FF transistor hfe?

[MarkDonMel]

Hey all,

Bought two sets of trans from small bear.  One set has gains of 101, and 166.  The others are 100, and 163.  Will there be much difference in sound?  

He was also gracious enough to provide the resistor values and fill my order with them exactly.

Instead of 33k, 8.2k, 470ohm, and 100k, his values were, 39k, 3.9k, 470ohm, and 150k, respectively.

I'm just curious if anyone who knows this circuit well, or just knows in general, what replacing the values back to orignial would do.  

And I'm not questioning steve's choice of values, I simply don't know any better, why were these values changed? I mean obviously for sound, but anyone care to elaborate?

Thanks much,  Matt.

[Peter Snowberg]

Hi Matt,

My guess is that there will be just about zero difference between them. For all real purposes they'll be identical.

Germanium transistors were highly variable in manufacture, so the different resistors are required to set them bias correctly. The standard procedure in making a fuzzface these days is to use pots instead of fixed resistors in some spots (the 33K & 8.2K) and then to adjust until the bias is proper. Buying the transistors from Steve comes with the added advantage of him doing the work for you to get the right values to start with. Germanium is so sensitive to heat that I would still use trimmers to fine tune things if you're a perfectionist, but his values should work fine. if you use the values you see on old schematics, you will have mis-biased transistors for sure and it will sound lousy, just like >95% of the production units apparently.

Oh, and the 150K for the 100K substitution will tweak the amount of feedback into a better place for the gains of your pairs. Mike Fuller posted his favorite gains a while back (with 100K in there). The gain values Steve sent you are a little bit higher and increasing that 100K will bring things back in line with a really optimum set of values. That circuit is all about balance.

You should be in great shape. Best of luck on your builds!

Take care,
-Peter

[MarkDonMel]

Right on Peter.  A lot of that makes sense to me now, at least the basics.

And yeah, it sounds great.  I mean I am really happy with it.  I was just curious about the why's, and you answered them for me.  Thanks.

I have enough parts for one more FF and I think this time I will try using trimpots and pots instead of the regular resistors.

Since I am a beginner, I am wondering if you or anyone has a site or info on which pots I would want to use to replace the 39k and 3.9k resistors with (33k and 8.2k originally).  And, would I replace the other two resistors, 470ohm and 150k, with trimpots?  Do I have to get specific trimpots for those, I mean will a tech at a retail store be able to know by the values of the resistors in the circuit?

One more thing, lol, would changing the value of the 150k to a higher or lower value muddy or clean up the sound, by decreasing or increasing it, respectively? I think I am starting to see what the parts do... I mean, that resistor is between the two transistors... so it will somehow effect the gain loop thing I have read about right?  Or, something like that lol...

Thanks agian peter, and all.  Matt.

[petemoore]

That's the feedback loop resistor.
 When I see that I know it's a FF.
 RG explains better lol but
 Basically the second transistor starts feeding the first transistor through that resistor. This causes what we call Fuzz. But certain harmonics get accentuated...and gain increases alot.
 Making that resistor causes the ckt to have more internal feedback...but that's controlled also by other things [guitar vol, ckts 1k gain knob etc].
 You can piggy back a 150k resistor parallel to the 150k [two 150s paralleled =75]  experimenting with larger resistors you can get closer to but under 150k.
 You could stick a 100k pot paralleled with a resistor [check the resistance range with a DMM and I would start with say 68k?]...the seriesed resistor and pot should go from 68k to 168k range. That way you could diddle and see exactly what it does.
 I like the extra pot on my Boutique FF though [far left of schematic]...
 The R value Steve prepared will probly get what you wamt/need with the gain knob and the guitar vol etc.

[javacody]

Pete,
   This thread brings up a few questions I have. Basically, you can replace the 8.2k,  33k, and 100k with trimmers (10k, 50k, and 100k in series with a 68k respectively) but how do you know where to start? Is there an order of tweakage here? Should you start with the values in the schem and start tweaking one resistor at a time? Which resistor would you start with?

Also, how does changing the fuzz pot to 2k change things?

[MarkDonMel]

yeah, thats what I meant  :wink:  lol

I am new to this, annd terms like "piggyback" and "parallel" are foreign to me.

I am going to assume, and you can correct me if you would...

