Fuzz Face and Rangemaster placement in the chain

[DougH]

I thought of an idea I may work on later this year- Get a couple chip amp kits from qkits.com and build up a stereo SS power amp. Then build up a stereo speaker cabinet- one cabinet with a speaker per side with a corner between them (90 deg apart). Someone mentioned doing this last year and IIRC it dispersed the stereo sound real well. For low volume and/or ease of transporting this could be a ported cab with a couple 8" speakers. Then run the tube amp (or amp and booster, RM, whatever "pre" amp pedals you want, etc)  with the attenuator as the dummy load & run the line-out through the stereo fx and out to the stereo SS amp/cab. Then come up with some kind of box, rack, whatever to put the tube amp, SS amp, and attenuator in for easy setup...

[frank_p]

That's a good idea...
But for me, if I enter an other cab. or combo in the house, I am gonna get killed. 
I can only use what is already there. 
Unless I sell or...  hmmm...  convert something.
I have two 8in. alnico speaker with closed back aluminium army cabs.
One speaker is broken.  I'll have to find something similar to the other one.

Doug do you think that a guitar could sound good in a stereo 8in. speaker with this attenuator setup ?   
I might need some kind of bass booster in the SS amps (?)

[DougH]

Well, I'm getting ready to build a simple ported 8" cabinet to see how it sounds. I experimented with the speaker in its original cab closed back and then open back with a cardboard box behind to get different sounds and I think I can get a decent low-volume sound out of it. So after that I'll think about the stereo idea some more.

[frank_p]

Cool !

[DougH]

Just to follow this up- I tried the "wet-wet" setup (using the attenuator as a load only and running the line out through the stereo fx buss to two amps) last night. It sounded real nice too. For stereo fx like the Tonecores, that do more than just divide the signal into wet/dry outputs, this works real well and allows the fx to do the "stereo imaging".

If I had a third speaker cab I would connect it to the attenuator and try "wet-dry-wet" too.

All these techniques are pretty hardware-intensive. I think it would be a nightmare to drag all this stuff around have to set it up and break it down. But for home/studio it works well and sure sounds great. And a side benefit is you really have control of your volume level. You can really get a "big" stereo sound at a reasonable volume level for "at home" playing.

[frank_p]

For me that drove (is that the right word) me to question myself about what's going on before and after the attenuator (and by extension, in the amp, but I don't have a scope for high voltage...).  I think I will do some scoping this week and compare the wave shapes before and after the attenuator, and that for different amps.  Playing in different output cabs (and SS amps) made me realise how raspy and fizzy some of my tube amps are sounding when listened at low volume with the attenuator.  I suspect that many of my tube amps have more of a rather "hard clipping" behavior than I think.  So I wonder if it's coming out of the amp or it is due to the attenuator.  It might also be the way that the ear (or brain) perceive distortion at different volumes.  Anyway, the question is that I am now in doubt in my belief that tube amps always have a softer clipping than the SS apparatus.  Perhaps what is going out of some of my amps is more clipped or "squared" than I would have guessed at a first tought.  So by cutting some hights on the output SS amp, I am just, perhaps putting together the same idea that what is in the tubescreamer (?). 

The THD hot plate is not a purely resistive device  as opposed to the Z airbrake.  I opened it a wile back and there are some inductances in there, I don't know if it's for filering or impedance matching or simulation of the voice coil.  I also remember that the Z Airbrake got better reviews than the Hot-Plate.  So I wonder if the complexity of it's circuit really bringed something to it in comparison with other attenuators that are less complex.

Now I am in front of so many variables that can affect why the distorted tone do sound a bit raspy at lower volume.  I really wonder if the wave shapes are similar out of the tube amp and out of the attenuator.  If so, my religious belief that tube amps all have smooth clipping in the power amp will come to an end.

Now, if the waves are really more clipped what would it be due to (?) : "bad" design of the power amp in relation to the power supply or the output transformer ?  Perhaps some of my amps are not well biased ?  I don't know for the moment.  But it's a moment of doubt.

