strange amplifier behavior, help needed

[zionz]

Hi all, im a newbie in electronics, i have been making some amplfiers, mainly using integrated circuits, but now im trying with transistors. Im currently building (trying to) an esp p3 ( http://sound.westhost.com/project03.htm ) , the problem is very weird (at least for me), when i start the amplifier i hear some audio at low volume and distorted, then sometimes (apparntly random) the volume rises very high and it sounds good (the distortion seems to dissapiar) but then in some seconds it start degrading (volume goes down slowly and distortion rises again) until the audio fully dissapears. This sequence repeats. I also noticed that when i touch the BD139 driver (the one connected to the bc559) causes some kind of effect, sometimes the sound come back by doing this.

Sorry for bothering but i have tryed many things with no success and im getting mad of it , any suggestions would be very appreciated, thanks you in advance.

ps: sorry for my bad english, i hope you could understand me.

[EchoWars]

Did you use Rod's board? Or make your own?

Sounds like a bad soldering job or a wiring error.

[dhaen]

Hi,

What you are describing is the amplifier moving in and out of it's operating point. I think you have an open-circuit, perhaps a bad solder-joint or forgotten connection.

Almost all transistor amplifiers rely on DC feedback to stabilize the operating point. In this amp the path is from the centre of the output transistors through the 22k to the right-side transistor of the LTP (Long Tail Pair or Differential Amplifier). You can see that this provides a little negarive feedback at AC (so the amplifier has gain at AC), but a lot of negative feedbcak at DC for DC stability.

Your English is fine

[zionz]

Hello, thanks you very much for your replies, i made the amp in a veroboard and it looks a bit ugly , maybe there's some wiring error, i will recheck the solders and connections again, i may be missing something. Another thing i noticed is that only the 2n3055 power transistor gets warm but i have tested (in a multimeter) both 2n3055 and mj2955 and they seem to work ok.

[janneman]

What is the output DC voltage? The fact that one transistor is hot points to an output stuck to a supply (the pos one in this case).
Can you measure the DC across the Re's of the output pair to see what the bias is?
What is the DC at each of the base's of the input pair?
Forget about the music for now, get it in order first.

Jan Didden

[dhaen]

You are right in thinking that the output transistors should be equally warm. What this means is that the voltage at the centre of the output transistors is not zero volts, so one device has more voltage across it. I suggest you measure the voltage on the output. It should be 0v.

By the way, it's best not to connect a speaker directly at this stage, as damage could occur to the speaker and amplifier. If you need to listen, for fault finding, wire a capacitor in series with the speaker. Anything from 100uF 35v would be fine. Best make it bi-polar by using 2 in series wired like this:
amp______ +CAP -______- CAP+______speaker

Veroboard does look messy, but once the it's working and the lid is on, it'll seem fine

One last thought: You have got both +35v and -35v rails connected?

[zionz]

hi again, after checking the circuit i found that the 2k2 resistor connected to the v+, the 100uf cap and the another 2k2 resistor was not making good contact, after fixing that i tryed again.I connected the amp using a 18+18 power supply and wihout audio input and measured the ouput dc that was +18v ! (ground to ouput), then i connected audio (but no output speaker) signal and the 220R resistor at V- started smoking . the driver transistors are now dead , im going to get some new ones to try again. And thanks you all for all the help.

[janneman]

... and when you try again, PLEASE don't connect any input signal untill you fixed it. What's the point?? You are only going to break more parts this way.

Your output was +18V and the lower transistor was warm. Somehow the top transistor therefor was either shorted or fully on. You need to measure the V across the Re's of the output and the other things as suggested and things like the Vb-e of the top transistor. We really want to help but that gets difficult if things continue to break.

Jan Didden

[zionz]

Hi again, i finally got the new components, i have replaced everything in the circuit(all the transistors and burned resistors), now how should i test it without risk of destroying it again?
Another thing, what could happen if i turn it on without the output power transistors connected to the drivers? (im sorry for my ignorance ) thnx in advance.

[Morello]

Always use small secondary fuses when testing amps. Try 250 mA.