Update -Source of buzz
Moving components back 1 by 1 has worked so far. The chain begining with digital out from SACD Player to Receiver, then pre-amp out from there to LM3875, and speakers..... sounds pretty darn good, no buzz sound.
When I connect the Coaxial line from the cable-company to the DVR or to my LCD-TV the buzz comes back. Note that none of these other equipments are even connected to the amp or receiver or mains, they just happen to be with-in a few feet of amps,...... but connect the antenna and it buzzes in the speaker.
I wondered if maybe the sheathe of the coaxial cable line wasn't grounded well enough. Grabbing a wire I held one end to the outside of the screw on connector at the wall with my finger, while touching the other end of the test lead to the nearest earth point I could find... the chassis of LM3875. This in fact made the buzz go away, but I don't think that's the right way to do it.
I went outside to make sure that earth was connected to the install point of CATV on the outside of my house. There is a big solid wire that connects from it to a metal-pipe that has been hammered into the ground decades ago. I had a big 12AWG solid copper wire, I stripped a good six inches off the end and stuck it a good 8 or 10 inches down into the topsoil right outside my house, connected the other end right to the cable jack outside the house (just to make sure LOL)... no change, still a buzz.
My options seem to be no TV (which the wife won't like), or ground the chassis of LM3875 to the sheathe of my cable TV antenna. The ladder just seems wrong to me; as if I am treating a symptom rather than a problem.
Any ideas?
Moving components back 1 by 1 has worked so far. The chain begining with digital out from SACD Player to Receiver, then pre-amp out from there to LM3875, and speakers..... sounds pretty darn good, no buzz sound.
When I connect the Coaxial line from the cable-company to the DVR or to my LCD-TV the buzz comes back. Note that none of these other equipments are even connected to the amp or receiver or mains, they just happen to be with-in a few feet of amps,...... but connect the antenna and it buzzes in the speaker.
I wondered if maybe the sheathe of the coaxial cable line wasn't grounded well enough. Grabbing a wire I held one end to the outside of the screw on connector at the wall with my finger, while touching the other end of the test lead to the nearest earth point I could find... the chassis of LM3875. This in fact made the buzz go away, but I don't think that's the right way to do it.
I went outside to make sure that earth was connected to the install point of CATV on the outside of my house. There is a big solid wire that connects from it to a metal-pipe that has been hammered into the ground decades ago. I had a big 12AWG solid copper wire, I stripped a good six inches off the end and stuck it a good 8 or 10 inches down into the topsoil right outside my house, connected the other end right to the cable jack outside the house (just to make sure LOL)... no change, still a buzz.
My options seem to be no TV (which the wife won't like), or ground the chassis of LM3875 to the sheathe of my cable TV antenna. The ladder just seems wrong to me; as if I am treating a symptom rather than a problem.
Any ideas?
AlexQS, glad you have had some success. Pay attention to the connectors on your cable TV wires. The "crimp" type ends are notoriously loud even when terminated properly. The leakage can be detected hundreds of feet (and more) away from the faulty connection.
Use "compression" style only to eliminate the problem. Here's a good link:
Coax Compression RG-6 Connector Termination.mpg - YouTube
Also - if you have junction boxes (splitters) or cable signal amplifiers, make sure they are approved by your specific cable provider. Usually they will provide that equipment at no cost.
Use "compression" style only to eliminate the problem. Here's a good link:
Coax Compression RG-6 Connector Termination.mpg - YouTube
Also - if you have junction boxes (splitters) or cable signal amplifiers, make sure they are approved by your specific cable provider. Usually they will provide that equipment at no cost.
Last edited:
Thanks Bob, good video clip.
All coax I have inside and outside is compression type, provided by cable co. No signal amp. Only junction is a 3-way splitter, also provided by, installed and earthed by cable company. It seems to be the LCD-TV mounted on the wall 3 feet above the amps that leaks rf. I just found out that with coax-cable connected to DVR, disconnecting HDMI from DVR to TV also makes buzz sound disappear.
I just really wasn't planning to seperate tv watching and music listening... I'd have to add a room onto my house to do that, LOL
Thanks again for your idea this morning, it really set me on the right path to discover the source of the problem.