If I wanted to do a frequency and amplitude sweep on my attenuator, in the goal of scoping what's going on out of there, would I have to drive it ?   If so, what would be a good method ?  A SS amp, a power opamp or a power transistor ?

I think I could go in endless experimentation with that stuff...
There are too much things I don't understand for now.

[JDoyle]

The gain of the Fuzz Face is roughly the 100k feedback resistor divided by the input impedence of the source signal (your guitar or the preceeding effect).

Because the feedback resistor 'starts' at the emitter of Q2, it is analogous to placing it from the collector to base of Q1, the only difference being the Vbe drop of Q2.

It may be an interesting experiment to install a DPDT switch that allows one to switch in/out a 50k log or 20k-25K linear pot, that way if you are going direct from the guitar you can bypass the pot, but if you are connecting it after an effect with a buffered output you can switch the pot in series with the FF input and manipulate the gain by raising the signal's input impedence to the base of Q1.

Regards,

Jay Doyle

[DougH]

Frank, from my experience a lot of it comes down to how the ear-brain perceives freq response at lower volume (less bass). More bass in the sound tends to warm it up and mask some of the highs, in my perception of it.

I did a comparison once with an attenuator using my mass motor vs. a purely resistive load. The mass load had more high-freq response, kind of like a speaker. The resistive load sounded rolled-off and more "middy". Maybe people like the high freq  rolloff when they turn it down and lose the perception of bass too in the Airbrake- maybe that "equalizes" it in their mind. Some of the trainwreck amp clips sound kind of spikey and raspy to me. Maybe the rolloff of the purely resistive load in the attenuator helps that, I don't know. That's just speculation on my part.

I agree that EQ-ing for less treble on the "slave" amps in the wet-dry or wet-wet configurations sound better. But then, you don't need "more EQ-ing" since you are already doing that with the amp that is being attenuated. Otherwise you are exaggerating the EQ (2 EQ's in series). I'm guessing that's why so many just use SS power amps as "slaves" in these kinds of setups- flat EQ and just there to drive the speakers. I'm speculating that may be a better solution than using a guitar amp. But guitar amps seem to work too, as long as they have enough headroom and you set the EQ fairly flat. I did "flatten" the EQ of my SS amp before I quit last night and noticed the sound vastly improved and smoothed out. I'm going to experiment with this some more.

[frank_p]

Frank, from my experience a lot of it comes down to how the ear-brain perceives freq response at lower volume (less bass). More bass in the sound tends to warm it up and mask some of the highs, in my perception of it.

I did a comparison once with an attenuator using my mass motor vs. a purely resistive load. The mass load had more high-freq response, kind of like a speaker. The resistive load sounded rolled-off and more "middy". Maybe people like the high freq  rolloff when they turn it down and lose the perception of bass too in the Airbrake- maybe that "equalizes" it in their mind. Some of the trainwreck amp clips sound kind of spikey and raspy to me. Maybe the rolloff of the purely resistive load in the attenuator helps that, I don't know. That's just speculation on my part.

Ok Thanks for your input.

I agree that EQ-ing for less treble on the "slave" amps in the wet-dry or wet-wet configurations sound better. But then, you don't need "more EQ-ing" since you are already doing that with the amp that is being attenuated. Otherwise you are exaggerating the EQ (2 EQ's in series). I'm guessing that's why so many just use SS power amps as "slaves" in these kinds of setups- flat EQ and just there to drive the speakers. I'm speculating that may be a better solution than using a guitar amp. But guitar amps seem to work too, as long as they have enough headroom and you set the EQ fairly flat. I did "flatten" the EQ of my SS amp before I quit last night and noticed the sound vastly improved and smoothed out. I'm going to experiment with this some more.

The slave amps I used this week were mono mixing-PA amps and my keyboard amp with different sort of cabs.  I don't have any SS guitar amps and never had one (except fof my Rolland mini cube).

