FZ-1 with 9V instead of 3V?

[fuzzy645]

Does anyone have a schematic for a FZ-1 using a 9V instead of the 3V?  If not, what do you think might need to change for this?  Seems like it would be a lot more convenient to mount a 9V than a pair of AA's.    The first dumb idea that came to my mind is to simply slap a voltage divider right off the 9V battery to trick the entire circuit into seeing only 3V. This way the circuit can be left original/vintage except for the extra voltage divider.     

[jrod]

There is a tutorial on the Small Bear Electronics web site. 

[smallbearelec]

The first dumb idea that came to my mind is to simply slap a voltage divider right off the 9V battery to trick the entire circuit into seeing only 3V.    

Not a dumb idea at all...perfectly logical. There were a number of design choices in this pedal that related to what was possible at the time that we would not make now. I did my best to fix all the "bugs" I could think of in doing the Fuzz E-One:

http://www.smallbearelec.com/Projects/FuzzE-One/FuzzE-One.htm

Do try building one. If you like Fuzz that's raspy and gnarly (in a good way) you will like this pedal. And you may already have the parts lying around...

[fuzzy645]

Cool....thanks!

[LucifersTrip]

http://www.smallbearelec.com/Projects/FuzzE-One/FuzzE-One.htm

Do try building one. If you like Fuzz that's raspy and gnarly (in a good way) you will like this pedal. And you may already have the parts lying around...

I did use that as a "check" after I build a few FZ-1 variants...but I was slightly confused at one thing since the voltages in the tweaking section of the .htm file are different than the pdf of the build.

Were the tweaking voltages supposed to be read with only the components installed up to that Q stage?

=====================
http://www.smallbearelec.com/Projects/FuzzE-One/FuzzE-One.htm

Q1...E should be .4 to .6 volts with R4 at 10K.

Q2 ... puts about 1 Volt on the collector of Q2.

Q3... put .6 to .8 volt on the collector.
=====================

pdf:
=====================
http://www.smallbearelec.com/Projects/FuzzE-One/Kit/FuzzE-One.pdf

Time to break out the multimeter and figure out where the problems are. Here are the
measured transistor voltages in the build you see here:
Collector Base Emitter
Q1 1.68 .824 .78
Q2 .53 .07 0
Q3 1.1 .02 0
=====================

[smallbearelec]

Were the tweaking voltages supposed to be read with only the components installed up to that Q stage?

That's what I did when building the perfboard version. When I did the PC board build for the kit, I just tweaked until I liked the resulting Fuzz and then measured. The setup procedure will get anyone into the ballpark. From there, it's a matter of whether you like what you hear as is or want to try other combinations of operating points.

[fuzzy645]

Another quick question.

I noticed the original circuit uses a couple of 20 uf coupling caps.  Since this value isn't as common nowadays, can you substitute a 22 uf, which is a more common value?

Thanks!

[smallbearelec]

can you substitute a 22 uf?

That's what I use...works fine.

[LucifersTrip]

Were the tweaking voltages supposed to be read with only the components installed up to that Q stage?

That's what I did when building the perfboard version. When I did the PC board build for the kit, I just tweaked until I liked the resulting Fuzz and then measured.

Sounds cool...so you did indeed measure the voltages in the perf version using only a partial circuit, one stage at a time...

In the instructions, did I miss the final voltages for that build? If not, what are they?
http://www.smallbearelec.com/Projects/FuzzE-One/FuzzE-One.htm

...and still a little confused by Q3's C voltage ("Q3... put .6 to .8 volt on the collector."). When you tweaked that, all 3 transistors are in place, so it should be the same (close at least?) to your measured voltage of the other final build (reported 1.1v)

thanx much

[PRR]

> the voltages in the tweaking section of the .htm file are different than the pdf

This is a VERY crude circuit. Bias of Q1 and Q3 is totally depending on transistor leakage... analyze these stages using "modern conventions" (negligible leakage) and it doesn't work at all. Bias of Q2 is very variable with the trim-pot in its base network.

Few of the voltages are at all critical. You only have to get close enough to slam. Once this thing starts slamming signal it's gonna do what it wants, no matter what "idle bias" you had before you play.

First get the "major mis-wire" voltages:

Supply rails are zero and 1.5V (1.7V in "9V" version). Check polarity against transistor type PNP or NPN.

The key "oops!" points are:
Q1: C=1.5V (or 1.7)
Q2 Q3: E=0V

The "lucky leakage bias" points:

Q1: E should be at least a few tenths of a Volt above zero.
Q3: C should be more than a few tenths up but nowhere near the +1.5V rail

These voltages are required so that Q1 and Q3 pass signal and Q3 has reasonable gain and output level. Within these zones, exact voltages are not too critical. Using random devices, getting even "near" these biases is mostly luck, or tedious resistor selection. I see SmallBear offers kits with resistors selected to work well with the supplied devices, a great idea.