In some of my applications I have had to sacrifice the convince of HDMI and use the combination of RGB cable (video) and SPDIF or optical (sound). All depends on the connections on your equipment.
And no - we are not going to start a group buy for your new room.😀
And no - we are not going to start a group buy for your new room.😀
This seems a good solution that worked, I have TV, and Sound, without the buzz.
SPDIF also buzzed, but I found that optical cable for TV sound is a non-buzzing solution here. (I never watch tv anyway!)... but the wife is happy, -correction We're all happy!!
Thanks Bob!! 😛
This is an excellent result Alex. You have persisted and got a result by actively experimenting with the location of the source of interferance.
It may be beneficial to continue your quest a little further now to locate and remove the grounding problem - if indeed there is one. I can't give you any help here as I am too, 'struggling' with the issue.
I can offer a little anecdote that may help with your interferance problem (if this is the issue):- recently I built an amp and it sounded great, until I housed it, when it began to hum, and I still dont fully understand why. But in an attempt to solve it, I kept poking around the circuit with a ground wire that was connected to the central grounding point. Of course I had to be very careful about where I poked that wire, but I was targeting every point of ground in the circuit including the chassis and wiring.
Even though I had gone to a lot of trouble to isolate the rca line-in from the chassis as per the 'instructions' I found that when I connected my pokie wire to the rca input gnd it killed the interferance. The input was already grounded 10cm away from the rca, but, none-the-less, that ten centimeteres of wire was enough to cause problems. Furthermore the pokie wire was about 30cm long, but it did the job nicely.
I just wanted to add one more piece of information for your next build, its all about saving money so please accept it in the spirit in which it is offered - a handy hint. Initially the only tool you will need is a scientific calculator of reasonable quality - like the ones you buy from K-Mart or an Office supplies shop for about $20Au (not sure about your currency).
OK; Here is a good calculation to learn and if you already know it I will apologise for going into detail.
For any amp that uses a sinusoidal signal (which is a good starting point but not neccessarily the same as the music-signal), the peak voltage (we will call it Vp) is equal = to the square root of (2 x the required power [P] times the load resistance [Rl])
For example; if we were designing an amp for the lm3886 chip and an 8 Ohm speaker the calculation would be;
Vp = Sqrt of (2 x 50 x 8) = 28VAC so therefore we would use a 56V centre-tapped or dual 28V transformer. This is the absolute maximum according to the datasheet and we would much better to run the amp at 90% or 45W into 8r. So a common 25-0-25 transformer would be ideal.
But how many amps of current do we need for that amplifier? Well, here is another sum that can help us:
The peak current (we will call it 'Ip') is equal to = the square root of [(2 x the required power 'P') divided by / (the load resistance, Rl)] or put another way; Ip=sqrt of [(2xP)/Rl] an example of this equation for the lm3886 circuit mentioned above would be;
Ip=sqrt of [(2x50)/8]=3.5A
So when I build my lm3886 amp I will need a 25-0-25 Volt transformer of 3.5A capacity.
So if I build this amp I will use a 160VA 50VCT transformer (or larger) but this is the point that I am trying to make;
I cant get blood from a stone, nor can I draw more power from an oversized transformer.
I like to be frugal and therefore I will use a 160VA transformer, or if I have a 200VA on the shelf, I will choose that one;- so that I have a little overhead and so that the transformer too can operate in some cool comfort.
I hope this has been of some use to readers including Andrew who asked for clarification.
Cheers and bye for now.
While prepping these, I used some super-fine grit sandpaper, got all the sharp edges off the corners, the burs from drill holes, etc and polished them up.
Of course this made them a mess of black aluminum dust all over. I washed them off with dish soap, and rinsed them well with hot water. The water I know is 120 degrees f. Of course the heatsinks got quite warm from being so long in the hot water, and stayed warm several minutes after they were dried. (Not quite as hot as something in the dishwasher though... It heats water to 140f before beginning its cycle.)
After listening as loud as I would ever want to, for a couple hours they got about as warm as they did when I cleaned them. I could hold my hand on them indefinitely at this temp, which I guess is under 120 f. I think I read that if I can't hold my hand on them for more than 10 seconds they are too hot. I don't have any infrared to read their surface temp, but I cook a lot, touch hot things etc, I think I have good instincts here.
