Couple of FF question regards RF filtering and pulldown resistors.

[carboncomp]

Hello, 2 questions for the price of 1.

Is there anything inherently bad about adding pulldown resistors to the Fuzz Face?

What's the best approach to filtering out potential RF/HF noise, a small cap on Q1 and Q2 C-E?

[antonis]

Is there anything inherently bad about adding pulldown resistors to the Fuzz Face?

Neither inherently nor acquired..

What's the best approach to filtering out potential RF/HF noise, a small cap on Q1 and Q2 C-E?

A 1nF (or so) cap between Q1 B-E and a 100nF ceramic shunt cap on power supply..

[carboncomp]

Neither inherently nor acquired..

Thanks Antonis. So, neither bad nor nessaserry.

A 1nF (or so) cap between Q1 B-E and a 100nF ceramic shunt cap on power supply..

So I can omit the 100nf if I already have a power filter going into my voltage inverter like this




So, just need a 1nf on B-E of Q1 like, think I'm getting confused on PNP and NPN....so on a PNP build I do like Cx here?

 

[antonis]

Don't bother about p-n-p or n-p-n 'cause RF rectification doesn't bother either..
(it always happens across B-E junction..)

Although attached schematic is tremendously small ( ), Cx is a Miller (feedback) cap which prevents Q1 & Q2 high pitch oscillations..
(it has nothing to do with "noise" elimination..)

[ElectricDruid]

Making big images display a bit smaller (ok, a lot smaller!) isn't too difficult. It looks like this:

Code Select Expand

[img width=width in here]image address in here[/img]

Just stick the img tag in as normal, then add in the width attribute to get it to something reasonable.

[carboncomp]

Don't bother about p-n-p or n-p-n 'cause RF rectification doesn't bother either..
(it always happens across B-E junction..)

:icon_lol:OK, that joke just lit a lightbulb in my head, thank you!

Although attached schematic is tremendously small ( ), Cx is a Miller (feedback) cap which prevents Q1 & Q2 high pitch oscillations..

Sorry, thought I would save anyone having to go get their reading glasses!

So for RF I would slap a cap here (marked RF)



And for oscillations, I would use Miller.....so this begs the question, why not both?

[carboncomp]

Making big images display a bit smaller (ok, a lot smaller!) isn't too difficult.

Thank you!

[antonis]

So for RF I would slap a cap here (marked RF)
And for oscillations, I would use Miller.....so this begs the question, why not both?

Is this some kind of rhetorical question..??

[carboncomp]

So for RF I would slap a cap here (marked RF)
And for oscillations, I would use Miller.....so this begs the question, why not both?

Is this some kind of rhetorical question..??

I'm not sure. To be rhetorical I think you need to know something, so that rules me out! 

I'm guessing from the little reading I have done on the topic (having only known about it for an hour  ) is that the Miller cap will reduce the gain as the frequencies go up, so how useful is that really going to be in a low-gain fuzz faces using PNP transistor, apposed to an NPN high gain transistor that is actually going to be able to produce those high frequencies? (that is a question as I have no clue  )

Also, guess blocking RF is more of a priority but unsure if there is a 'one size fit all' approach, the internet is full of contradicting advice, and guess if there was an excepted fix it would be on all the clones?!

[Dormammu]

Installing a 1-10k resistor in series with the input cap will often eliminate HF hiss.
Sometimes this is also required for the output cap.

[antonis]

I'm guessing from the little reading I have done on the topic (having only known about it for an hour  ) is that the Miller cap will reduce the gain as the frequencies go up, so how useful is that really going to be in a low-gain fuzz faces using PNP transistor, apposed to an NPN high gain transistor that is actually going to be able to produce those high frequencies? (that is a question as I have no clue  )

Hmmmm..
(An ambigous answer to a rhetorical question..)

Miller caps reduce stage gain, as the frequencies go up, which gain is mainly dependent on particular configuration..
i.e. a Miller cap in a BJT CE amp of given (high) gain has the same effect regardless of particular BJT h_FE..

P.S.
I presume this (or similar) kind of disscusion is more creative than a "yes/do" or "no/don't" alike..

[antonis]

Installing a 1-10k resistor in series with the input cap will often eliminate HF hiss.

Excuse me but No.!!
(it's for oscillation prevention..)

[carboncomp]

P.S.
I presume this (or similar) kind of disscusion is more creative than a "yes/do" or "no/don't" alike..

"yes/do" or "no/don't" only teaches you to follow instructions, not understand what you are doing or why!

[Dormammu]

Installing a 1-10k resistor in series with the input cap will often eliminate HF hiss.

Excuse me but No.!!
(it's for oscillation prevention..)

IDK about your experience, but my experience is — yes. Installing such resistors often saves me a lot of problems with different noises.
No Excuses .!!   

[carboncomp]

Miller caps reduce stage gain, as the frequencies go up, which gain is mainly dependent on particular configuration..
i.e. a Miller cap in a BJT CE amp of given (high) gain has the same effect regardless of particular BJT h_FE..

OK, this bit is confusing me a little, stage gain as in the ratio of the output divided by the input and a function of amplifier typology itself in this fuzz face example, a Common Emitter PNP amplifier opposed to the individual BJT h_FE in the circuit?

Is what we are doing by adding the miller cap akin to attenuating frequency range amplified by making a sort of sudo RC filter using the input impedance as the R and the miller cap as C meaning we filter out the high-frequency oscillation? 

if so (big assumption as I am sure I got most of that wrong) why is it not commonplace, think the only place I've seen this used of a fuzz face typology is pedals like the DAM meathead, and the Colorsound One knob fuzz (both using high gain transistors, so maybe that's why I made the assumption about the individual BJT h_FE being relevant, when in fact its just smoothing out the fuzz regardless of if its PNP/NPN, high h_FE, low h_FE?

That's a lot of question marks for more of a stream-of-consciousness diatribe on how little I actually understand.   

[antonis]

NFB (Miller) cap is widely used in high gain circuits, despite the origin of the gain..
(look at all those pF range caps between op-amps output and inverting input..)

Although your approach about RC filter is quite right (as long as C absolute value is multiplied by stage gain plus unity) it should be better to face it as a frequency dependent variable resistor inside NFB loop..
(like the one implemented for C-B feedback bias configuration)

What I'm trying to tell you is the need for Miller cap is irelevant to particular device gain..
(as logn as, of course, that particular device gain doesn't involve in overal stage gain..)

[ElectricDruid]

Is what we are doing by adding the miller cap akin to attenuating frequency range amplified by making a sort of sudo RC filter using the input impedance as the R and the miller cap as C meaning we filter out the high-frequency oscillation? 

In general, you want to only apply gain to the bandwidth of interest. If we're talking about audio, we don't want to amplify any subsonic thumps and bumps, and we don't want to amplify any ultrasonics interference or RF noise or whatever. We want our gain limited to a range somewhere between 20Hz and 20KHz (and we can choose the exact limits based on what we intend and what the job is). Well-designed circuits will include "extra" components that modify the frequency response like this and limit the gain for anything outside our required bandwidth. This helps keep noise down and makes the circuit more reliable. This is the same for op-amps, transistors, whatever.