audio amp.

[herm0016]

i built a 20watt per channel audio amp. kit tonight. lots of fun. got it all wired up and such. tested it.. the left channel has a buz that seems to be related to the fan on the atx 250w powersupply i am using. you can here the difference in the rpm of the fan when you impede the airflow. the right channel is absolutly perfect. i have this kit: http://www.chaneyelectronics.com/products/amplifiers/c6442.htm i dont really understand how one channel can have problems, while the other is clear unless my ic amp is bad? not sure... i am using an old enlight 250 watt atx powersupply, eventually i am going to mount this in my current rig so i can have speakers connected drictly to the computer, but still have them amplafied. all in my case. help! ill have pics of the schmatics up tomorrow.

 

[bobsmith1492]

Did you check your solder connections? A loose connection could cause high resistance allowing supply noise to enter on the input and be amplified to the speaker. Check the input lines, especially, on that channel.

The other thing could be that, by the looks of the layout of that amp, it is getting supply noise on the channel because of the way the board is layed out. Is the buzzing channel the one furthest away from the incoming power line? Try putting a decoupling cap on the supply lines right under the noisy amp chip.

 

[herm0016]

http://i14.photobucket.com/albums/a317/ajherman/Pictures013.jpg

http://i14.photobucket.com/albums/a317/ajherman/Pictures012.jpg

http://i14.photobucket.com/albums/a317/ajherman/Pictures.jpg


here are pics of the board, mask and schematic. im not well versed in elctronics design, but can build anything i have a schematic for. if someone wants to show me where i need to put the decupling cap i would be greatly apprecitive. i think i have an idea.. but dont want to hurt anything. thanks.

 

[bobsmith1492]

You need to put it between the +V input and the GND input (the power inputs). Follow the traces up to the chip (the ground is probably the big, fat trace) and try to put a capacitor right there under the chip on the bottom of the board. Something like a 0.1 uF and/or a 10uF capacitor (rated for about 1.5x your input voltage) would be good.

Also, did you check for bad connections (cold solder joints)? I couldn't really tell from that picture. If the solder is really dull and nasty looking or is balled up away from the board, it is a bad connection.

 

[herm0016]

my soldering is good. i know that much. and i will get a cap to try tomorrow. otherwise i have tried with my actual computer and the interference is just crazy bad, but thats what you get with a switching powersupply. i tried just puting a cap across the fan leads also, it did help, but not much. i am going to buy a good transfromer powersupply to power this project. it means another cord, but it will make the audio a lot better. probably still ad caps to the powersupply fan and the board. mabye a trio. a tantalum, and elctolitic and a mylar. each has diffrent rf/em charcterestics.

 

[bobsmith1492]

A decent switching supply shouldn't cause much noise; switching frequencies are usually far above anything you could actually hear. Do you have access to a scope to check what frequency/frequencies the noise is at? Also, what makes you think it is the fan from the supply causing the noise?

 

[herm0016]

i know part of it is the fan because when you change the rpm of the fan the noise changes also, and when you stop the fan teh noise stops. i think i will also try addeing inductors to the fan leads to filter out the high frequeces. im gonna be backpacking for a week. back next sunday, i will continue working on it then and bump this when i have more info.