Big Muff Pi HFE

[Jdansti]

I'm finally going to build a BMP, and the version I'm doing calls for all transistors to be 2N5088s. I've got two batches from two manufacturers, and one batch's HFEs are ~350. The other batch's HFEs are ~600.

If I'm looking for a more "classic" rock/ blues fuzz, should I try the ones around 350 first?  I'll probably socket the trannys and see they sound like, but I wanted to see what other people think about HFE levels on the BMP.

[LucifersTrip]

my favorite ones have always been made with 150-300 hfe..the triangles & ram's heads.

you'll notice that's there's a big change in collector voltage from 50 to around 100 hfe, and a much smaller change afterwards.
http://www.diystompboxes.com/smfforum/index.php?topic=98421.msg860294#msg860294

so 350 to 600 won't be that big of a deal, but 50 - 100 would be.

edit:
but there's a bigger difference with Q4
http://www.diystompboxes.com/smfforum/index.php?topic=98421.msg860638#msg860638

[Mark Hammer]

Transistor gains are not fixed, in the way that preamp tube gains are (and even there, that can be varied).  Their voltage gain will depend on the biasing and components surrounding them.

What that implies is that the hfe is not the determinant of the gain, and consequent clipping, in that stage.  And, if you look at any of the sites that show the schematics of different BMP issues, you'll see that not only do the transistors change from issue to issue, but the emitter and collector resistances also change.

In theory, one could use a very broad range of transistors and hfe specs to arrive at the identical gain/tone, by adjusting the corresponding resistors.  When folks make a recommendation for this or that transistor type or hfe range, it is important to consider they are making that recommendation for a given context, biasing scheme, etc..  What that also means is that you may not need to drives yourself silly in search of different transistors, but may come to the same tonal goal merely by tweaking those associated resistances.

[Govmnt_Lacky]

@John

The best and only advice I would give you is to follow your EAR! 

After all... do you really think that EHX "matched" transistors for HFe? Hell, they are well known for using whatever parts they had on hand! 

Good Luck! 

[Jdansti]

Thanks, guys. I looked up several schematics, and yes, the emitter and collector resistances change from version to version. The capacitors in the clipping stages also vary. I think I'll breadboard it first and play with the resistor and capacitor values to find what sounds best to me, as Greg suggests. I'll probably try different clipping diode configurations as well. So many permutations!  Thanks again!

[iq01221]

Trimpots maybe?

[Jdansti]

^I initially thought of that, but would I need two trimmers per transistor?

[Mark Hammer]

Twer I, I'd use a two-pin socket for the collector resistor, and a 500R trimmer for the emitter resistance to ground.    That would strike a reasonable balance between simplicity and flexibility.  You know the emitter resistance is going to be smaller than 500R, but the collector resistance could vary more widely, and widely enough that a trimpot might be a little twitchy.  A socket lets you pull a 12k, stick in a 15k or 18k, and then twiddle the emitter trimpot to taste.

[Jdansti]

Good idea Mark, thanks!  As it turns out, I put it on the breadboard and had a go at it when I got home from work today. I spent several hours building it and twiddling with the emitter resistors, the input caps for each gain section and the diodes. Tomorrow I'll play with the collector resistors (which will probably require more emitter resistor twiddling) and the transistors.

I found the silicon diodes to be harsh and blatty. I pulled both pairs of diodes and placed germaniums in the second clipping stage and it sounded much smoother and better to my ears.  The decay seamed to fizzle early with no diodes in the first clipping stage. I tried silicon, germanium, and LEDs in the first clipping stage and settled on 3mm clear reds. It seemed to improve the sustain-either that or I was hallucinating after a four hour stretch of swapping parts while I banged on my guitar. I'll check it tomorrow with fresh ears and brain.

I'll post a schematic of my results and a sound clip when I'm finished.

BTW, I've been using this web page as a reference of the different types of BMPs out there.  http://www.bigmuffpage.com/Big_Muff_Pi_versions_schematics_part3.html

[Jdansti]

I'm back with a final report.

I wasn't happy with my breadboarding results and I was about to give up when I remembered that I had built JMK's SCUBA Muff a little over a year ago as a Christmas gift for my brother in law and I really liked how it sounded. I had built it on a PCB from JMK. I didn't have another PCB, so I modified an existing vero layout to match JMK's values and I made modifications to accommodate JMK's (1) input cap selector, his (2) clipping diode switcher in the first clipping stage, and (3) his deselector switch for the second clipping stage.

JMK's diode switcher switches between two sets of clipping diodes, but I used a 4 throw rotary switch and used the following combinations:

1) Two 1N4148
2) Two 3mm LEDs
3) Two Ge
4) Three Schottkys  arranged asymmetrically

Here's the vero layout that I started with:
http://www.sabrotone.com/wp-content/uploads/2010/04/Big-Muff-Pi-Creamy-Dreamer.gif

Here's JMK's schematic and BOM:
http://www.jmkpcbs.com/JMK_PCBs/Store_files/Scuba%20Muff.pdf

Here are my photos. This has to be one of my ugliest wiring jobs, but it sounds great!  This is my second pour-on epoxy job. Thanks for all of the ideas, guys!





The vero board looks like hell because I made grid lines with a Sharpie, and then when I was finished, I cleaned the back of the board with acetone. The Sharpie ink ran into the board's holes and made it look like I had burned the hell out of the board. Lesson learned!

[Gus]

Read LucifersTrip post and the linked threads in the post to get a better understanding of what matters
I agree the Q4 stage is the more sensitive one with different beta
The other stage transistors matter but not as much the 1st more than the 2nd and 3rd with larger input signals, again read the threads linked
The 2nd and 3rd stage have the output swing clamped by the diode clipping and the gain is less than 470K/8.2K

look up transistor gain vs collector current
Note that with the same value collector resistor if the collector voltage is closer to ground(+9VDC circuit) the transistor has a higher collector current
Read R.G. treble boost article note the transistor is set to 7.2VDC

Something I posted before

[Jdansti]

Thanks a lot, Gus. I've still got a lot to learn about this circuit. One thing I've found out is that the transistor types and their associated resistors are just one parameter that determines the resulting sound. The caps between the gain stages and within the clipping sections as well as the diode types and configurations also have a huge effect on the sound. After a about eight hours of trying different transistors, resistors, caps, and diodes, I got to the point where I felt like I was trying to reinvent the wheel.

I still would like to tinker with the circuit on the breadboard, but take a different approach. One thought is to place a diode clipping circuit on all four gain stages and either switch them in and out, or place pots in the clipping stages, a la Orman: http://www.muzique.com/lab/sat2.htm. Maybe someone's already done that to the muff circuit.

[bluebunny]

Looks great, John.  I closed my eyes for the gut shot...   

[digi2t]

Came across this the other day, with gains listed for each transistor.



A bit late, I know. Sorry. Very interesting site though. I`ve been going through it intensely of late, because I`m planning a build as well.

[Jdansti]

Thanks Marc and Digi2t.

^That's an interesting page with good information in it.

[LucifersTrip]

Came across this the other day, with gains listed for each transistor.

I have to admit, it's tough to believe those 2N5133 button style transistors from the 60's - early 70's have hfe's in the 600-700 range, but who knows. Those batches really varied, especially back then.

The vintage 2N5133 and 2N5134's I own rarely have hfe over 150, but the data sheet for the 33 shows 60 -1000.

[Arcane Analog]

I have measured 2N5133s from the 30s to the 500s.