Bias, gain, and output level of Q2 are all adjustable with the trim-pot. As Q2 stage is similar to Q3 stage, the same initial check values apply: C should be more than a few tenths up but not jammed against the +1.5V rail, and the C voltage should vary with the trim pot.

Except the Q1 C and Q2 Q3 E "oops!" voltages, none of these numbers are carved in stone. What is the worst can happen? Well, gain could be "low' but the eXtreme overdrive means some part of the wave will pass a high-gain zone. Mostly: it will distort. Which is what we want. It will distort different for different bias, though again the eXtreme overdrive means the wave will pass through all distortion zones. The designer gave you the trim-pot which shades the distortion one way or another to user's taste.

[fuzzy645]

> the voltages in the tweaking section of the .htm file are different than the pdf

Bias of Q1 and Q3 is totally depending on transistor leakage...

So just for purely educational purposes, I would like to understand the reason for this.  To keep things as simple as possible for starters, lets just take Q1.  I see a 100K resistor (R1) in series with a .01 uf cap (C1) which then feeds a 1M resistor to ground (R2) as well as the base of the 1st transistor (Q1).

In discussing the bias of Q1, I suppose we would need to determine both the current and voltage presented to the base of Q1.      Now, if C1 wasn't there, the 2 resistors would look like a voltage divider to me, but since C1 is in there, I really don't know what to do with this. 

So, my questions are as follows:

1.  What is the math to calculate the bias of this transistor?

2.  Suppose the Q1 had almost zero leakage (and I realize this is a false assumption).  What would then be the issue?  Would it be that not enough current flows into the base of Q1 so that Q1 would never turn "on" so to speak?

Thanks

[PRR]

> In discussing the bias of Q1, ....      Now, if C1 wasn't there...

For discussing bias, the no-signal DC condition, the capacitors are not there.

A capacitor is Open-Circuit for DC. It can be replaced by a no-connection.



> What is the math to calculate the bias of this transistor?

WWJD? Joseph The Carpenter could figure many things, but he had no math and arithmetic so awkward it was only good for counting.

Using 33BC figuring but 1970AD transistors, Joe would think thus: For NPN, current must flow down from the Collector supply INto the Base to bring the Base more Positive than Emitter. In the FZ1a there is no path for such a current. Therefore the Base Current is "none". The Collector current is many-many greater than Base current, but many-many of none is none. Therefore nothing happens.

If you have a concept of "zero" and handy multiplication, still nothng happens.

So how does FZ1 work? Pre-1963 (and later surplus) transistors leak Collector-to-Base. How much? Depends on impurities, price, temperature, light, even lead-stress.

The data-sheet parameter could be Iceo. This is, not the collector-base leakage current, but Icb multiplied by whatever current gain the transistor has at that Icb, which gives the collector-emitter current.

If Iceo is, say, 0.3mA, and that current flows in Re 10K, we math "3V". But that won't happen with only 1.5V supply. Ah... as Vb rises, Rb 1Meg diverts some Icb, so does not get current-gained by the transistor.

We actually want an Icer spec, where "r" is very close to 1Meg (and Vce is about 1V). Even when we bought these new, we didn't get that spec (except on custom-order... one reason Steve may be finding clusters and voids in his Ge collection.)

We could pretend that Ge leakage is similar to an internal resistor C-B with values from 10Meg to 10K. Yes, leakage does vary that much, and no the leakage is not really a resistor.

Lookit. If you want to understand practical transistor bias, do NOT study this circuit.

[LucifersTrip]

> the voltages in the tweaking section of the .htm file are different than the pdf

Few of the voltages are at all critical. You only have to get close enough to slam. Once this thing starts slamming signal it's gonna do what it wants, no matter what "idle bias" you had before you play.

These voltages are required so that Q1 and Q3 pass signal and Q3 has reasonable gain and output level. Within these zones, exact voltages are not too critical.

Yes, you are correct, you can get a fuzz in a pretty large range, from maybe ~ .5v - 1.2v, but the specific spot you settle on makes a pretty big difference. The .htm instructions suggests .6 - .8 (the cliche 50% supply, which was recommend by others when I built variations), but I have almost always gotten a better tone in the 1.1 - 1.2v range in all my FZ-1 variant builds...as noted in pdf instructions file.

Using random devices, getting even "near" these biases is mostly luck, or tedious resistor selection.

Correct...though, this circuit is so well known we do have known hfe's/leakages to start with. With that knowledge, resistor selection is very easy.

...which leads me to my unanswered question from a related thread:
http://www.diystompboxes.com/smfforum/index.php?topic=95071.msg823237#msg823237

I was repairing an FZ-1 variant (LRE Fuzz) and needed to replace Q2 & Q3. I measured the removed Q2 & Q3 from the original and they had gains of 53 & 62 respectively, with very low leakage (~ 50-75uA), which leads me to my only question:

Knowing that Q3 has to have high leakage (~300uA+), could that pedal ever have sounded good ?  With the original Q's, it produced only a barely audible gated blat when you bashed the strings....

