Issues with a picky fuzz circuit

[stonerbox]

Here is a little fuzz I have been playing around with. Built one and it worked perfectly. I am on the second build now, and it refused to act or sound the way the first one did. The issue..? Q3 HAS to be this BC549C (can't tell what manufacturer CLOSE UP PICTURE). I ordered BC549CTA (from Mouser) CLOSE UP PICTURE and whenever use one of these, Q2 fizzles out and the volume becomes weak and the character of the fuzz is completely lost. Could anyone educate me on why this would occur?

[Rob Strand]

Measuring the DC bias voltage on Q3 in each case would be useful.

The other thing that stands out is R5 is quite a high value and R6 is a low value.  If Q3 doesn't have enough gain there may be a very fine balance of getting Q3 to saturate if Q3's gain isn't high enough; or visa versa.  In this case you might have to adjust R5 to get Q3 to saturate the "right" way.

[stonerbox]

I will get you the DC bias tomorrow.


One more clue, when the BC549CTA is in Q3, Q2 starts low in volume but the signal drops very quickly after the first strum, like instantly, and so does it's Collector voltage.

[R.G.]

Here is a little fuzz I have been playing around with. Built one and it worked perfectly. I am on the second build now, and it refused to act or sound the way the first one did. The issue..? Q3 HAS to be this :BC549C

At a general high level, any time a device absolutely HAS to be something or other, the design is critically dependent on some feature/parameter of the part - and as such, it's not reproduceable (or manufacturable!) without strenuous identification of what the dependency is and careful testing/matching/selection of parts.
Can you  tell us more about how you know the device has to be the BC549C? Is it this fizzing you speak of?

[...] Q2 fizzles out and the volume becomes weak and the character of the fuzz is completely lost. Could anyone educate me on why this would occur?

Got a scope? Could be parasitic oscillation above audio. "Fizzles out and volume becomes weak" is one symptom of self oscillation.



[/quote]

[Rob Strand]

One more clue, when the BC549CTA is in Q3, Q2 starts low in volume but the signal drops very quickly after the first strum, like instantly, and so does it's Collector voltage

If you are getting blocking then sometimes that can be fixed by adding resistors in series with the coupling capacitors.  Also, it's work playing with the value of the coupling capacitor.

You could be running into a very fine issue where the swing of the Bazz-fuzz type stage is right on the limit of causing blocking and maybe the differences in voltages transistor's VBE and diode drops are too high or too low.   If that's the case the circuit isn't really a practical one and it would be best to try to bend the circuit away from that type of dependency.

[stonerbox]

Measuring the DC bias voltage on Q3 in each case would be useful.

The other thing that stands out is R5 is quite a high value and R6 is a low value.  If Q3 doesn't have enough gain there may be a very fine balance of getting Q3 to saturate if Q3's gain isn't high enough; or visa versa.  In this case you might have to adjust R5 to get Q3 to saturate the "right" way.

edit: problem solved, check my post at the bottom.

Tests are in!

Interesting results, since the two variants of BC549C should have the same specs. I guess factories do not make the exact same product, no surprise there. And under extreme cases like this, it shines through.


With Musikding's BC549C in Q3

Q3
C 8.95
B Idle 0.534 Playing: 0.300
E 0


Q2 2N5088
C Idle: 0.671  Playing: 0.299-0.310
B Idle: 0.509 Playing: 0.323-0.400
E 0


With Mouser's Onsemi BC549C in Q3

Q3
C 8.29
B Idle 0.567 Playing: 0.567
E 0

Q2 2N5088
C Idle: 0.668  Playing: 0.040 (!)
B Idle: 0.494 Playing: 0.128
E 0


With 2N3904 in Q3, the thing got loud as hell with this one, as the numbers downbelow allude to.

