Leaky Q3 in Maestro FZ-1A and variants...

[LucifersTrip]

I've built the FZ-1A and numerous 1.5v variants (Fuzz King, Olson New Sound, etc...) and I didn't run into a any hiss problems at all until I breadboarded the LRE Fuzz to troubleshoot a vintage one:
http://www.diystompboxes.com/smfforum/index.php?topic=94768.msg820869#msg820869

...and I've listened to many sound clips over the years and haven't heard any prominent shhhhhhhh...

So, a few questions:

Is it common to get hiss (shhhhhhhh...) in the FZ-1A and it's variants since a very leaky Q3 is needed to bias it correctly?

or...

Is the hiss accounted for somehow by design? ...and therefore there is no hiss like you'd get using a leaky transistor in a circuit that shouldn't (ie, Fuzz Face) have one?

Did the makers of the original actually sort the 3 transistors by leakage and throw the leakiest one into the Q3 slot?

or...

Is that something DIY-ers discovered relatively recently that would give the best sound? ...and the only really good sounding vintage ones had a randomly thrown in leaky Q3 ?


I've built so many of these but never really thought about this until the hissy LRE. I can't get anything below ~ 1.4v on Q3's collector with any transistor with low to medium leakage. The second I put a high leakage one in there (300+ uA), the voltage drops to .6 - .8v, the fuzz & the sweep are excellent, the sustain & decay are seem better, but I get a shhhhhhhhh...

[smallbearelec]

Is it common to get hiss (shhhhhhhh...) in the FZ-1A and it's variants since a very leaky Q3 is needed to bias it correctly?

Is the hiss accounted for somehow by design? ...and therefore there is no hiss like you'd get using a leaky transistor in a circuit that shouldn't (ie, Fuzz Face) have one?

Did the makers of the original actually sort the 3 transistors by leakage and throw the leakiest one into the Q3 slot?

We have sorted a good number of sets by now. Q1 should be relatively low gain/low lealage, so hiss does not usually originate there. A hissy Q2 is bad, because Q3 will amplify any noise in that stage and add its own. We have had pretty good sucess in producing sets we are happy with, maybe because we have a lot of raw stock from which to sort. Try doing an NPN set with an Amperex 2N2430; they are not expensive.

[PRR]

> Did the makers of the original actually sort the 3 transistors by leakage

Yes.

I don't know this specific product. But in general there were Good Spec parts sold for High Prices (military, computers), and "reject" parts unqualified for high-price sales which were thrown into barrels and sold by the pound.

A transistor with hFE 50-100 and Vcbo>30 could sell for $7. Folks who needed 30V breakdown, or a specific range of hFE, "had to pay".

But there was a lot of cost-sensitive gain-tolerant 3V and 6V stuff being made.

Transistor radio makers certainly sorted barrels of "rejects". An added stage of low-gain trumps fewer stages of hi-gain IF the price comes out less. In fact lo-gain transistors were not bad because the radio said "7 Transistors!" on front which is clearly better than a "6 Transistors!" radio at the same price. You needed a low-gain low-volt high-frequency tranny for the front end, some gain and frequency for the IF, and trannies too slow for any "serious" work were good enuff for the audio stages. Regency dabbed paint, IIRC blue was an IF and pink was a 1st AF transistor, no matter what shape or original part-type.

Lower-volume makers might turn to Steven's predecessors, garage-ops who bought a barrel for $100 and paid neighbors $1/hour (circa 1960) to poke leads and sort by speed and gain (not so much breakdown.... low-budget work was all low-voltage). These are the same guys who bought mil-surplus crates of tubes, sorted the in-demand types, and listed the rest in the back of the magazines "33 cents any tube!* (*on this list)".

[LucifersTrip]

We have sorted a good number of sets by now. Q1 should be relatively low gain/low lealage, so hiss does not usually originate there. A hissy Q2 is bad, because Q3 will amplify any noise in that stage and add its own. We have had pretty good sucess in producing sets we are happy with, maybe because we have a lot of raw stock from which to sort. Try doing an NPN set with an Amperex 2N2430; they are not expensive.

Thanx...I didn't want to waste space in this post, but in my build report (linked above), I did note that I tried numerous Q1's & Q2's with no effect to the hiss problem.

I have more ge's that I will ever use in a lifetime, so finding one with a specific gain & leakage is really no problem.

After a bit more experimenting, I am no longer certain that Q3's collector voltage should be ~.75v on this one.
I just remeasured Q3's collector on a couple of my excellent sounding  FZ-1A variants (Olson & Fuzz King) and both measured 1.25v+. The Fuzz King is very similar to the LRE (see schem link below).

So, for the hell of it I just changed the 40k base pull-down resistor (which was set to put .8v on Q3's collector) back to the stock 5K. The voltage shot straight to 1.4v as expected and most of the hiss disappeared. I don't get as smooth of a saturated fuzz but it does sound more like the dirty fuzz I'd expect and actually more like the YouTube sound file.
http://www.youtube.com/watch?v=5zHrMdykbF0

In addition, the higher strings are not as harsh as at .75v and are much more playable for single note runs.  

