Why are audio circuits so fussy with copper layout?

[HandyMan]

Hi,

I'm trying to build myself a basic amplifier to use with my PC as some may already know from my questions about a couple of ICs to do the job previously.

I've now switched to the TDA2005 IC to do the job. Yet again, though, although it amplifies, there is tremendous noise and pulsing coming from the speakers. It is away from interference and I've tried alternative power supplies and audio inputs. So, three different amplifier ICs and the same trouble each time...

With this IC, there is a hobby circuit available all over the net which uses a similar circuit to the one I was copying, so I tried that - same problem. Then I noticed that on one site selling the kit, they said this:

"We spent a lot of time getting the tracks right on this single sided PCB to get the lowest possible THD. Getting the separate tracks to the earth pin is the secret."

Could someone please explain what THD is and would it sound like what I have complained of: noise, pulsing, but mainly loud static noise?

Anyway, I'm now wondering if the problem with all the circuits has been my layout. I'm using Eagle to do it, and it autoroutes with thin tracks which I thicken but it doesn't allow for easily making areas of copper on the PCB, and I haven't done so anyway as I didn't know it was necessary or where it was necessary.

I am going to try and do my homebrewed PCB with the exact component placement and as similar copper layout as I can see on the hobby kit's PCB picture as possible in the hope this will do the trick. But why are these amplifier circuits so fussy? I gather it's the grounding that's the trouble - my thin tracks running all over the place to ground components dotted about randomly is not good, instead big thick ground parts running around the edge of the PCB are needed with the component sides that need grounding actually imposing into this copper area where possible. But why? Whether this way or a thin track running to it, there's still an electrical connection, isn't there?

Someone please help educate me :roll:

Thank ye

 

[audioguru]

Hi HandyMan,
THD is total harmonic distortion and usually sounds just a bit rough. Your severe problem is probably caused by the TDA2005 oscillating at a very high frequency.
It is important with audio amplifier pcb's to keep high supply or ground current from flowing through the printed tracks that connect to the input, to avoid positive feedback.
A ceramic supply decoupling capacitor is recommended and the same for the RC networks across the output.
Didn't you use the pcb layout that is shown in the datasheet?
https://www.electro-tech-online.com/custompdfs/2005/03/tda2005.pdf

 

[Sebi]

THD: total harmonic distortion. I mean, this parameter depend from IC construction, not strongly from PCB layout.
All amplifier fussy with bad layout, extremely thin GND. Here flow a few amperes in peak, and make a voltage feedback to input = oscillation.
Make a big GND area as possible, use 1..2000uF elko and 100nF decoupling caps nearby the IC supply pin, the output RC component legs mus be short as possible, the R must be inductorless.

 

[HandyMan]

Hi,

Thanks for the advice from you both. I understand better now, although I forgot to ask what "bootstrapping" is, as I've seen this IC with the bootstrap pins coupled via a capacitor to the speaker outputs and also with those same pins just going straight to VCC (not using the bootstrap). What's that all about?

I didn't copy the layout from the datasheet as the circuit is different to the one I want to use (stereo outputs) and so I used the schematic from a hooby kit site, but no board to compare to.

Here's my new circuit which I will print out on to transparency, use to expose my photo-resist board and then make the amplifier. Before I do, could I just ask if this looks good enough?

**broken link removed**

Thanks

 

[David Bridgen]

Hi,
Could someone please explain what THD is and would it sound like what I have complained of: noise, pulsing, but mainly loud static noise?

T.h.d. stands for total harmonic distortion.

No, it wouldn't sound like noise, pulsing, but mainly loud "static" noise.

Have you got a fairly large value capacitor connected across the supply voltage? The absence of this could be the cause of your problem.

Separate tracks to a common point, usually at the amplifier's input terminal/s, for each component which needs to be connected to the circuit ground is a common technique to try to minimise instability. The wider the tracks, the lower their impedance. (Another important factor.)

 

[panic mode]

post audio sample so we can hear it...
most problems i've seen is when people don't use proper shielding
for the input cables or when they create ground loops.

 

[Sebi]

The output RC components must be closer to the output pins.I recommend to use here SMD parts.

 

[audioguru]

It looks like you have the feedback capacitors at pins 2 and 4 connected through resistors to the wrong opposite amplifier's output.

 

[HandyMan]

David - thanks for the explanation and advice. On my board then, I should have each ground connection from each component going separately to the ground pin of the audio input (bottom middle of my board)?

Loud static noise is about what I've got. Had a 4700uF cap across the supply voltage, no difference whatsoever.

Panic Mode - That's a good idea. It crossed my mind a while back to do that, but I thought it might not be something people wanted to bother with. I take up so much time of those who are kind enough to help that I felt like I'd be taking the mick a bit. Anyway, will reconnect the amp built using the old thin-track circuit and get the noise on it recorded and on here ASAP.

Sebi - the board layout is copied from a kit which is available using the identical circuit, where they claim that it works well... so I thought it'd work for me too. I have no SMD parts and wanted to make this cheaply using bits I have here.

Audioguru - you've got keen eyes there and I can't thank you enough for spotting that most embarrassing error!

You really must have taken some time looking over the board, and that is really appreciated. I'm sometimes surprised by how much effort people will put in to help someone they have never even met, it's very nice indeed.

Here is the board with error rectified and with the power input smoothing caps positioned before the IC VCC pin, instead of after it. The circuit diagram is below it too.

**broken link removed**

**broken link removed**

Thank you to all who spend their free time helping me. All the best to you all.

 

[audioguru]

Hi HandyMan,
C4, C10 and C11 should be ceramic, not poly.
C4, C10 with R7 and C11 with R8 should be as close to the IC pins as possible, perhaps to the left side of the IC. :lol:

 

[HandyMan]

Hi,

Just finished building the amplifier using a very slightly altered version of the board shown in the pic above.

It works perfectly! :shock:

I'm listening to some music on it now, as it's connected to my PC and it's running a couple of JBL car speakers. They're expensive powerful buggers and I wanted to use them for something now got new car.

When no input is connected, the amp just makes the slightest hum like all amps seem to. With audio going in, it amplifies brilliantly and the sound is superb.

Seems that the thin tracks and lack of a proper single ground path caused the troubles experienced before.

Here's a couple of pics of the finished board before population but after using my inkjet printer + transparency, UV face tanner + photo-resist single-sided board, developer + warm ferric chloride PCB making process. It gives perfect results (ignore the bad edge on the board, I can't cut well with the saw...):

**broken link removed**

**broken link removed**

So there you have it. Thanks to everyone for their help and advice, it's made this project work at last! (fourth time lucky, but hey, I wanted to waste lots of photo-resist board. I enjoy it. Really.)

James 8)

 

[Sebi]

The board look perfect. Congratulations!

 

[monkeytree]

how do you get such good boards mine arent usable

 

[Nigel Goodwin]

It's a good job he hasn't got his location filled in! - those boards look so good I'd be tempted to go round and break his legs :lol:

 

[HandyMan]

Thanks for the compliments, but mainly for all your help and advice with producing a board which works.

Nigel, I've now entered the country in my profile, just to tease you.

Monkeytree, I'll write my method in your new topic, which might hopefully work for you too.

James

 

[aehandasa]

please send me project source on my mail
thanks

 

[aehandasa]

THANKS HandyMan

 

[Pyroandrew]

noise in analog systems is a real *****, especially audio. there is a vast black art getting audio sytems to sound perfect, it is especially difficult if you have digital componants as well as analog. one small alteration can bugger everything up, i.e. going from prototype to PCB. Well done on the boards by the way.