The first SS combo I bought was a Keyboard amp (Peavey KB 60) because at the time I was playing a diatonic nylon strung Paraguayan harp (36 strings).  So the salesman at the music store sold me that: it was perhaps a better idea to mike the harp in a keyboard amp.
Anyway, I never succeded to amplify the harp convinently.  The damn harp is a feedback monster;  with 36 strings and a spurce top that is so thin that you can see the sunlight when looking in the sound holes, it's really difficult to play in an electric band with it.  I have a plan to build an electric harp "in the future".

The PA-mixer with graphic EQ in it is great for experimentation with the attenuator.

Thanks Doug for all your time explaining me some stuff.

The gain of the Fuzz Face is roughly the 100k feedback resistor divided by the input impedence of the source signal (your guitar or the preceeding effect).

Because the feedback resistor 'starts' at the emitter of Q2, it is analogous to placing it from the collector to base of Q1, the only difference being the Vbe drop of Q2.

It may be an interesting experiment to install a DPDT switch that allows one to switch in/out a 50k log or 20k-25K linear pot, that way if you are going direct from the guitar you can bypass the pot, but if you are connecting it after an effect with a buffered output you can switch the pot in series with the FF input and manipulate the gain by raising the signal's input impedence to the base of Q1.
Regards,
Jay Doyle

And for the Fuzz Face, it's opened on the operation table.  Thanks also Jay. 

My experiments continues...

[DougH]

Thanks Doug for all your time explaining me some stuff.

Hey, no problem Frank. A lot of it is just speculation so take my advice for what it's worth (i.e. what you paid for it  ).

Just to clarify one point- I don't think there's anything special about SS vs tube on the slave side, and I believe the "flat EQ" is probably a good idea either way.

[Ben N]

Just to clarify one point- I don't think there's anything special about SS vs tube on the slave side, and I believe the "flat EQ" is probably a good idea either way.

Two qualifiers:
1) There is nothing special about tubes assuming you are running pretty much in the linear region, either because your SS slave has plenty of headroom (like 100 watt hifi amps that are functionaly equivalent to 2x6L6, because God forbid the SS power amp ever hit saturation), or because the dynamic range of the signal going in is limited. I think it is a good idea to include some kind of limiter (perhaps switchable) at the input of the slave--maybe something like the old Vox SS amps? Personally, I think limiting (that tube amps do organically) is actually preferable to high headroom, since a lot of dynamic range is not necessarily a good thing in a performance (as opposed to a listening) situation.
2) Flat EQ--again, assuming either of two situations: (a) that your slave is driving a cab that rolls off frequencies pretty much the same as the master amp does or would (a la EVH's 80s setup, with the power amps driving 4x12s), or (b) if your cab is full-frequency, like a PA cab, that your line out has some filtering to roughly simulate a speaker cab. When I used the line out on my old Deluxe reverb, it was unusable into a PA, but worked ok into the fx return of a guitar amp. I guess the most versatile solution would be a full range cab with a switchable cab sim built into the amp.

[JDoyle]

...I think limiting (that tube amps do organically)...

Just to be nitpicky - all amps 'limit' as you are using it, it's just that tube amps do so in a much more pleasing manner than solid state. Plus, especially in the realm of hifi, the amount of negative feedback is the determining factor in the abruptness of the onset of the 'limiting'. Because the transfer curve of a tube is much more linear than either FETs or BJTs, less negative feedback is needed to correct the distortions created in the amplifying process. BUT there are other advantages to negative feedback, especially in the case of hifi, such as desensitizing the gain across the frequency spectrum, increasing the bandwidth and altering the input and output impedences up or down (which is what makes 'emitter/source degeneration' similar to, but in the end, different from, negative feedback; not that anyone cares).

And as an aside, placing a limiter of some sort in an amplifier restricts the available output power. Baxandall has a good article about this in Wireless World from 1980, I think...

Jay Doyle

[Ben N]

...I think limiting (that tube amps do organically)...

Just to be nitpicky - all amps 'limit' as you are using it, it's just that tube amps do so in a much more pleasing manner than solid state.

