmed signal will go to the other effects like clipping stages, EQ, reverb and other stuff.

[sajy_ho]

Almost I forgot! Vp should be less than 1/4 of the Vcc, otherwise the gain will be less than unity! So for Vcc of 9V it should be less than 2.25V.
Lower Vp----> More gain - Less headroom
Higher Vp---> Less gain - More headroom
The choice will be yours, Just play around with Fetzer calculator and see what is the best for you.

[PBE6]

Thanks sajy_ho!!

[PBE6]

Started building this one, but I already have some confusing results.

I opted for a Fetzer input stage, that is working fine as is the James EQ section I put after it. I was breadboarding the JRC4558 section to see what kind of voltage I would be getting once the signal had been clipped, and instead of getting about 0.6 V as I expected, the voltage was about 1.6 V (with a 1 V input and ~100x gain).  This is probably workable, but I was planning to put a Fetzer stage at the end as well. The Vp for my 2N5457 is about -1.75 so this signal is still within the no-clipping range, but hotter signals won't be.

Does this voltage seem right? If so, I am definitely confused. Any insight would be appreciated!

[Pojo]

This probably needs someone more knowledgeable than me to really comprehend what's going on, but here's my take. The thing with non-inverting opamp stages with diodes in the feedback loop is that some unclipped signal gets through. If you listen close you can usually hear some clean tone sort of 'riding underneath' the clipped tone. So with a pair of 4148's, the clipped signal will be clamped to +-.7V for a total voltage swing of 1.4V. Seems reasonable to me that the blended clean tone could easily add another .2V to the total swing.

[PBE6]

Thanks Pojo..I should clarify that the test signal was +/- 1 V (2 V ppk) and the final signal was +/- 1.6 V (3.2 ppk), but that makes sense based on your explanation (1 V signal + 0.6 V from feedback = 1.6 V).

Can anyone explain the equation for non-inverting opamp gain with feedback diodes to me? I tried figuring it out earlier today but I got stumped =\

[PBE6]

Finally got around to finishing this pedal, it sounds promising but I think some changes could make it better (famous last words!!). I ended up using the following stages:

1. Fetzer input (2N5458, Vp = -1.94, Av = 1.4)
2. James EQ (pre-)
3. Opamp amplifier (to counter the EQ losses)
4. Opamp gain stage with silicon clipping diodes in feedback loop (2 forward, 1 back)
5. James EQ (post-)
6. J201 Fetzer (Vp = -1.0)
7. Second J201 Fetzer (Vp = -1.0)

(The last two Fetzer stages are there to counter the losses from the second EQ stage as well as add some faux-triode character.)

Overall it has a nice Fender-like bite to it, definitely changes the character of my solid state practice amp for the better, even with gain rolled all the way back.

The pre-EQ bass control really changes the character of the distortion, getting fuzzier with increasing bass, but the treble control is a bit too subtle (I think the opamp feedback capacitor is already taking out highs starting around 3 kHz so the treble control may not have much to work with). The post-EQ controls seem to work well, especially when set opposite to the pre-EQ. Dialing back the pre-distortion bass and then adding it back in post-distortion (and vice-versa with the treble) gives it a focused sound with some muscle but none of the fuzzy buzz.

The only other problem is that it's a bit noisy with the gain cranked, although changing the 1M gain pot to a 500k one tamed this fairly well without giving up too much grit. I'm guessing that the passive EQ sections and subsequent boosts are contributing the majority of the hiss. Is there a better way to arrange these stages to minimize noise? Or would a different EQ strategy be more effective?

Thanks to all who commented, I appreciate the help!

[sajy_ho]

Hi PBE6, I'm glad you did it
Your pedal is using lots of amplifier stages so it will be very sensitive to noise, but if you use shielded wire in the signal's path and also use only one point to connect your ground to chassis(star ground), the noise will be at minimum level.
You need a emmiter-follower/source-follower stage before the EQ to overcome high-frequency loss in the long signal path.
Use one of these:http://www.muzique.com/lab/buffers.htm

[PBE6]

Thanks sajy_ho.

The pre-EQ section should be OK with regard to impedance, the preceding stage is the Fetzer with an output impedance of roughly 1k (the ideal resistor values are quite low for a Vp of -1.94 V) and the following stage is an opamp amplifier so it should have very high input impedance. I used Duncan's Tone Stack calculator to find the James EQ values and the treble response was up by 20 dB at 10 kHz but but only up by about 6 dB at 3kHz. I think the opamp filter might be swamping it out. Either that or I wired something incorrectly I that stage (wouldn't be the first time!).

Actually, I'm embarrassed to admit that I screwed up the opamp wiring initially. I couldn't get the gain stage to amplify the pre-EQ signal, and for the life of me I couldn't track down the problem. I ended up using the other half of the opamp to boost the signal first. When that part didn't work either, I realized that I had wired the ground resistors/capacitors of both opamp to their outputs instead of their inverting input by mistake (double d'oh!!). Instead of ripping everything out again and destroying the perfboard, I figured I would just live with the additional gain stage and add an extra Fetzer at the end to bump up the juice.

I wonder if the noise would be significantly less if I had set up the circuit as originally intended?

In --> Fetzer (1.4x gain) --> EQ1 --> gain/clipping (20 dB plus clipping gain) --> EQ2 --> clean gain (20 dB) --> Fetzer (1.4x gain) --> out

Or does that look like it would still sound like a bag of snakes?

[sajy_ho]

I think that would help; but using shielded wire, metal chassis and proper grounding will also help.

[PBE6]

Very true!