I don't know if this guy had a good working one, but his reported Q3 collector voltage is 1.44v, but he does note that the low base pull-down R does make it gated:
http://www.guitar-pedals-effects.com/FuzzKing.gif

Unfortunately, I don't have any of my FZ-1A's handy now, but you are sure that Q3C should be ~ .6-.8?
Has anyone built any of the obscure variants and measured those voltages ?

thanx much

[LucifersTrip]

> Did the makers of the original actually sort the 3 transistors by leakage

Yes.

but did the makers of these low budget fuzzes (especially the lower budget knock-offs) actually take the time and energy to "throw the leakiest one into the Q3 slot?"

A transistor with hFE 50-100 and Vcbo>30 could sell for $7. Folks who needed 30V breakdown, or a specific range of hFE, "had to pay".

when I first started to pick up some vintage transistors, I did marvel at the old adverts showing transistors selling for $20+ ...in the 50's & 60's

Transistor radio makers certainly sorted barrels of "rejects". An added stage of low-gain trumps fewer stages of hi-gain IF the price comes out less. In fact lo-gain transistors were not bad because the radio said "7 Transistors!"

It is funny that you noted that. I have a "contains 10 transistors" minimum rule in mind when buying old transistor radios for salvage. The last 2 I destroyed said "all transistor" (I took a gamble on count) and "12 transistors". I've read  accounts where it was stated that transistors were thrown in for count, with no function.

thanx for the info

[LucifersTrip]

I am finally done tweaking this thing. Since there seems to be little interest in this FZ-1A variant, I'll make this brief.

Q1 E  approx -.6 to .65v, hfe 150+ low leakage
Q2 C approx -.8 to .9v, hfe 80-100 medium leakage
Q3 C approx -1.12 to -1.2v, hfe 70-80, high leakage

Q2: As I stated, I chose this after Q3 and it was the most difficult. It needed just the right gain & leakage and was much easier to choose once I got a reasonable sound with a good Q3

Q3: If low leakage, then Q3 C approx -1.4 to -1.5v
High E string dull until voltage drops below -1.3v. If you increase base pulldown the hiss steadily increases as the voltage drops. The quality of the fuzz is no better at -.75v so I only increased the bass pulldown to 10K bringing Q3 C to approx -1.12 to -1.2v.  (With a 40K you'd hit -.75 but it's way too hissy)

The only changes I made were: Q3 base pulldown 5K to 10K. Q2 collector resistor 10K to 5.6K

========================

Once I had this going, I started on the original, which I posted my first findings here...again, little interest, but:
http://www.diystompboxes.com/smfforum/index.php?topic=94768.0

I will post details in that thread, but the bottom line is that without making any changes, swapping out Q2 & Q3 and putting in similar transistors to what I have on my breadboarded version above gives me a really good sound....with the exact same moderate hiss as on my breadboard.

I tried a load of combos of gains & leakages in Q2 & Q3 and only got it to really fire up with Q's similar to my breadboarded version.

Final:

Q1 E  approx -.62v
Q2 C approx -.85v
Q3 C approx -1.14v

There's probably no reason to change any resistors.

I measured Q2 & Q3 from the original and they have gains of 53 & 62 respectively, with very low leakage (~ 50-75uA), which leads me to my only question:


Could this pedal ever have sounded good?

[kvb]

I'm interested in the project, but I've got no time to mess with it. I think I still have the FZ 1 on a breadboard. I was even able to find a few 270n transistors. I worked on it back in June.

It did not work well at all, but it did pass signal. I know something wasn't biased right. But even then I had too much going on to worry about it much.

These things had to have sounded decent enough for a company to have invested the effort into producing them.

Thanks for posting your findings. When I have time to revisit the project, I'll definitely read this again.

[LucifersTrip]

I'm interested in the project, but I've got no time to mess with it. I think I still have the FZ 1 on a breadboard. I was even able to find a few 270n transistors. I worked on it back in June.

It did not work well at all, but it did pass signal. I know something wasn't biased right. But even then I had too much going on to worry about it much.

Small Bear has a guide that has helped a bunch of people:   .4-.6v on Q1E, 1v on Q2C, .6-.8v on Q3C. it's under Fuzz E-One.   Though, I did mine by ear and trial and error.   My results for this one are pretty close, with Q3C being the furthest off.

These things had to have sounded decent enough for a company to have invested the effort into producing them.

The reason I asked that question is because I could not get any useable sound with anything but a leaky Q3 (and that is the general consensus), and the ones in the pedal were very low leakage.  Transistors leak with age, not the reverse, so I wondered how it could have ever worked well with the originals. I also answered one of my first questions above, "Did the makers of the original actually sort the 3 transistors by leakage and throw the leakiest one into the Q3 slot?" ....obviously not!

Thanks for posting your findings. When I have time to revisit the project, I'll definitely read this again.

cool...hope you do get around to it...it's a completely different project than the FF and FuzzRite variants.