OK, Jay, you know that's what I meant, but you did put it more precisely.

And as an aside, placing a limiter of some sort in an amplifier restricts the available output power. Baxandall has a good article about this in Wireless World from 1980, I think...

Even if the limiter really just precedes the amp?

Apropos of this discussion, http://www.diystompboxes.com/smfforum/index.php?topic=73342.0;topicseen--the "passive speaker sims" discussed are passive filters used to tone down the line outs of various amps so they are usable. This simplest, of course, is just a RC.

[DougH]

For my purposes I'm not interested in purposely getting distortion in the slave units. I'm using guitar speakers so I'm not interested in the slaves doing EQ-ing either. I'm mainly looking for a lot of headroom and enough power to play "loud enough" while the power slaves stay clean. Basically I just want to amplify a facsimile of the loaded tube amp at the beginning of the chain (with/without additional fx processing), then let it get EQ-ed by the guitar speakers. So for me, I don't really care what the form of the slaves are as long as they meet those requirements. (In this instance I don't think limiting is necessary for SS.) I think an SS power unit is probably the easiest way of doing it esp if you want to make it a compact unit for transport. Other than that I don't think it makes much difference to me.

[frank_p]

2) Flat EQ--again, assuming either of two situations: (a) that your slave is driving a cab that rolls off frequencies pretty much the same as the master amp does or would (a la EVH's 80s setup, with the power amps driving 4x12s), or (b) if your cab is full-frequency, like a PA cab, that your line out has some filtering to roughly simulate a speaker cab. When I used the line out on my old Deluxe reverb, it was unusable into a PA, but worked ok into the fx return of a guitar amp. I guess the most versatile solution would be a full range cab with a switchable cab sim built into the amp.

I think that is why it sound good in the bass cab on my side.

BUT there are other advantages to negative feedback, especially in the case of hifi, such as desensitizing the gain across the frequency spectrum, increasing the bandwidth and altering the input and output impedences up or down (which is what makes 'emitter/source degeneration' similar to, but in the end, different from, negative feedback; not that anyone cares).

Ouch Jay...  Do you have some reading to suggest me.  With graphs and equations.   There is a lot of stock in that phrase...

For my purposes I'm not interested in purposely getting distortion in the slave units. I'm using guitar speakers so I'm not interested in the slaves doing EQ-ing either. I'm mainly looking for a lot of headroom and enough power to play "loud enough" while the power slaves stay clean. Basically I just want to amplify a facsimile of the loaded tube amp at the beginning of the chain (with/without additional fx processing), then let it get EQ-ed by the guitar speakers. So for me, I don't really care what the form of the slaves are as long as they meet those requirements. (In this instance I don't think limiting is necessary for SS.) I think an SS power unit is probably the easiest way of doing it esp if you want to make it a compact unit for transport. Other than that I don't think it makes much difference to me.

I think I didn't explain myself clearly.  I just said I don't have any SS GUITAR amps.  I used SS pa-mixer amp and SS keyboard amp (No SS guitar amp as slave).
I need to cut high in any case.  Tube guitar amps do it by themselves it seeems, not the keaboard amp (thats why the keyboard combo don't sound good).  I think that when everything is set "flat" on a SS guitar amp, it's not a flat response that you finally have (well, less than a keyboard amp).  That's perhaps why it sound better with a guitar amp: there is already some treble cut.

The best result I had was with the SS PA-mxer amp and bass cab (to cut high requencies at low volume). 

[DougH]

I think I didn't explain myself clearly.  I just said I don't have any SS GUITAR amps.  I used SS pa-mixer amp and SS keyboard amp (No SS guitar amp as slave).
I need to cut high in any case.  Tube guitar amps do it by themselves it seeems, not the keaboard amp (thats why the keyboard combo don't sound good).  I think that when everything is set "flat" on a SS guitar amp, it's not a flat response that you finally have (well, less than a keyboard amp).  That's perhaps why it sound better with a guitar amp: there is already some treble cut.

