diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Construction Tips (http://www.diyaudio.com/forums/construction-tips/)
-   -   TEA2025B - Cannot resolve white noise. (http://www.diyaudio.com/forums/construction-tips/226872-tea2025b-cannot-resolve-white-noise.html)

Chozo4 2nd January 2013 08:44 AM

TEA2025B - Cannot resolve white noise.
 
Greetings all,

(Bah, lost a whole post because the cat backspaced it to the last page...)

In advance - I do apologize if I may sound far too green behind the ears when it comes to electronics so do bare with me here.

I've since built my first actual device while still admittedly not knowing much about electronics so no luxury as a multimeter at the moment other than thick solder, a soldering iron, soderwick, and [vintage] NoKorode Soldering Paste. I've since changed sound cards in my computer from one that had a 2w TEA2025b amplifier on it to one with a 0.5w? preamp (forget the model number at this point) and as a result needed an amplifier to mount in my Small form factor case. Used the soundblaster mounted on my wall wired up directly to the tea2025b amplifier but figured it was about time I actually built a tea2025b amp that I could mount behind the faceplate of my Small-FormFactor PC case. Have to admit - hanging a large soundcard to the wall while amusing was but annoying to glare at.

So - I've got it built but the problem is that I simply cannot get rid of the white noise in the background (more like a mild hiss that doesn't change with volume). At first I had the issue of a heavy hum until I grounded to the inputs rather than the mains. I've tried star-grounding all the inputs and the noise is still there irregardless as well as swapping in and out components on both channels. I've suspected it may be due to the full gain but I've seen others built that didn't have any complaints from others of the noise resulting from it. On top of that I cannot test that theory due to having no smaller than 100ohm 1watt resistors (all of the resistors are 1 watt) and using a 100ohm completely silences the feedback but the noise is lessened and no input is heard.

The layout is based on the Radioshack General Purpose PCB #276-168B for the sake of a compact build within 2x3inch size. Schematics used are the TEA2025B stereo in the tea2025b documentation and the layout presented on the soundcard that had it. I do apologize about the crappy MSPaint layout - it's the best I could do right now but hoping it's clear enough. The parts I used are all I could gather so was hoping they are at most within spec enough to not garner any problems but here we are with white noise as the end result (but it works fine otherwise.)

http://rwinds.no-ip.org/~archive/TEA2025B_Layout.png

Sources used are a pair of headphones for output and computer/pc audio for input (the intended platform to use the amp on) with a 5v testing voltage off the USB. Will use a simple Panasonic 9v Regulator on completion to drop from 12v to 9v within the PC.

NOTES:
1) The noise is encountered on grounding to input
2) hum is evident from grounding to mains (so not grounding there)
3) star-grounding did not resolve the noise on either end.
4) Adding a secondary capacitor (104 ceramic) to the power ground did not make much of a difference (noticed it used on a Chinese schematic hoping it would help).
5) The soundcard a majority of the parts were based off didn't have quality issues that I could tell.

Resolutions:
1) Changed power source from USB(with LED Activity light) to 5v 1A wall wart - most of the noise resolved
2) Adding a 100ohm resistor to the output ground resolves the remainder of the noise.

.. in progress so far:
3) Bypassing the RC series which used a polar electrolytic and connecting the inout directly to pin2 seems to have resolved the low top-end and drowning out. May need to put the monolithic cap back in its' place and see how it goes from there.

epicyclic 2nd January 2013 11:13 AM

Hi Choz

try putting a 100 ohm resistor in series with each feedback pin and there respective feedback capacitors ( feedback pins are pin 6 and 11 ) this will reduce the gain from 46dB to 36dB ( 36 appears to be the minimum recommended ) .

Chozo4 2nd January 2013 11:26 AM

Quote:

Originally Posted by epicyclic (Post 3307775)
Hi Choz

try putting a 100 ohm resistor in series with each feedback pin and there respective feedback capacitors ( feedback pins are pin 6 and 11 ) this will reduce the gain from 46dB to 36dB ( 36 appears to be the minimum recommended ) .

That's what I tried and ended up with a muted output instead (putting it in an RC series at the feedbacks). Will try again and see if things change somehow but need to reassemble it from scratch. Ended up tearing off one of the leads on the side of the board from swapping components so much (Have PCB but not breadboard) as a result of too many heat/cool cycles with the iron. 100 ohm is the lowest value I have available so here's hoping anyhow.

Was testing it with 5v input - perhaps that could be why 100ohm was silencing it? Not enough voltage to overcome the 100ohm feedback reduction? Anywho - after a rebuild tonight will reply back with the results as I feel I could have better planned out a more uniform format. Rather glad I got the larger PCB - gives more room for error and retries.

Chozo4 3rd January 2013 08:47 AM

Haven't gotten the chance to modify things but just realized something I missed from the documentation. The Capacitor from pin 2/15 to ground I'm using is smaller than 0.15uf (0.10uf actually) and according to the documentation anything smaller poses a risk in causing oscillation (Verbatim: "DANGER OF OSCILLATIONS"). Perhaps if I switched to something larger instead as the closest I have is the 274 (0.27uf) I have at the inputs. Putting a reversed capacitor of 1uf in the inputs place for now (or removing altogether since it's to buffer input static from volume controls) and putting the 274 in place of the 0.10uf at 2/15->GND could remedy the issue.
Additionally the documentation also states not to use a an IC socket for stability sakes but I'm using one for the build to avoid damaging the IC from any mishaps. Rather doubtful however of that being be a lead cause (no pun intended) in the noise.

Will see once I get a chance tomorrow hopefully to rebuild on a fresh PCB section.

Chozo4 3rd January 2013 04:55 PM

Cannot seem to edit posts after a certain amount of time....

After going nuts longer I realized that a majority of the noise was the result of my power supply. Was getting it over my PC's USB as it is but also the USB cable I was using had a blue LED to show power usage which was adding some heavy amounts of feedback over the supply pin. After breaking down and using a 5v 1A DC wall wart most of the noise is.. pretty much gone. The small amount that's there is hardly noticeable until an input source is added. which is somewhat noticable but is rectified by adding either a 10k ohm resistor between the input lines (already have one from input to ground..) or adding a 1k ohm to the output ground. Adding the resistor to the output seems more economical overall until I can get a resistor of less than 100ohm to reduce the gain as 100ohm still seemed to kill off whatever sound there was..

Aside from that - turning up the input volume beyond half causes the bass and then sound itself to sound drowned out and muddled like maxing out blown speakers. That though is most likely the result of not enough voltage as I am using a 5v input afterall and the 100uf voltage bypass capacitor I'm using temporarily..


All times are GMT. The time now is 08:07 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2