Hey guys, i want to build a regular Cmoy amp and insert blinking LEDs in the system using a Tip 31 transistor. The problem is whenever i connect the LEDs i get loud 50hz(amp powered by 18v wall transformer) noise. Also i get massive massive distortion, to the point every genre of music sound the same. If i disconnect the transistor the amp works fine. I would really like to implement the LEDs and i have a feeling that it`s not that hard to fix, but i just can`t figure out. P.S. i tried a voltage divider to individually power the amp and the transistor+led`s but the voltage divider alone brings enormous distortion.
Here is i pic of my scheme!
https://dl.dropboxusercontent.com/u/45794359/trans.jpg
Here is i pic of my scheme!
https://dl.dropboxusercontent.com/u/45794359/trans.jpg
Yeah, it`s supposed to flash with the audio signal, i assume is amplitude on lower freqs. I tried adding a resistor right before the LED's V+ and it didn`t change anything. Now the thing is if i disconnect the transistor from the virtual ground, the LEDs will be fully on, and i`ll get noise on the channel that's connected to the transistor and no distortion. If i hook up the transistor to vGround(in order to flash to audio), i get distort and noise on both channels.
It would be easier to follow if it were drawn out conventionally. It looks like there is continuity on the drawing from the virtual ground point (junction of the 4K7's) "down and round an up" to the emitter of the transistor and then back to the negative of the battery... shorting it out. Hmmm...
Getting LED's to flash with music sounds so easy and in practice its not. The reason is that the transistor needs around 0.7 volts across base and emitter to conduct, and in headphone terms thats loud. Anything below that and the LED's are off. Running on batteries (battery life) is a major design consideration. You need ultra high efficiency LED's run on very low current.
Its all do-able but its quite a major design exercise to get it right, the circuitry would be far more complex than the CMOY itself.
Here's something along the lines of what you want, posts #25 and 30
http://www.diyaudio.com/forums/soli...t-guitar-preamp-led-mirror-2.html#post3512623
Getting LED's to flash with music sounds so easy and in practice its not. The reason is that the transistor needs around 0.7 volts across base and emitter to conduct, and in headphone terms thats loud. Anything below that and the LED's are off. Running on batteries (battery life) is a major design consideration. You need ultra high efficiency LED's run on very low current.
Its all do-able but its quite a major design exercise to get it right, the circuitry would be far more complex than the CMOY itself.
Here's something along the lines of what you want, posts #25 and 30
http://www.diyaudio.com/forums/soli...t-guitar-preamp-led-mirror-2.html#post3512623
Look at your transistor, brown and black should not be connected together. Disconnect either one based on your calculation of the LED voltage (or current or however you did it).
Add: And one more thing... have you considered the bias voltage of the transistor i.e. the DC voltage that will appear on the base or gate of the transistor and will be fed to the sound channel input. Especially for BJT which has the gate voltage pretty much fixed to emitter + Vbe (around 0.5~0.7V).
Add2: Yea I said DC, which is usually blocked by the capacitor. However DC is never fully DC and will have a 50Hz component from the transformer due to imperfect filtering and coupling/ground/etc, and this is usually the source of a 50Hz.
Add3: Either way, make the bias circuit for the transistor, there should be a DC-blocking capacitor between the transistor and the input. Because feeding the bias voltage back to the source isn't a good idea.
Frankly I think you should get the VU meter IC and redo your circuit.
Add: And one more thing... have you considered the bias voltage of the transistor i.e. the DC voltage that will appear on the base or gate of the transistor and will be fed to the sound channel input. Especially for BJT which has the gate voltage pretty much fixed to emitter + Vbe (around 0.5~0.7V).
Add2: Yea I said DC, which is usually blocked by the capacitor. However DC is never fully DC and will have a 50Hz component from the transformer due to imperfect filtering and coupling/ground/etc, and this is usually the source of a 50Hz.
Add3: Either way, make the bias circuit for the transistor, there should be a DC-blocking capacitor between the transistor and the input. Because feeding the bias voltage back to the source isn't a good idea.
Frankly I think you should get the VU meter IC and redo your circuit.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.