Frank, I think my comments were directed more at Ben and maybe Jay. I think the reason your pa and keyboard rigs need highs cut is probably due to the speakers. That's why for me I like to use guitar speakers as they already provide the rolloff.

[frank_p]

Exactly.  OK sorry, Doug.

For the Fuzz Face:

It may be an interesting experiment to install a DPDT switch that allows one to switch in/out a 50k log or 20k-25K linear pot, that way if you are going direct from the guitar you can bypass the pot, but if you are connecting it after an effect with a buffered output you can switch the pot in series with the FF input and manipulate the gain by raising the signal's input impedence to the base of Q1.
Jay Doyle

Just like late 60's Fuzz and with a switch, OK.  I don't know why I didn't remembed that.  RG explained it already in the technology of the fuzz face.  (the Fulltone 69 mod).

Making the front end like a current source instead of a voltage source.  This is perhaps a good project for measuring and doing some maths.

Like this (late 60's on GGG):
http://www.generalguitargadgets.com/pdf/ggg_ff5_sc_b69.pdf

I think I'll just build a new one.

[JDoyle]

Just like late 60's Fuzz and with a switch, OK.  I don't know why I didn't remembed that.  RG explained it already in the technology of the fuzz face.  (the Fulltone 69 mod).

Gotta admit I didn't remember R.G. talking about it in the Tech of... article, or that fuller had done it either. It honestly has been quite a while since I've looked at commercial schematics or read that article. I just know that the fuzz face topology is an old circuit, a really old circuit. Ghausi (I think I'm spelling it right) wrote a few articles delving into it back in the 60s; it was considered to be a 'wideband' amplifier (if you think about it, because the signal at the emitter of Q2 is almost exactly the same as the signal at the collector of Q1, the parasitic Miller capacitance of Q1 has the same signal appear at both ends of it, and therefore its effect is reduced considerably, extending the bandwidth). The only thing missing in the Fuzz Face schematic is an input series resistor, and that can be replaced with either the impedence from the preceding effect or that of your guitar.

Multi-transistor feedback threory can get dizzyingly heavy on the math (I'm an ignorant American, math is singular to me  ) but one of the better articles, if you are up for it, is here:

http://users.ece.gatech.edu/~mleach/ece3050/notes/feedback/fdbkamps.pdf

As for the other comment about you wanting me to post some stuff about my negative feedback comments earlier, I don't have anything I can find online about neg. fdbk vs. emitter degeneration. My statement stems from the definition of feedback, in that it does those things I've mentioned, where as emitter/source degeneration only reduces the gain and just because something exhibits one aspect of something else does not mean that it IS that something else. Even The Art of Electronics states that emitter degeneration is a 'form' of negative feedback (but then admittedly gets a bit foggy later on). Further, the basic definition of negative feedback requires a portion of the output signal to be sent to the input, and while the action within a transistor EFFECTS what happens at the input, the OUTPUT of the transistor, the portion that has 'gain', is NOT sent back to the input by adding a emitter/source resistor. Therefore, inserting an emitter/source resistor may reduce the gain, but any reduction in distortion is secondary to that effect; and all of the other postitive aspects of negative feedback are essentially missing. But don't worry, it's probably all semantics anyway. 

ANYWAY:

As for articles on distortion in power amplifiers, the following really deal with solid state HIFI amps, and amplifiers utilizing negative feedback, but are the best available online by far and should keep you or anyone busy for quite some time.

