ionomolo said:
Whether or not, just another good reason for good filtering. I'd much rather over-filter it with coils and caps then have the amp compensate instead.
Most IC amps need a HF filter cap right at the power pins to prevent oscillation. Descrete amps you can get away with the HF filter cap further away.
Supposedly a descrete amp has less PSRR than most IC's.
Thanks for the replies guys, didn't notice I had more than 1 (and I'm still under forum moderation hence the delay in replying)
In that thread, carlosfm suggests "A 300~330pf cap across the inputs of the LM3886 will solve the problem" - which inputs exactly? Sorry I'm a bit of a noob, I just went and put together the chipamp.com LM3886 kit according to the (very simple) instructions without paying much attention to the circuitry.
I just don't want to do anything stupid like put it across the signal input when it's supposed to be across the power input or vice versa. Also is what jaycee suggesting the same thing?
I haven't tried the amp in a different part of the house, but it happens even with light switches on a different floor and literally at the other end of the house. I find it interesting since lower floor lights, upper floor lights, and powerpoints are all on different fuses.
The amp is inside a fully-enclosed aluminium case which is grounded to earth, and connected via a powerboard which supposedly has RFI filtering.
Peter Daniel said:
jaycee said:
In that thread, carlosfm suggests "A 300~330pf cap across the inputs of the LM3886 will solve the problem" - which inputs exactly? Sorry I'm a bit of a noob, I just went and put together the chipamp.com LM3886 kit according to the (very simple) instructions without paying much attention to the circuitry.
I just don't want to do anything stupid like put it across the signal input when it's supposed to be across the power input or vice versa. Also is what jaycee suggesting the same thing?
ashok said:
I haven't tried the amp in a different part of the house, but it happens even with light switches on a different floor and literally at the other end of the house. I find it interesting since lower floor lights, upper floor lights, and powerpoints are all on different fuses.
The amp is inside a fully-enclosed aluminium case which is grounded to earth, and connected via a powerboard which supposedly has RFI filtering.
ok just soldered one of these filters in-line with the power going to my gainclone; unfortunately it didn't help the popping at all 
is someone able to answer my question from my previous post - which are the 'inputs' on the lm3886 that i need to solder the cap across?
is someone able to answer my question from my previous post - which are the 'inputs' on the lm3886 that i need to solder the cap across?
You need to do a bit more investigation to see how the noise is getting in.
I would be putting a scope on the power lines and then switch the light on and off to see if it is on the power lines.
If so then you might need to add some 100nf across the power lines or across the transformer outputs.
I use a filtered mains socket and my fridge switching on and off still gets through but not quite so bad.
If your amp is in a metal box and earthed then I think the noise has got to be coming down the mains.
Unless you have a seprate pre amp and wires outside the box picking up the radiated emissions.
I would be putting a scope on the power lines and then switch the light on and off to see if it is on the power lines.
If so then you might need to add some 100nf across the power lines or across the transformer outputs.
I use a filtered mains socket and my fridge switching on and off still gets through but not quite so bad.
If your amp is in a metal box and earthed then I think the noise has got to be coming down the mains.
Unless you have a seprate pre amp and wires outside the box picking up the radiated emissions.
the amp case is earthed, yep.
i think adding any other capacitors is not my preferred choice of action at the moment, since i've already added so many! here's a pic of my powerboard:
it's a fully-grounded metal unit with quite a few caps and stuff in the top end (couldn't see the parts clearly.) there are 3 caps and an inductor (i think) in the filter i just added so who knows how many caps i have now!
i thought just then that perhaps there's too much filtration going on (if there is such a thing) so i just plugged the amp straight into the wall... pop is still there.
what makes it easy to debug is that the RCA inputs can be disconnected and the pop is still there... only power and speaker need to be connected.
a quirk with having such large caps (10,000uf) is that my amp plays music for a good 4 seconds after the mains power is cut. so just then i flipped the mains switch, and ran to turn off the light in my room before those 4 seconds were up and guess what... it popped again!!! i think i'm getting close.
the switch that i flipped cuts both the live and neutral supplies to the power supply board, so the only electrical connection between the amp and the house is... the earth. (had i done this simple test earlier i wouldn't have bought the mains filter... argh)
i think adding any other capacitors is not my preferred choice of action at the moment, since i've already added so many! here's a pic of my powerboard:
An externally hosted image should be here but it was not working when we last tested it.
it's a fully-grounded metal unit with quite a few caps and stuff in the top end (couldn't see the parts clearly.) there are 3 caps and an inductor (i think) in the filter i just added so who knows how many caps i have now!
i thought just then that perhaps there's too much filtration going on (if there is such a thing) so i just plugged the amp straight into the wall... pop is still there.
what makes it easy to debug is that the RCA inputs can be disconnected and the pop is still there... only power and speaker need to be connected.
a quirk with having such large caps (10,000uf) is that my amp plays music for a good 4 seconds after the mains power is cut. so just then i flipped the mains switch, and ran to turn off the light in my room before those 4 seconds were up and guess what... it popped again!!! i think i'm getting close.
the switch that i flipped cuts both the live and neutral supplies to the power supply board, so the only electrical connection between the amp and the house is... the earth. (had i done this simple test earlier i wouldn't have bought the mains filter... argh)
chylld said:
Yes, pretty much. You want the capacitor across the inverting and noninverting inputs of the LM3886, that's pin 9 and 10.
It could also work if you solder 470pF across R2 (22K). From looking at the design of the chipamp.com board this might be the simpler option. This effectively makes a low pass filter that rolls off -3dB at about 320kHz
OMG i can't believe it... it worked!!! i soldered a 470pF cap across each R2, not the best soldering job i've done but good enough... tried 10 times in a row to recreate the pop and it's gone!!
i can't thank you more than enough.
actually if i put my ear up to the speaker and turn off the light switch real slowly, then i can hear a quiet 'bzzt' but i wouldn't class that as audible from a sensible listening position. turning off the light switch normally now has no pop at all.
i can't thank you more than enough.
actually if i put my ear up to the speaker and turn off the light switch real slowly, then i can hear a quiet 'bzzt' but i wouldn't class that as audible from a sensible listening position. turning off the light switch normally now has no pop at all.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.