Fuzz face problem

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've done a Fuzz face with this schematics:

An externally hosted image should be here but it was not working when we last tested it.


BUT IT DOESN't WORKS!!!

The effect is just passing a signal trought(actually sound gets a bit muddy) and i can't get a distorted tone. I used AC188 germanium PNP transistors.

Here are the pictures:
An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.

An externally hosted image should be here but it was not working when we last tested it.



If you guys can help me that would be awesome.
 
Just a couple of things, without actually tracing out the circuit:

1) Note that this is a positive-ground circuit, which is unusual by modern standards. Make sure you've got the + side of the battery (or power supply) connected to ground.

2) Are you sure that's a 100K resistor next to the transistors? It looks (on my monitor, at least) like the multiplier band is red, which wouldn't be right.

Reid
 
1) I've done this.

2) Multiplier is orange. Have you compared the color with 8,2k tranzistor (which has two red bands)?


is this correct connection of the transistors???
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.



When I'll know how to connect transistors, i'm going to make it with PCB.
 
jancooo said:
2) Multiplier is orange. Have you compared the color with 8,2k tranzistor (which has two red bands)?

Of course I did, but not all resistors use the same shades of color. (Look at the difference in the brown on the 100K and the brown on the 470, for example).

Speaking of brown... since the two visible bands are brown and (now I know) orange, it can't be a 100K resistor, so what is it? If it's much too low, it could cause the signal to be low and undistorted by passing too much negative feedback current to Q1. If it's 91K (white brown orange) it's probably close enough.

If the pinout posted by elementx is right, then he's correct that you've got the emitter and collector swapped. (His diagram is missing the arrowhead on the emitter.) Have you tested the transistors and measured the gain on them?

Reid
 
To me it looks like the 100k (brown, black, yellow) resistor is 10k (brown, black, orange), and the 33k resistor might be 2.2k. I could be wrong; as you say, red and orange can be hard to differentiate.

But we shouldn't be guessing this. Get youself a multimeter and measure! Doing electronics without one is like bird-watching without binoculars: possible, but very error prone.

Rune
 
pnp

the pnp ac188 is great gremanium transistor!!!!!!!!!!!!!! i have few of them
you just got the collector and emitter mix up
after that should work great
you only reverse the pins for pnp transistor
if had ac187 it would be way the schematic show
you might wana get some transistor sockets sometimes the heat form solder can damge them
 

Attachments

  • pinout.jpg
    pinout.jpg
    13.4 KB · Views: 113
transistor testing

or you could use a multimeter with a transistor test socket

just an example http://cgi.ebay.com/3-1-2-LCD-Multimeter-AC-DC-LCR-Frequency-hFE-diode_W0QQitemZ300052305536QQihZ020QQcategoryZ66991QQssPageNameZWDVWQQrdZ1QQcmdZViewItem?hash=item300052305536


it say transistor testing
i know thiers other way but i use
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.