Q3
C 5.10
B Idle 0.658 Playing: 0.500
E 0

Q2 2N5088
C Idle: 0.667  Playing: 0.420-ish
B Idle: 0.498 Playing: 0.388-0.400
E 0

[stonerbox]

Here is a little fuzz I have been playing around with. Built one and it worked perfectly. I am on the second build now, and it refused to act or sound the way the first one did. The issue..? Q3 HAS to be this :BC549C

At a general high level, any time a device absolutely HAS to be something or other, the design is critically dependent on some feature/parameter of the part - and as such, it's not reproduceable (or manufacturable!) without strenuous identification of what the dependency is and careful testing/matching/selection of parts.
Can you  tell us more about how you know the device has to be the BC549C? Is it this fizzing you speak of?

Got a scope? Could be parasitic oscillation above audio. "Fizzles out and volume becomes weak" is one symptom of self oscillation.

Sorry, poor wording on my part. What I meant is, it drops in volume (mainly) and character changes. Fizzles out , when I stop and think about it, may be more about loss of bass content and a thin and "spitty" sound, and that is not it. Sorry.

[stonerbox]

Problem solved: In my original build, I had put my old BC549C in backwards, and they functioned. My fresh Onsemi's BC459C, which also has reversed Collector and Emitter pinout, did absolutely not. I am still curious as to why.

[Rob Strand]

Problem solved: In my original build, I had put my old BC549C in backwards, and they functioned. My fresh Onsemi's BC459C, which also has reversed Collector and Emitter pinout, did absolutely not. I am still curious as to why.

I'm not sure what's going on there.

European transistors like the BC549 etc tend to have pinouts reversed from the US transistors like the 2N3904. That's fairly well known and common.

Onsemi/Motorola transistors mostly use the 2N3904 type pinout.  However, I don't think they produced a BC549 until they bought out Fairchild; I'd have to dig through my old databooks.   Fairchild did produce some BCxxx parts.

From the Fairchild BC549 datasheet, the pinout looks like the normal European pinout for these transistors,
https://www.mouser.com/datasheet/2/149/BC547-190204.pdf

The actual onsemi BC549 datasheet (2001), also follows European pinout.   
https://www.mouser.com/datasheet/2/308/1/BC549B_D-2310299.pdf

The format of the datasheet is Onsemi/Motorola so maybe they did make BC549C's at some point.

If you have BC549's with 2N3904 pinouts then maybe they are fakes/remarked.
FWIW, transistor do operate with C and E swapped.   They have much lower hFE.     Also the voltage on the emitter (collector on normal ckt) cannot be much over 10V otherwise the reversed BE junction will breakdown.

[duck_arse]

I'm just guessing, but I think OP might mean the BC pinout is reverse to the 2N pinout, so BC needs to go "the other way" with respect to his "native" parts. I did wonder about pinouts when I read his first post

[stonerbox]

Problem solved: In my original build, I had put my old BC549C in backwards, and they functioned. My fresh Onsemi's BC459C, which also has reversed Collector and Emitter pinout, did absolutely not. I am still curious as to why.

I'm not sure what's going on there.

European transistors like the BC549 etc tend to have pinouts reversed from the US transistors like the 2N3904. That's fairly well known and common.

Onsemi/Motorola transistors mostly use the 2N3904 type pinout.  However, I don't think they produced a BC549 until they bought out Fairchild; I'd have to dig through my old databooks.  Fairchild did produce some BCxxx parts.

From the Fairchild BC549 datasheet, the pinout looks like the normal European pinout for these transistors,
https://www.mouser.com/datasheet/2/149/BC547-190204.pdf

The actual onsemi BC549 datasheet (2001), also follows European pinout. 
https://www.mouser.com/datasheet/2/308/1/BC549B_D-2310299.pdf

The format of the datasheet is Onsemi/Motorola so maybe they did make BC549C's at some point.

If you have BC549's with 2N3904 pinouts then maybe they are fakes/remarked.
FWIW, transistor do operate with C and E swapped.  They have much lower hFE.    Also the voltage on the emitter (collector on normal ckt) cannot be much over 10V otherwise the reversed BE junction will breakdown.

