Fetzer valve article and dr.boogey

[siore]

Greetings!  It took me a couple of days, but I just digested the fetzer valve article on ROG.    Following the principles mentioned on there, it seems the dr.boogey should be done differently?  At least the more popular layouts on it?

1.  The 1uf source bypass caps.  Going by the 3/2 power rule, there should be no bypass caps to get the triode emulation.  Although it was mentioned in section "10:fetzer valve deluxe" that a small 'presence' cap could be added, the range of values are between 68-100nf!  What's going on here?  The 1uf cap seems to be blindly copied from the dual recto schematic (cathode bypass caps).

2.  Source resistor.  If the bypass caps are omitted, then the source resistors must be chosen, not taken in stone as 1.8k.  1.8k was based on a j201 with Vp=.75 and Idss=.35ma, but based on the actual measurements on fifteen j201's in my stash, there's a lot of variation on Vp's and Idss'.

3.  Miller caps.  I've seen schematics where the miller caps are shown gate to source, but the article recommends gate to ground.  When did people start doing gate-to source miller caps?  Why?

4.  The C21 input cap.  What does it do?  I don't see it on the dual recto schematic, and find no mention of adding it on the fetzer article?

Anyway, I should've started on my dr.boogie, using the gaussmarkov layout, but I decided to look into FETs info first, and that led me to the fetzer valve article.  Guess at least, I'll be having mods ready right?     My intention here is not to question how things are  done, so I apologize in advance if some guys get offended when they have been following the layouts.  Rather, i hopefully start a discussion on the items I've mentioned.  From the tech guys who know what's going on circuit-wise, and the guys who've breadboarded these mods (maybe the schematic is best, they've found out?).  It seems to me that quite a bit of research was done by mr.dimitri danyuk, who's work is apparently a huge chunk of that ROG article.

Thanks in advance for any input you may have.

[stm]

You present some interesting questions which I'll try to address based on my practical experience, therefore, I'll present just my personal view on these matters, not the absolute truth, so I don't expect to step on anyone's toes.  There is a difference between "trying to emulate tube behavior accurately" v/s "trying to get good tone with whatever works".  Some people get passionate over the first statement, others are pragmatic and go along with the second--this is a personal choice, of course.

1) With respect to the three-halves power Fetzer Valve: this is a great sounding circuit by itself, however its use as an amp stage replacement is not written in the stone.  Personally, I've found that for very low to medium overdrive I prefer the 3/2 fetzer valve over the square law stage that you get when you bypass the source resistor with a large cap, as I tend to dislike the harmonic structure that you get with the latter.  Nevertheless, I've also found (to my surprise) that high gain circuits do sound well with the added gain (and harder clipping and stronger 2nd harmonic) that you get by the use of a source bypass capacitor, so the use of a source capacitor it is not to be dismissed at all!  In addition, in case of the Dr. Boogey, the source bypass capacitors are mandatory for reducing the low end flabbiness, so they are key to the proper sound of the circuit.

2) The 1k8 resistor in the original amp is a more than adequate value for the source of a self-biased J201 stage.  Even though these resistors might have been blindly lifted from the amp's schematic, this can be seen as a happy coincidence that works quite good as it is, and at the same time allows using the same values for the bass-thinning caps as in the original amp.

3) With respect to the miller capacitance, it really comes from plate-to-grid (or drain-to-gate) capacitance multiplied by the voltage gain of the stage, plus the grid-to-cathode (or gate-to-source) parasitic capacitance of the device.  This may sound complex, however a similar effect is achieved by adding an equivalent capacitor between gate and GND--this is the so called "miller cap" in these circuits.  Notice that the Dr. Boogey has the miller caps mounted between gate-and-source, however considering that the source is already bypassed to GND by a 1uF capacitor (which has an impedance around 4500 times lower than that of the 220pF caps), it can be considered that at high frequencies the so called miller caps are in fact shorted to GND.

4) The input capacitor is just there to prevent upsetting the biasing of the stage when there is some DC leakage from another pedal placed in the chain.  Its value is high enough not to reduce low frequency response.  You may omit it if you like, but personally I'd keep it just in case.

5) I won't go too deep with regards to the third amp stage (the one with the 39k cathode resistor and the 1n plate capacitor).  It is enough to say that the JFET implementation of this stage is far from replicating the original amp stage, nevertheless the Dr. Boogey still sounds great.

In summary, I suggest you build the Dr. Boogey stock using a proven layout like GM's with short pot wires in order to avoid oscillation issues as this is a high gain pedal.  After that, you may try some mods and see how it goes.

[siore]

Sebastian, thank you very very much for the detailed reply!  Exactly what I was hoping to learn.  I was afraid Im gonna have to breadboard the changes, but you just shed some valuable light on all items I was wondering about.     I have further questions on the source resistor.  For my J201's, I'm getting 1.3k for a group's Rs value (I know they are approximations and should be trimmed further though, and Vs/Vd is more important), but I take it the 1.8k is also there to match some cap values?  Do I benefit at all from nit-picking this source resistor value, when 1.8k already gets me in the ballpark?  On the other items, those are all clear to me now, especially on the miller caps.  Thanks again!

[stm]

I don't see a problem with using 1k8 instead of 1k3.  My guess is that it won't make much of an audible difference (provided you keep the cap resistor-to-cap ratio unchanged.  I've read some people have reduced one or more of the 1u caps (don't recall exact value, maybe it was 0.68uF).  So, if you go with the 1k3 resistor and leave the 1uF cap unchanged you are attaining a similar effect, so this could be advantageous.  You'll need to listen to the changes to see what suits you best.

Unfortunately when you are optimizing a circuit at this level of detail it is not possible to determine best values without listening.  It will also depend on your guitar, amp, personal taste, and also upon the drain voltage at which you decide to bias each stage.  Look for a thread named "ultimate Dr. Boogey".  It contained links to many related threads.  Anyway, my advice is before you start spending time thinking about how to tweak the circuit, first build it stock, then determine what you'd like to change and proceed accordingly.

[siore]

Agreed.  I can always experiment later on with it, especially now you mention that it's a similar effect to lowering the bypass caps value.  That one I've read about (dschwartz on his dual rectal thread I think).     Was just worried changing source resistors would mess up the drain bias if I then already removed the trimmer for fixed resistors, so I asked this one before the build.  Anyway, that's about all I need for now, it's time to take out the hot iron.