I need to learn to read the data sheet to be more sure of specs for heat. Stepping me through those equations is helpful for me to begin learning more.
If I were to do it again I would either go up to 1 inch thick, or make them longer. I'd like them to stay supercool all the time.
So I'm glad that I did not need to resort to grounding the amp to the cable. That would've been a real hak way to solve problem in my opinion.
At this point I think it would be more of a shield problem than a ground problem- maybe those problems go hand in hand?
Glad it works now. 🙂
Far,
these were the targets I am asking you to define.
Alex, nice to hear that the problem has been identified.
Another solutions may be fitting an isolator in the TV cable feed. I'm not sure if it should go before or after the first cable box. Try both.
Alex, I have also been trying to use "bulk" metal as heat sinks in an integrated MyRef build. The frame of the chassis has two 17" x 3" x 3" x 1/4" aluminum angle components connected with aluminum "U" channel runners.
During the winter months that configuration worked just fine. Yesterday was my first "open the windows and crank-it-up" day as the temps were around 80F outside. After about four hours of loud and obtrusive "Big Band" music, I could hear some slight dropouts that sounded like the LM3886 temp protection circuit was about to trigger. My heat gun read ~ 105F on the LMs and at the metal between the chips.
I now believe the bulk approach is not enough and plan to attach some finned heat-sinks to the chassis to add some convection air flow as an assist. Something like this:
It's seems just a matter of surface area that's exposed to cool air. The back of the chassis also has a coat of primer that will probably need to be completely removed to eliminate the insulating properties of the paint.
Of course a small DC fan could be added but I don't think that should be necessary on any "chipamp".
Hope you don't box yourself in with the HS approach you are using.
During the winter months that configuration worked just fine. Yesterday was my first "open the windows and crank-it-up" day as the temps were around 80F outside. After about four hours of loud and obtrusive "Big Band" music, I could hear some slight dropouts that sounded like the LM3886 temp protection circuit was about to trigger. My heat gun read ~ 105F on the LMs and at the metal between the chips.
I now believe the bulk approach is not enough and plan to attach some finned heat-sinks to the chassis to add some convection air flow as an assist. Something like this:
It's seems just a matter of surface area that's exposed to cool air. The back of the chassis also has a coat of primer that will probably need to be completely removed to eliminate the insulating properties of the paint.
Of course a small DC fan could be added but I don't think that should be necessary on any "chipamp".
Hope you don't box yourself in with the HS approach you are using.
Last edited:
That little heatsink will help.
But I doubt it gets anywhere near my recommendation to use double what National suggest in the datasheet graph.
But I doubt it gets anywhere near my recommendation to use double what National suggest in the datasheet graph.
I searched all over the place for a deal on heatsinks, and finally I checked out heatsink usa. They are really reasonable. If I remember correctly I paid $28 for two 6" x 6" heatsinks shipped. That was much cheaper than any I found on ebay or even the metal scrap yard!
Hi Andrew,
That was just a small piece from another project used to show the location. I'll be fabricating from the full unit that's 10" x 3" x 1 5/16". It also has a 1/4" web that will double the thickness in addition to the fins.
I've been very happy with this supplier.
10.000" - HeatsinkUSA, LLC Store
That was just a small piece from another project used to show the location. I'll be fabricating from the full unit that's 10" x 3" x 1 5/16". It also has a 1/4" web that will double the thickness in addition to the fins.
I've been very happy with this supplier.
10.000" - HeatsinkUSA, LLC Store
Just a note for this type of application: The back-plate will be removed and stripped of connectors. Once the shape and size is finalized, the plate and the HS will be seated together by using super-fine polishing compound and lots of rubbing of the parts against each other. One can get a semi- mirror finish that enhances the effectiveness of the heat transfer paste.
Working good
I have been enjoying listening for a couple weeks it seems.
I did notice that I can tell when things are getting warm by the way it sounds. After about 90 minutes or two hours of high volumes it just doesn't sound as good, not as crisp or sharp. Jazz trio, vocals, etc allow for cool running at high volumes for several hours. Rock, pop, blues, etc can warm things up quickly depending on the song.
