Still have noise.

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've got a BrianGT NI3875 that is very minimalist and I have been having some noise issues. (FM, I can hear the announcers with my ear up to the 94db sensitive speaker) I asked about the same problems a few months ago, and a few potential solutions were presented. I didn't do much at the time, as I found moving stuff around helped.

As suggested I replaced the 1M law faking pot with a 25k log (also PEC carbon, now available from digikey.) The pot didn't make the slightest difference (Thorsten was right, the 1M pot was fine, but the 25K log does track better at low volumes.) He was also right, in that the problem is coming from my speaker cables.

Their position dramatically affects the noise level. However, it is a big pain to always find the right spot. Even then it isn't entirely quiet. I tried a zobel at both ends of the cable. It didn't do anything but sound worse. I've also tried some higher inductance cables (though not that high,) and they help some. Do I have any options beyond adding an inductor at the outputs? If I do, what type and value are best?

I've left the grounding scheme alone, though I will change it to Carlos' if it might help. I also haven't added RF filters at the board inputs, though since the problem comes from the outputs I don't see how it would help.

thanks for any advice,

Paul
 
neutron7 said:
how can speakercables pick up radio? does it have something to do with the feedback/ground in the amp, because i dont see how a normal radio signal could drive an 8ohm load without any amplification.

i cant possibly answer that, but ive seen it happen before. after months of trail and error, the customer unhooked their coax from their tv and the system was silent...
 
I don't have any TV coming into the system. I'll try to find a cheater plug and lift the ground, but I don't see how moving the speaker cables would affect the noise so much if it were caused by a ground problem.

I think it is being amplified through the feedback loop. Ferrites are the obvious solution, but everyone says they sound horrible unless perfectly matched to the cable. I don't have any around to try. Any idea how to pick the right ferrite?

thanks for the suggestions,

Paul
 
Ex-Moderator
Joined 2002
I hope that doesn't mean you're going to lift a safety earth? That is a very bad idea.

As for the ferrites, I used to live about 1/2 a mile from the main TV broadcast tower for SE England at Crystal Palace, and had real problems with RF breakthrough. I used cheap clip on ones, and noticed no sound degradation at all. Who is "Everyone"? ;)
 
wasnt there a thread on this forum somewhere where a guy's speaker cables were acting like antennas and picking up tv and radio signals? he ended up using ferrites and it stopped it?

i think it was in the chip amp or class d forum. i remember it being like 10 pages or so and they guy couldnt figure out what to do. its probably in the class d forum, because i dont typically read stuff in the chip amp forum.
 
It looks like the amplifier IC is oscillating. Note that oscillation is a requirement in order to pick up AM radio.

It may happen due to poor PCB layout, poor input filtering, too much capacitive loading of the speaker output, poor wiring where signal wires go together with speaker or power wires, etc...
 
Pink Mouse,

Why is lifting the Earth Ground a big deal? I bought a descent subwoofer and it had a terrible hum that I couldnt get to go away. Finally, I called Martin Logan direct and the suggested the cheap plug converter to lift the saftey ground and bada bing...the hum was gone.

He said it wouldnt cause any problems, but its a $3000 sub, so if you have some inside knowledge... I want in!LOL

Dominick
 
it happens alot

when i used ti install PA systems, it seemed like every church i went to in downtown indy would have radio coming through from one place or another. altough it usually came in through mic runs going through the attic or walls. the ferrites did the trick most of the time. although we never left that as a perm. solution. we always got some cable made by kanari (sp?). heavy braided shield that was a $%#@* to unbraid to get a lead for soldering. then each of the 2 conductors were twisted pair. so for a three conductor mic line there was the shield and 4 conductors in the cable. the theory was that all of the shielding helps keep things from getting to the conductors and the 2 wires for each conductor twisted around each other (and insulated individually) canceled what did get ine. kinda makes since but kinda doesn't. but it worked every time.
 
Do you live in an older home, say from the 1950's or so? I did at one time and the earth grounds (3rd prong) were not hooked up at all. I had to have an electrician come and take care of the issue to help protect my computer equipment. Just a thought. Also what someone else said about the cable tv is very true, even if not hooked up to your system a faulty cable/connector on CATV can cause it to act like a transmitter. Definately not a good thing to have in the house.

Just to clarify, the house had several 3-prong outlets and they did have all three wires attached. It is just that the 3rd wire was not connected to where it needed to be... ie earth ground.
 
I'm going to order some assorted ferrites and see what happens.

The amp isn't oscillating. The noise is FM not AM. The load is easy as can be. If it were oscillating the chip would probably run warm if not hot, and a zobel would help.

If the chassis became hot (loose wire or something?) then a lifted ground would be an issue. However, if that were a common possibility, surely the UL would require that all components have safety grounds. My 2yr old disc player doesn't have a safety ground, and and I fail to see how that is different. If it were grounded, then I'd have a loop. I don't mean to be argumentative, but I don't see why putting a cheater plug on an amp is a big deal. I have my sub amp ground lifted (sawed the ground prong off the cable) otherwise it hums.

Thanks for the advice. It'll be a week before the ferrites arrive, but if anybody thinks of something in the mean time. . . .

Paul
 
Duelbox,
Sorry, you posted while I was working on mine. My home is from circa 1915. The wiring is a mix of new and very old. The ground does work on the circuit I'm using. The breaker box is grounded to the cold water inlet. That's not entirely safe. I should probably speak to the landlord about that. There is an unhooked cable right next to my stereo. The far cable picks up more noise, so I assumed that wasn't the culprit, but I'll pull the cable down into the basement. The cable (disconnected) enters the house below the systems power outlet. That can't help.

thanks for the thoughts,

Paul
 
I disconnected the cable at the splitter. I still had the same amount of noise, so I pulled out my laptop to post. Then I realized I have cable internet. Doh!

I think that this noise is also the result of listening to very sensitive speakers at 10ft in a quiet house. I can't hear the noise from my chair unless the furnace and fridge are off, there is no traffic, etc. It still annoys me to death. Perhaps I need to learn to ignore it.


pj
 
Brian's pcb is pretty debugged by now but it doesn't cover all applications and situations.

The LM3875 is very quiet, around 2 uV input noise which means a S/N of over 110 dB, dead silent in other words.

1 Remove the input connection and short the input to ground, preferable at the pcb.

2 Connected a speaker with a short cable

3 Power up one channel at the time

Noise now?

Do you have a picture over your setup?

Have you decoupled the IC properly?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.