AMZ Mini Booster too much gain

[samrsmiley]

I've just build an AMZ Mini Booster which seemed to work properly when I tested it, albeit with a tone generator and a solid state amp.  When I put it onto my board though it has significant gain-almost splatty distortion-not clean whatsoever.  The weird thing is that it seems to get more gain as the effect is turned on, the longer it's on the more gain it has. 

There are a fair amount of pedals in the chain, but all were off.  The only buffered pedals are placed before the AMZ.  After the AMZ there is a Hall of Fame reverb which is on. 

Just so it might help with any diagnosis Gtr-Fender Tuner pedal-Cusack Screamer Fuzz (buffered)-Nobels ODR (buffered)-Wampler Euphoria-AMZ MINI BOOSTER-Wampler Ego Comp (off)-Wampler Faux Echo (off)-Rogue Tremolo-TC HoF verb (on)-amp (custom built based on a Fender Deluxe).

The tubes on the amp are pretty old, but I also tested into my Vox AC30 with the same results.  All pedals other than the AMZ and HoF were off, and all the ones not marked buffered are true bypass as far as I know.

Any help would be really appreciated!

[amz-fx]

Increase the value of the source resistor on the lower jfet to reduce gain... it is R5 in the online schematic, and is shown as 1K.  Make it 10k and see how that sounds.

Best regards, Jack

[samrsmiley]

thanks Jack, is it Ok to use slightly different value?  I'm not sure if I have any 10k's around.

[Ice-9]

You could also replace the 1k R5 resistor with a 10K potentiometer. This would give you a  controlable amount of gain. At its maximum resistance of 10k the volume pot would give a nice clean boost just beginning to distort at is maximum level and with the 10k pot at the minimum altering the volume pot would give you a distorted sound all the way up to its max level. Consider it a Dual Booster.

[amz-fx]

thanks Jack, is it Ok to use slightly different value?  I'm not sure if I have any 10k's around.

Sure, any value between 1k and 10k.  You should remove C6 also.

I would take out C6 first, then start increasing the resistor value as needed.

regards, Jack

[PRR]

Jack-

I see on your AMZ Mini-Booster page some history of the gain-cell traced to the 1970s.

For what little it is worth, that NatSemi cite was shown in a class I took musta been 1975-1976.

I've just noticed the same core amplifier in a 1966 book.

Book scan: http://i.imgur.com/vLkOx.gif (42KB GIF image)

Analysis And Design Of Transistor Circuits
Laurence G. Cowles
Van Nostrand, 1966
309 pages

Touted as a DC amp because instrumentation work was big at the time.

The NatSemi plan 'forces' op-point to the center, the 1966 plan wants critical matching/trim. For signal there's very little difference (1966 plan has a touch more gain _if_ it biases).

And the 1966 plan shows the buffer which is essential if MAXimum stage gain is a primary goal. (In audio we often do not want MAX gain.)

Knowing Cowles, this isn't his invention, though he doesn't give individual references, just a general bibliography for the chapter. I would look first at Evans' Biasing FETs for Zero D-C Drift, though I never heard of that journal so it may be a long hunt. The RF Shea is sure to be worth reading even if it does not cover this amp.

[amz-fx]

Paul,

There is a similar version of it in a patent from 1964 where it was listed as "prior art". Check it out:

http://www.muzique.com/news/mu-follower-history/

It was relatively well known even at that time.

regards, Jack

[samrsmiley]

Finally getting around to making this change-I can't believe I was actually writing back and forth with the creator of the circuit!  Very cool, thanks for the help Jack, and I'll let you know how it does.

[samrsmiley]

Those two things (taking out C6 and increasing R5 to 10k) worked! 

Just curious as to why those are in the schematic? 
The pedal has to be about 9 o'clock to get much of a boost, but that's totally fine since I'm looking for a bit of a boost and to keep it really clean.  Thanks for the help!

[Paul Marossy]

Those two things (taking out C6 and increasing R5 to 10k) worked! 

Just curious as to why those are in the schematic? 

R5 is the resistor on the source of the lower FET. Messing with the value of it changes the amount of gain. C6 is in parallel with R5 and having it there also increases the gain. Just removing C6 should make a noticeable difference in the amount of gain.

[duck_arse]

can anyone tell me if there is a formula for the output impedance of the minibooster circuits? and if so, what it is.

[PRR]

> can anyone tell me if there is a formula for the output impedance of the minibooster circuits? and if so, what it is.

The base circuit is "nearly infinite".

Much depends on the FETs' internal construction and the bootstrap cap's size versus frequency. But hundreds of K.

(This also means the gain is high but hard to predict. Which may be why samrsmiley's result was overpowering.)

As described by Jack, there is a 100K pot at the output. We see by inspection that if the pot is way down, Zout is low. At 10%, it can't be over 10K, or under 9K if the FETs were instead zero Zout. At 50%, again less than 50K but probably way above 25K. Full-up, max is 100K, min Zout is unclear although with typical FETs probably very close to 100K (say 90K, and who cares?).

[duck_arse]

so an impedance buffer following is a good idea for consistency?

[PRR]

Why?

[fuzzo]

I've built two mini boosters, both had too much gain . I just removed the bypass cap and since i've no complaints .

For the impedance , isn't it Rinternal/Rload ? but the upper JFET makes it different  because that's not a resistor 

[duck_arse]

prr, I would have thought "nearly infinite" output impedance was bad for any audio circuit.

the high output z means the response is more dependant on the circuit that loads it, surely, and therefore less predictable.

[PRR]

Two arguments:

#1- It is a lot of gain. In most useful situations the output pot will be less than half up, output impedance under 50K. 50K is a fairly low impedance relative to most *guitar chain* inputs you will drive (because a naked guitar is often over 50K and they don't want to load that).

#2- Jack's had it posted for years and years. It has probably been built many times. If it "needed" a buffer, he would have shown it buffed-out.

Go ahead and try it. Experimentation is the best teacher. And it is trivial to buffer this. But I doubt there's any "wow!" change.

[duck_arse]

I'm only asking because I came across someone's circuit with a note that the high output impedance of the miniboost meant they had added a buffer. I'll see if I can track it down.

[wavley]

Two arguments:

#1- It is a lot of gain. In most useful situations the output pot will be less than half up, output impedance under 50K. 50K is a fairly low impedance relative to most *guitar chain* inputs you will drive (because a naked guitar is often over 50K and they don't want to load that).

#2- Jack's had it posted for years and years. It has probably been built many times. If it "needed" a buffer, he would have shown it buffed-out.

Go ahead and try it. Experimentation is the best teacher. And it is trivial to buffer this. But I doubt there's any "wow!" change.

Folks most likely get the buffer idea from R.G.'s article, he output buffers one in the 2nd schem down.

http://www.geofex.com/Article_Folders/modmuamp/modmuamp.htm