Soon I will be implementing upgrades to fix the things I learned from the mistakes I made. i.e. -Larger heatsinks. I finally have that sorted out in my mind (I think), the smaller the C'/W the better. Heatsink USA has some stuff I like.
Also, though I am not having any buzzing problem, I do think that I should have twisted together the pairs of wires, and I didn't. As I "transplant" this to a better chassis I will do that better I hope. I'll let you know if this new line from PSU to AMP makes the amps imune to the buzz that occures if the cable-tv signal makes a loop between DVR, TV, & Receiver. Cable is fine to go to one or the other, but if it going to both had not worked (even if the amps playback another source seperate from tv.)
The goal is to have one chassis for powersupply and transformer, and another chassis for amplifier board. Both x2 as this is dual mono, so I'll end up with four chassis.
I made this, and I think it will be an improvement over the non-twisted wires I used initially. I took a pair of wires for V+ & PG+, and twisted them together with a drill. Also twisted together a nice pair for V- & PG-. Then there is a single conductor that will take earth from mainsplug in the powersupply chassis to the amplifier chassis. I then braided those three things, and fed them into a tech-flex jacket. I don't expect this kind of jacket to provide any rf-shielding, but I am thinking that with the way the wires are twisted that I won't need it. Do you think this cable will be improvement?
I did notice that my receiver's pre-amp-out section doesn't sound any better than my computer soundcard driven by Windows7 & iTunes.
Do you think a B1-Buffer and a DAC together would make a good pre-amp to match these dual mono lm3875 ?
I have been enjoying listening for a couple weeks it seems.
I did notice that I can tell when things are getting warm by the way it sounds. After about 90 minutes or two hours of high volumes it just doesn't sound as good, not as crisp or sharp. Jazz trio, vocals, etc allow for cool running at high volumes for several hours. Rock, pop, blues, etc can warm things up quickly depending on the song.
Soon I will be implementing upgrades to fix the things I learned from the mistakes I made. i.e. -Larger heatsinks. I finally have that sorted out in my mind (I think), the smaller the C'/W the better. Heatsink USA has some stuff I like.
Also, though I am not having any buzzing problem, I do think that I should have twisted together the pairs of wires, and I didn't. As I "transplant" this to a better chassis I will do that better I hope. I'll let you know if this new line from PSU to AMP makes the amps imune to the buzz that occures if the cable-tv signal makes a loop between DVR, TV, & Receiver. Cable is fine to go to one or the other, but if it going to both had not worked (even if the amps playback another source seperate from tv.)
The goal is to have one chassis for powersupply and transformer, and another chassis for amplifier board. Both x2 as this is dual mono, so I'll end up with four chassis.
I made this, and I think it will be an improvement over the non-twisted wires I used initially. I took a pair of wires for V+ & PG+, and twisted them together with a drill. Also twisted together a nice pair for V- & PG-. Then there is a single conductor that will take earth from mainsplug in the powersupply chassis to the amplifier chassis. I then braided those three things, and fed them into a tech-flex jacket. I don't expect this kind of jacket to provide any rf-shielding, but I am thinking that with the way the wires are twisted that I won't need it. Do you think this cable will be improvement?
I did notice that my receiver's pre-amp-out section doesn't sound any better than my computer soundcard driven by Windows7 & iTunes.
Do you think a B1-Buffer and a DAC together would make a good pre-amp to match these dual mono lm3875 ?
Attachments
When I build my gainclone (four years ago) I used microphone cable triple shielded for the inputs to potentiometer with excellent results. I would not use so many chassis (four) if you use one chassis for power supply you can always separate the inside with a thick piece of aluminum and the same goes for the amplifiers boards. if I could only build one chassis with thick aluminum enclosures inside that way I can separated all boards, power supply etc it would be a perfect amp chassis for me. But I do not have the right machine to cut aluminum and beside...I am bad cutting and building a chassis from scratch
PS: I bought from Heatsink usa and the service and heatsinks are very high quality.😉
- Status
- Not open for further replies.