(Plus, lots of graphs, math, and experiments! )

Here is a great series of articles by Baxandall:

http://www.linearaudio.nl/Documents/Baxandall%20power%20amp%20design.pdf

And a long series from Douglas Self adapted from his book "Self on Audio":

http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=205207238&queryText=distortion+in+power+amplifiers
http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=205601405&queryText=distortion+in+power+amplifiers
http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=205801115&queryText=distortion+in+power+amplifiers
http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=205917273&queryText=distortion+in+power+amplifiers
http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=206100517&queryText=distortion+in+power+amplifiers
http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=206105223&queryText=distortion+in+power+amplifiers
http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=206503400&queryText=distortion+in+power+amplifiers
http://www.audiodesignline.com/showArticle.jhtml;jsessionid=QJI2ZCP0FRZ5KQSNDLPCKH0CJUNN2JVN?articleID=206800813&queryText=distortion+in+power+amplifiers

And here is one that (first) hit on the fact touched in each of the above, that while negative feedback is overall a good thing, if you are going to use it, a little bit is worse than a lot:

http://www.linearaudio.nl/Documents/feedback%20fallacy%20scroggie.pdf

Enjoy!

Jay Doyle

P.S. - Ben N.: you are right, if the limiting is at the beginning, it will only effect the dynamic range of the input signal, not the output ability of the amplifier itself. I apologize, I read your post incorrectly. - JD

[DougH]

Another thing worth mentioning, Jay, is when you mention negative feedback I think you are referring to global negative feedback. 

Yours in nitpickery,

Doug

[frank_p]

OK, guys...  I can't reply... For now...
I am reading Mr. Doyle's documentation.
And I am overwhelmed.

And I am REALLY happy with it.  There will be no negative feedback from my side.

Didn't find find Ghausi's articles on wideband amps on the net, but hey, I have already a lot of stuff to read.

Plus found more things on feedback in Ch. 8 of my Sedra Smith's book.
And:
http://en.wikipedia.org/wiki/Asymptotic_gain_model#Two-stage_transistor_amplifier
http://en.wikipedia.org/wiki/Negative_feedback_amplifier#Two-port_analysis_of_feedback

Sedra
Ch 8.2:
- Gain Desensitivity (desensivity factor)
- Banwith extension
- etc.

Multi-transistor feedback threory can get dizzyingly heavy on the math (I'm an ignorant American, math is singular to me  Wink ) but one of the better articles, if you are up for it, is here:
http://users.ece.gatech.edu/~mleach/ece3050/notes/feedback/fdbkamps.pdf

I have some bases in maths, but I am rusted... And I know that even with knowledge of maths, those circuits can be very hard to analyse.
If you're an ignorant, you'll be happy to know that I live in an igloo and I eat beavers up in Canada.

Great !  There are even examples at the end of the paper. Plus: the link at the top:
http://users.ece.gatech.edu/~mleach/ece3050/notes/feedback/FBExamples.pdf

And more (to read till I die) at the same site:
http://users.ece.gatech.edu/~mleach/ece3050/

As for the other comment about you wanting me to post some stuff about my negative feedback comments earlier, I don't have anything I can find on-line about neg. fdbk vs. emitter degeneration.

I found nothing either in french on that subject.  I don't even know if I can translate it by "dégénerescence d'émetteur" or it's an other term or expression in french.  All google is giving me is psychology papers...

Even The Art of Electronics states that emitter degeneration is a 'form' of negative feedback (but then admittedly gets a bit foggy later on). Further, the basic definition of negative feedback requires a portion of the output signal to be sent to the input, and while the action within a transistor EFFECTS what happens at the input, the OUTPUT of the transistor, the portion that has 'gain', is NOT sent back to the input by adding a emitter/source resistor. Therefore, inserting an emitter/source resistor may reduce the gain, but any reduction in distortion is secondary to that effect; and all of the other postitive aspects of negative feedback are essentially missing.

I have the book, I have to search (more) in it...

But don't worry, it's probably all semantics anyway.

Yes, electrons are negative also.


I'll continue tomorrow, it's getting late, and it's been four hours I've been on you post Jay. 

It's really OVERLY GENEROUS : these papers are REALLY great !  I do not understand everything, but enough to understand their values.

I'll continue to read the papers on feedback in power amps tomorrow (and in the next weeks).
All this will be usefull.

Big thanks.

Hugo-Franc. P.

P.S.: Yes, V_be is not VBE...  But can I write V_be ?