Thank you for digging into this and doing some research!

Aha, so that is what it is called! Then, I am used to American pinouts? Always thought of that as "standard pinout".

Code Select Expand

Is this called American Pinout?
______________
|      ____    \
|     |          \
|     |           \
|     |            \
|     |____         |
|     ____           |
|     |   |          |
|     |___|          |
|     |   |          |
|     |___|          |
|     _____          |
|     |             |
|     |____        /
|     |           /
|     |____      /
|______________/

The interesting part is that these BC549C works pretty okay backwards, though, lower  output  (like you pointed out) and forwards (produces proper gain) but the Onsemi's BC549C are nearly unfuctional when put in backwards.

[R.G.]

Most bipolar transistors will work with collector and emitter swapped, at least a little.
In the first bipolars, collector and emitter junctions were more or less identical, so swapping the two made no difference, and the frequency response and gain were both equally poor either way. Better semiconductor physics let to making the collector-base junction be much higher reverse voltage than the emitter-base, and the base region much, much thinner, so as they got better for normal use, they got worse for reverse use.
"Beta" for current gain is called that because reversed-mode current gain is called "alpha".
Reversed mode is still used in special low voltage situations for analog switching, or was until the last decade or two, when JFETs and MOSFETs got cheaper and more available. Reversed mode has a lower "on-resistance" if you're analog switching.
Most modern bipolars have a reverse voltage spec of 5 to 7 volts, so they might, possibly work in a 9V circuit that kept the voltage seen by the emitter-acting-as-collector down. This would be very dependent on the actual detailed diffusion process the manufacturer was using the day that transistor was made.

[PRR]

"Beta" for current gain is called that because reversed-mode current gain is called "alpha".

You may have forgotten cuz you were so young? 

Early theory focused on Common-Base. Current gain is usually less than one, and is Ic/Ie. (Some processes would get >1 current gain at higher Vc.) Note that Ib is not on the table.

Common-emitter turned out to be more useful especially for more modern processes. Beta is Ic/Ib. (By implication, if beta is large, then Ic is near same as Ie, our modern assumption. I had not needed to know for years, and I had to look-up which was referenced in beta.) (And Google is swamped in fraternities, investopedia, and bad coursework.)

If a 2N9876 has beta near 100, then alpha is near 0.01. OTOH if grandpa's 2N22 has beta near 2, alpha is near 0.5 (really 0.33).
https://www.petervis.com/Education/transistor-alpha-beta/transistor-alpha-beta.html
https://www.electronics-tutorials.ws/transistor/tran_1.html

[R.G.]

Could be. That part of the archives is pretty dusty.  :-)

[PRR]

GE Transistor Manual 5th ed 1960 has fairly authoritative and near complete list of definitions.

https://www.steampoweredradio.com/pdf/ge%20transistor%20manual/GE%20Transistor%20Manual%20Circa%201960%201.pdf   50MB PDF

[Rob Strand]

I'm pretty sure alpha = IC/IE and typically applies to common base mode.

IC = beta*IB
IE = (1+beta)*IB
so alpha = IC/IE = beta / (1+beta)  = 1/(1+1/beta)    which is just under unity.

The reverse beta (and reverse alpha) is it's own thing.  It cannot be written in terms of forward parameters: beta, alpha, hFE, hFB; whatever flavor you like.

In spice models the forward beta is BF and reverse beta is BR.   There are many reverse parameters in spice.  It's very often the case that spice models have dodgy reverse parameters.

At the end of the day you can measure the reverse beta with a hFE tester by just flipping C and E.   However, like hFE the reverse hFE (hfC?)  varies with current.

[antonis]

"Beta" for current gain is called that because reversed-mode current gain is called "alpha".

That reminds me an old joke about "Early" and "Late" voltages..

[Rob Strand]

hat reminds me an old joke about "Early" and "Late" voltages.

In spice the forward and reverse parameters have their own Early voltage (VAF, VAR).