why does my brownsource sound like s**t > with audio example

[ulysses]

i think i may have found the problem

when i tested two brand new 1m2's (from the same pack as the ones soldered to the board) on their own voltage divider circuit they actually gave a split reading of 2.8v which is the same reading as i was getting on the emiiter of the tranny.

have i got dud resistors? in the same test circuit 2x 100k resistors split the voltage to 4.38v

ill try with the 100k resistors. is there a particular reason that 1m2's were used to split the voltage?

cheers
ulysses

[ulysses]

i changed the 2x 1m2 resistors to 100k

the readings on the tranny are 3.5/3.93/8.37

unfortunatly the 5th leg of the opamp is still at 2.5v

i tested the 330k resistor with my multi. it checks out 100% 330 on the dot..

cheers
ulysses

[Doug_H]

The 2 or so volts on the buffer base and emitter should be okay. The transistor Zin is very low compared to 1M2 and is therefore drawing enough current across the upper 1M2 to cause a larger voltage drop. That's ok, as the guitar signal is probably .5v max (typically) anyway. I'm more concerned about the low bias v on the opamp. There is a problem there, maybe it's wiring or something. It may or may not be the cause of your distortion problem.

[John Lyons]

I believe the 1.2M value was chosen to keep the impedance of the input/buffer high. The ratio of the resistors is what matters most. Using a 100K,200K 300K etc etc shold keep the same ratio and voltage. It's just suppling half voltage, half way between ground and +9v. Unless I'm mistaken.
Take your metter and check the resistance of the 1.2M reisitors. As long as they read close to 1.2M then there isn't any other factor that could be throwing off voltage. Check all the resistors in the circuit for that matter. It's a good practice to get into when you are inserting parts initially. This way you will be sure that everything is correct as far as values.

John

[Doug_H]

I believe the 1.2M value was chosen to keep the impedance of the input/buffer high. The ratio of the resistors is what matters most. Using a 100K,200K 300K etc etc shold keep the same ratio and voltage. It's just suppling half voltage, half way between ground and +9v. Unless I'm mistaken.

Actually, those large values are more typical with a FET buffer than with a BJT buffer. FETs have a high Zin that you don't want to compromise with the bias network. BJT buffers are typically designed such that (in this case) R3 || R4 would be much lower than hfe * R5. So for an hfe of 100, for example, you would want R3 || R4 to be much lower than 390k. This is to insure that enough current is supplied to bias the BJT properly. With R3=R4=1.2M you can see why- the current drawn through the bias network is causing a 7 volt drop in the base voltage, "compromising" it if you were intending on having 4.5v at the base.

I still don't understand why the designer chose to use a buffer at all. If the op amp was wired as a non-inverting amp you could set the input impedance wherever you want. From a purely electronic standpoint, a lot of this (as well as other parts of the circuit) seem like poor design choices. There may be a method to the madness though. For one thing, for a guitar signal 2v is plenty. Perhaps the the assymetry gives it a more pleasing sound when driven by a hotter signal from another pedal (?). Who knows? I haven't tested that and am just speculating.

As for the Zin, how much is enough? Is the difference between the 1.2M and something more reasonable (like 100k, as Ulysses found) a significant difference?  On the version I'm building I dumped the buffer, used a non-inv amp configuration and am using a 470k from + to Vr. That sounds good to my ears. Is the buffer really needed? I don't think so, but there may have been a reason.

[ulysses]

i did a test. two 22k resistors split 9v and then pass through a 330k resistor. the result 3.35v. sounds right.

i was getting 2.5v at the opamp leg 5.

so i took the opamp out of the cradle and tested it. the result. 3.15v

so the circuit drops the voltage slightly, im inclined to believe that 3.15 is the correct voltage delivered to pin5.

with the opamp in it drops to 2.5v so the opamp is using .65v.

if two 22k resistors passing through a 330k resistor = 3.35v then how on earth are you guys getting more than 4 volts on the 5th leg?

cheers
ulysses

[ulysses]

i tested all the 1m2 resistors i had. none of them were 1m2 in value. it was like trying to match jfets. they were all over the place.

dodgy jaycar.

i have since been around the board to confirm that all the other resistors are right. they are.

cheers
ulysses

[John Lyons]

On my PCB and schematic pin 5 is the first opamp which get Vref straight from the 22K/4.7 voltage source.
Pin 3 has the 330K connected to it. The Vero and PCB layouts swap the opamp sides. So your pin 5 is my pin 3.

John