[fuzzy645]

> In discussing the bias of Q1, ....      Now, if C1 wasn't there...

For discussing bias, the no-signal DC condition, the capacitors are not there.

A capacitor is Open-Circuit for DC. It can be replaced by a no-connection.



> What is the math to calculate the bias of this transistor?

WWJD? Joseph The Carpenter could figure many things, but he had no math and arithmetic so awkward it was only good for counting.


Using 33BC figuring but 1970AD transistors, Joe would think thus: For NPN, current must flow down from the Collector supply INto the Base to bring the Base more Positive than Emitter. In the FZ1a there is no path for such a current. Therefore the Base Current is "none". The Collector current is many-many greater than Base current, but many-many of none is none. Therefore nothing happens.

If you have a concept of "zero" and handy multiplication, still nothng happens.

So how does FZ1 work? Pre-1963 (and later surplus) transistors leak Collector-to-Base. How much? Depends on impurities, price, temperature, light, even lead-stress.

The data-sheet parameter could be Iceo. This is, not the collector-base leakage current, but Icb multiplied by whatever current gain the transistor has at that Icb, which gives the collector-emitter current.

If Iceo is, say, 0.3mA, and that current flows in Re 10K, we math "3V". But that won't happen with only 1.5V supply. Ah... as Vb rises, Rb 1Meg diverts some Icb, so does not get current-gained by the transistor.

We actually want an Icer spec, where "r" is very close to 1Meg (and Vce is about 1V). Even when we bought these new, we didn't get that spec (except on custom-order... one reason Steve may be finding clusters and voids in his Ge collection.)

We could pretend that Ge leakage is similar to an internal resistor C-B with values from 10Meg to 10K. Yes, leakage does vary that much, and no the leakage is not really a resistor.

Lookit. If you want to understand practical transistor bias, do NOT study this circuit.

Thanks for your reply.  Your picture in particular has helped me because I was looking at the 100K as somehow part of the bias, when in looking again at the circuit that appears to be more of an input load resistor type of thing.  

So really in looking at the circuit again, Q1 is supposed to be an emitter follower to set a high input impedance (aka a "buffer").  A more "modern" emitter follower (although this is NPN) would be as depicted by Jack Orman here:



In theory now, (speaking in "object oriented" terms as I am a programmer by trade), couldn't we just simply replace the entire "leakage dependant" Q1  part of the circuit here with a PNP version of the Orman buffer, and then we could use a totally non-leaky (heck, even silicon) PNP transistor for this part of the circuit.    Wouldn't that be more reliable?  

To my (untrained) eyes, Q2 and Q3 look pretty "run of the mill" for a fuzz and not entirely dis-similar from (lets say) the Fuzz Face.

Am I smokin' crack here?

I do have one other question though. I am wondering why you indicated 1.5V at the top as opposed to 3V?  What part of the circuit is a voltage divider that is chopping the 3V in half?

[PRR]

> ...depicted by Jack Orman here:

Please do not hot-link to images on Jack's pages. It annoys his server and now I am banned from his site.

I have asked the Moderator to remove the link in your post.

[PRR]

> why you indicated 1.5V at the top as opposed to 3V?

FV1 was produced in both 3V and 1.5V versions. Take your pick.

> simply replace the entire "leakage dependant" Q1  part

There are many ways to do it. When this box was new, "reject" Ge parts sorted by leakage were sold by the barrel, and this was mass-producable. Unfortunately the technology improved.

Note also that Q3 has no explicit base bias and idles at uncertain Collector voltage. This is perhaps no big deal, since "any" input signal is amplified in Q2 to many tenths of a volt, and Q3 Collector is sure to be slammed floor to ceiling (zero to +1.5V/3V).

[fuzzy645]

> ...depicted by Jack Orman here:

Please do not hot-link to images on Jack's pages. It annoys his server and now I am banned from his site.

I have asked the Moderator to remove the link in your post.

I removed the linked image myself, no worries.

Thanks for your reply

[aron]

>It annoys his server and now I am banned from his site

Is that even possible?? Wow?

[amz-fx]

There is no ban mechanism for linked images on my site.  I'm not sure what the problem is, but I'm in contact with my web host to see what can be done.

Sorry for any inconvenience this has caused.

Best regards, Jack Orman

[PRR]

Jack is looking into it; however I have found that my _IP_ is not "banned" because I can see his site with FF or Chrome. I'm wrestling with IE, even changed from IE7 to IE8 (this is not an upgrade, IE8 is buggier than my compost in July).

A remote thought is that my doofus ISP is caching; however they are not smart enough to know what browser and cache differently I have so that's unlikely.