Piggybacking-  I would just add another resistor connecting the leads to the exact same leads as the resistor already there.. so it would be like a dual resistor?  This would mean that instead of piggybacking two, say.. 100k resistors, I could use one 200K?  Knowing nothing about this, this is all I can fathom using general logic, with no real experience about it.  I am probably totally wrong.

Parallel- totally beyond me...  Head me towards a link?  Or some other info?  Or, if its easy enough and you have a minute, would you explain these to me?

I really, really, appreciate the help.

Thanks a lot,  Matt.

[javacody]

Mark:

Series, one lead of a resistor is connected to another lead of a resistor, looking like this:

A----====----BA-----====------B

With A connected to the circuit, then B connected into A of the second resistor, and B of the second resistor connected back to the circuit.

Parallel resistors would look like this:


A |-----====------| B
A |-----====------| B


with A connected to A and B connected to B.

To calculate the value of series resistors, you add the two, to calcuate the  value of parallel resistors, read the following resistors in parallel

[MarkDonMel]

Bitchen,  thanks javacody.

Piggyback is the same as series?

regardless, that helped A LOT.

thanks  

[idlefaction]

piggyback isn't a word commonly used in electronics, it's just the way he was describing it at the time to try and put it in laymans terms    so don't stress about it.

he was meaning piggyback to dcescribe how the resistors look when they're wired in *Parallel* in this case.

also, all the transistors i've ever got from small bear have had those resistor values    do some tweaking yourself, follow RG's FF-tweaking article if you like, and your ears will be the final test.

have fun, it's a wicked circuit  

[brett]

Maybe this will help too...
That 150k resistor (as you already worked out) completes a "feedback" loop in the circuit.  A "feedback loop" introduces an output signal to the input of a circuit so that it cancels out part or all of the input signal.  This is critical in fuzzface circuits (the circuit would not work properly without it).  Now, because the amount of cancelling that happens depends on the amount of current flowing through this resistor, a smaller value resistor (e.g. 100k) allows more cancelling than a large resistor (e.g. 150k).  So the output with your 150k resistor is greater than with the 100k resistor.  That sounds good, but it's not all milk and honey.  Much of the sound of a fuzzface comes from having strong feedback - that's why most people use a 100k feedback resistor, and some even use a 68k resistor.  Partly, the best resistor also depends on the amount of output (i.e. "gain") from the transistors.  At first I thought it seemed a little odd that you've been given a 150k feedback resistor when your transistors will produce a little more gain than the "usual" ones (Hfes around 70 and 120 are considered "ideal").  So if you had high gain transistors, why up the gain even more by reducing feedback with a 150k resistor?

But then I thought about the last FF clone that I built - it had transistors with Hfe=150 and 160, and it sounded great (with a 100k feedback resistor).  High gain (maybe a bit too much when turned all of the way up) but a really good fuzz all the same.  Think also of the silicon FFs - they used very high gain transistors (Hfe>200), and many players with a very good "ear" for tone (like Jimi Hendrix) used them.

Which leads me to conclude that there's considerable latitude in what makes a good fuzzface.  So there's bound to be some value in experimenting and tweaking those resistor and transistor values.

[MarkDonMel]

Great.  Thanks a lot, I know feel comfortable in understanding the basics of that.  I just read some really cool beginner info on a link that was provided above as well.  I think with my next one, I am going to try some mildly different resistor values in that feedback loop to hear what difference in sound there is.  This is so much fun.  I am very pleased with all of your willingness to help and share knowledge. This board is awesome.

Thanks once more,  Matt.  

[petemoore]

I think you got parallel
 ._______________.
 ._______________.
 And Series
                         .__________..___________.   [arts not that great but connecto dots.]
 Easuer for me to find the connection on the top of the board...in this case I would look for brown green yellow [gold] thingy on the board or the 150k resistor.
 start messin around with the bottom of the board...with all that sticky copper and solder there...re-do's there can be done but if they for any reason get messy theu're harder to clean up.
 Sometimes I just sacrifice the installed resistor and use the leads to wire another in it's place...gives me somthing to bend the new leads on,and I already know it connects where Im wanting. [watch your soldering heat/time here or you'll loosen the connections if the heat has time to sink to the boards bottom connections.
 I don't like redos on board bottoms cause it slways tends to want to shake when the solder melts or the part comes free, splattering solder
 top of the board I can clean that with nail in 1/2 second...bottom of the board is quite another matter.
 I use the term and it actually deserves to be heeded as a term 'piggybak' as most of us remem 'riding piggyback' as kids and the relationship of two bodies [2 human or 2 resistor] ones of the bodies legs supporting both, I think is a useful and appropriate term for paralleling resistors [caps etc?] on the top of the board.

