Hello Guys! After wasting several hours trying to fix my issue, I decided to try asking gor your help. Sorry for long post, but I like to explain my issues thoroughly.
I've built a lm3886 gainclone recently, based on schematic from BrianGT's guide, with a bit simpler psu. Everything works fine, even gives next-to-nothing noise and sounds very well.
I realized that sound gets improved with addition of a proper preamp, so i decided to put something simple but good enough in one case with gainclone so i would get great amplifier with preamp in one box.
For starters, I took out a board with a headphone amp I built years ago but couldn't get to work, even managed to fix it, but I've got a few issues with it:
- *the noise that comes out when you have nothing plugged in and raise the volume* - it was audible even with a source plugged in and playing, but i managed to fix it after i sorted out the grounding and put the board in metal case
- another noise, sounds similar to pink noise. It was louder at the beginning, but after i fixed the issue above it moved to left channel and is present only a little in right channel. It's not very loud(but louder than is should be), and it's barely hearable when playing music, but it's still an issue.
I thought that something was wrong with the preamp board(it wasn't working at the beginning so that was the first thought), but decided to try powering it with two 12v batteries. And it got almost totally quiet.
So the issue must be with psu. It should be fine in theory, but it is not, so I'm asking you for help.
Here's the psu:
there are two parts, one on the left, which powers up gainclone, and the second board which powers preamp. Initially I made it with lm317 only, but with the noise present, i thought maybe going down from high 37v to 15v was causing it, so i added 7812/7912, but not much changed. I tried raising lm317 voltage, so it got 37v->18v->12v, still nothing. I even connected whole psu tomiddle tap in transformer(its ~27v with middle that gives about ~17v) to check if lower voltage will cause less noise, but it stayed the same. I'm running out of ideas to solve it.
What do you think?
I've built a lm3886 gainclone recently, based on schematic from BrianGT's guide, with a bit simpler psu. Everything works fine, even gives next-to-nothing noise and sounds very well.
I realized that sound gets improved with addition of a proper preamp, so i decided to put something simple but good enough in one case with gainclone so i would get great amplifier with preamp in one box.
For starters, I took out a board with a headphone amp I built years ago but couldn't get to work, even managed to fix it, but I've got a few issues with it:
- *the noise that comes out when you have nothing plugged in and raise the volume* - it was audible even with a source plugged in and playing, but i managed to fix it after i sorted out the grounding and put the board in metal case
- another noise, sounds similar to pink noise. It was louder at the beginning, but after i fixed the issue above it moved to left channel and is present only a little in right channel. It's not very loud(but louder than is should be), and it's barely hearable when playing music, but it's still an issue.
I thought that something was wrong with the preamp board(it wasn't working at the beginning so that was the first thought), but decided to try powering it with two 12v batteries. And it got almost totally quiet.
So the issue must be with psu. It should be fine in theory, but it is not, so I'm asking you for help.
Here's the psu:
there are two parts, one on the left, which powers up gainclone, and the second board which powers preamp. Initially I made it with lm317 only, but with the noise present, i thought maybe going down from high 37v to 15v was causing it, so i added 7812/7912, but not much changed. I tried raising lm317 voltage, so it got 37v->18v->12v, still nothing. I even connected whole psu tomiddle tap in transformer(its ~27v with middle that gives about ~17v) to check if lower voltage will cause less noise, but it stayed the same. I'm running out of ideas to solve it.
What do you think?
Possibly ground loops, something you cannot see in a schematic but only in how you wire it all together.
You should have protection diodes in the regulator cct. When the power section pulls down the supply to the regulator, the charge in the 1000uF caps will discharge back through the regulator(s). This is not good. Read data sheets.
1000uF seems a bit overkill in a regulator cct.
data sheets for these types don't suggest values this high.
Instead of a 12 regulator, calculate (or measure) a resistor to drop a couple of volts and isolate the regulators from the 1000uF capacitive load. If the noise is coming from the LM3x7 regulators this will help isolate it as a R-C filter.
You should have protection diodes in the regulator cct. When the power section pulls down the supply to the regulator, the charge in the 1000uF caps will discharge back through the regulator(s). This is not good. Read data sheets.
1000uF seems a bit overkill in a regulator cct.
data sheets for these types don't suggest values this high.
Instead of a 12 regulator, calculate (or measure) a resistor to drop a couple of volts and isolate the regulators from the 1000uF capacitive load. If the noise is coming from the LM3x7 regulators this will help isolate it as a R-C filter.
One recommendation is to shunt R2 and R3 with an electrolytic like 47uF. But that's probably not going to fix it.
To which gropund is your speaker return connected? Is it connected directly at GND_ between the main supply caps or to the GND_ at the far right of the circuit, or someting in between?
jan didden
To which gropund is your speaker return connected? Is it connected directly at GND_ between the main supply caps or to the GND_ at the far right of the circuit, or someting in between?
jan didden
Thanks for reply!
I used 1000uf because i've got too much of them laying around, and i go by the rule "a bit too much uF won't hurt".. i guess i was wrong
rectifier goes directly between lm3x7 in and out? i think i've seen that before. gotta try. Voltage was intended to be 15v, but i tried 12v regulators out of desperation, and raised lm3xx voltag to give them some space.. i guess i'll just go back to LM only.
About the ground: from the GND on schematic one wire go to one GC followed by another, and second wire go to preamp psu, which then go to preamp, and further to pot and input connector. should I make a wire from every gnd point to one in the chassis or somewhere?
I used 1000uf because i've got too much of them laying around, and i go by the rule "a bit too much uF won't hurt".. i guess i was wrong
rectifier goes directly between lm3x7 in and out? i think i've seen that before. gotta try. Voltage was intended to be 15v, but i tried 12v regulators out of desperation, and raised lm3xx voltag to give them some space.. i guess i'll just go back to LM only.
About the ground: from the GND on schematic one wire go to one GC followed by another, and second wire go to preamp psu, which then go to preamp, and further to pot and input connector. should I make a wire from every gnd point to one in the chassis or somewhere?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.