[Peter Snowberg]

This is what I would suggest for trimmers....

100K --> Use a 22K resistor in series with a 250K linear pot

33K --> Use a 22K resistor in series with a 25K linear pot

8.2K --> Use a 2.2K resistor in series with a 10K linear pot

The series resistors will stop you from blowing the transistors if you turn the control too far.

Have fun.

-Peter

[bwanasonic]

I'd be interested in seeing some people's subjective terms for the various adjustments made by raising/lowering values in the FF circuit. For example, I have found I prefer around 6v at the collector of Q2 to the *reference* of 4.5v (I use a trim at the collector). Lower sounds *woolier* to me. What descriptive term would you call the feedback trim pot?  

Kerry M

[MarkDonMel]

lThanks pete and peter.

Piggyback is a good term for that, I think I understood you from the beginning.

And I am understanding more now this whole thing.

I have two quick questions.

1.  Will those pot and resistor values work with the new values I am using?  39k, 3.9k, and 150k?

2.  If  they go between Q1 and Q2, which goes first?

Just so I understand this,  the Base of Q1 is going to one of the leads of the 150k gain loop resistor, the other lead of that resistor (R3) is going to the emitter of Q2.  Since I don't really know how to map the current direction, does it matter which goes on which side?  ...i.e. Q1 base to resistor to pot to Q2 emitter?  Or should it be pot then resistor?  Does it matter?

In addition, what about for the other resistors?  does it matter which "side" the pot and resistor go on as long as one end is connected to the board where the one resistor lead goes and the other to the other spot on the board where the resistor goes, I mean in the parallel pot/resistor conduit or channel, or current line or whatever you call it?

Are there three leads on the pots?  if so is one a ground?  In that case it would go right to ground right?


Thanks, Matt.

[Peter Snowberg]

If you make up the combinations above and then put your meter across them, you can dial each one up to a "starting" value, replacing the fixed resistor. Mark the position with a pen or scratch so you know where 39K, 3.9K and 150K are in the rotation.

The placement of the parts doesn't matter. Since they're in series you add the values so it's like saying 1+2=3 and 2+1=3.

I hope this ASCII art does the trick:

Code Select Expand

-------/\/\/\/\/\/\--+---/\/\/\/\/\-------
            ^       |
            |       |
            +-------+
.    ___
.   /   \
.  /     \
. |   0   |
.  \ _ _ /
.   # # #
.   | | |    ___
.---+ +-+---(___)-----Just connect the fully clockwise end to the wiper and then to the resistor. The fully counter clockwise end goes to circuit point A and the other end of the resistor goes to circuit point B. No grounds connected at all. This configuration is actually called a rheostat rather than a potentiometer.

The wiper is connected to the CW end for "safety". If the pot doesn't have wiper contact for some reason, the resistance just jumps up the max value rather than to an open circuit.

Take care,
-Peter

[petemoore]

It doesn't matter.
 Whatever seriesed resistances you have [R+R or R+pot etc.] will measure the same from one end to the other, or the other end to the one [either direction].
 I haven't heard anything that would mess with this logic...yet.
 So as far as I know and can tell it makes no difference whatsoever how the seriesed resistors are placed between two points.

[javacody]

Peter, that is an interesting breast.... er pot there in that ascii art.   :wink:

[MarkDonMel]

Right on!

Thank you so much, I totally understand how to do this now and feel comfortable modding my FF now.  Thanks!

I am guessing that is the top view of the breast, right?  Or I mean "full frontal view"? lol

Those were very easy instructions to understand, thanks peter and pete.

Matt.

[javacody]

Matt,
   Don't quote me on this, but it seems like usually when people draw pots. they draw them looking from the bottom.  Some pots have a nipple... er, indent on the bottom like that.