tda2030 amplifies noise

[Zak28]

Recently made a 12v single supply tda2030 amp and its very noisy. All non polarized caps are 0805 or 1206 MLCCs the board has footprint for 4700uF rail capacitor instead of 220uF - haven't tried 4700uF but circuit with 220uF made noise lots of noise with 2.3R and 8R speakers powered by a meanwell GST60A12-P1J 12v 5A adapter.

What can be done to remedy the noise issue?

https://i.ibb.co/KwZn2tZ/circuit.png


https://i.ibb.co/tHj3ZXd/tda2030-Top.gif


https://i.ibb.co/5nKSsmd/tda2030-top-fill.gif


https://i.ibb.co/GvpRcSr/tda2030-bottom.gif


https://i.ibb.co/7gdgGb6/tda2030-bottom-fill.gif

 

[Easy peasy]

try shorting the input and vary the volume pot from zero to full - if the noise is there and much the same as you vary the pot - then the noise is coming from else where

if the noise goes up as the pot turns up - then the noise is the TDA2030...

try adding 100pF across R4

 

[Zak28]

Perhaps heatsink is a major cause of noise as there is no copper under the heatsink and its not isolated from TDA tab. Going to fab this board since I discarded the others for their issue with staggered leads. The inline leads on this board will permit lm1875 also along with a few silk screen adjustments for the terminal blocks and larger rail filter cap. Not sure whether the 4700uF cap will improve the quality.

What about the layout can be improved?

 

[betwixt]

What kind of 'noise' is it?
Background hiss like white noise or something containing regular tones.
Note that the TDA2030 is over 30 years old - amplifiers have improved a lot since then!

I doubt the diodes do anything useful, if they are intended to protect against back-emf from the speakers they need to be faster responding than 1N400x series.

Brian.

 

[Zak28]

What kind of 'noise' is it?
Background hiss like white noise or something containing regular tones.
Note that the TDA2030 is over 30 years old - amplifiers have improved a lot since then!

I doubt the diodes do anything useful, if they are intended to protect against back-emf from the speakers they need to be faster responding than 1N400x series.

Brian.

It appears to be high frequency pitched tone, TDA2030 is excellent for audio amplification, there is class A and I have many TTA1943(Q) and TO-3 parts for hifi if required.

 

[betwixt]

The tone implies you have instability. I last used those devices in the 1980s and I can remember them being difficult to keep stable. They are not particularly good devices. Try removing the two diodes and also move C3 so it is directly between pins 3 and 5 with shortest possible track length.

Brian.

 

[Zak28]

I redid the board with film caps implace of any capacitors dealing with input audio. The only MLCCs are low value bypass caps. Is this how the suggested 100pF should be?

https://i.ibb.co/CMCX732/img.png

 

[Easy peasy]

yes, you may want to try a 2k2 - 10k res in series with the 100pF too, on the LHS of the cap ...

 

[Zak28]

yes, you may want to try a 2k2 - 10k res in series with the 100pF too, on the LHS of the cap ...

Not sure were to place suggested resistor - would a small variable resistor be better? Perhaps it can be adjusted to notch out the noise.

 

[Easy peasy]

noise is inherently wide band - there is no notching it out - most noise "filters" are low pass filters, I'm sorry I can't clarify further than " put a resistor in series with the 100pF cap " with the resistor at the pin 2 end or LHS, ( left hand side ) of the cap.

- - - Updated - - -

please don't forget to post your results

 

[Zak28]

noise is inherently wide band - there is no notching it out - most noise "filters" are low pass filters, I'm sorry I can't clarify further than " put a resistor in series with the 100pF cap " with the resistor at the pin 2 end or LHS, ( left hand side ) of the cap.

- - - Updated - - -

please don't forget to post your results

Going to fab the board after laying it out further. Is this proper arrangement with the suggested capacitor and pot?

https://i.ibb.co/PhQPkMj/img.png

 

[betwixt]

The capacitor is just to roll off the high frequency response, there is no point in making it adjustable. If you are getting a continuous tone out of the amplifier (post #5) it implies it is oscillating and although dropping the HF gain may help it doesn't fix the underlying problem of a feedback path somewhere. You may have to use a capacitor so big that the audible high tones are attenuated before instability stops and with such a big value you may run into other problems.

Brian.

 

[Zak28]

Are PNP class a amplifiers less prone to amplifying noise as did this specialized audio IC?

The noise was not a tone, it was high pitched static and was constant regardless of how much the input pot attenuated input signal.

I remade the amp board the only MLCC is the 100nF power rail capacitor, seems MLCCs small phase margin caused oscillation. Also the 100nF bypass cap which was suggested to be moved closer to tda IC was further in previous board which could have been an issue along with an increase in power rail capacitor from 220uF to 4700uF which is likely to attenuate ripple from power adapter.

 

[KlausST]

Hi,

Note:
In post#5 and post#13 you talk about class A, but the TDA2030 is class AB.

The "class" (mainly) defines the output stage of the amplifier..and is not related to noise (which mainly is caused by the input stage.

But your problem is not "noise" it rather is a stability or GND loop problem

Klaus

 

[Easy peasy]

If the power supply rails are "noisy" this may well translate to noise at the output - you mentioned the 12V power was from a meanwell power supply ... LC filtering probably a good test to see if the power supply is the noise source ...

 

[Audioguru]

The TDA2030 is old and was cheap so of course it produces hiss. It is not made or available anymore. The more modern TDA2030A also is obsolete and is not made or available anymore.
An LM3886 is more modern and was designed for high fidelity with a spec'd maximum very low amount of noise.

 

[Zak28]

The TDA2030 is old and was cheap so of course it produces hiss. It is not made or available anymore. The more modern TDA2030A also is obsolete and is not made or available anymore.
An LM3886 is more modern and was designed for high fidelity with a spec'd maximum very low amount of noise.

Whats form of products were tda2030s typically employed in? They seem fine for mic amplification but for audio it would require proper component selection such as metal foil resistors and proper layout.

 

[Audioguru]

I have never seen a TDA2030 or TDA2030A. I deleted the datasheet of the old TDA2030 but you can look it up to see if it spec's noise level like the TDA2030A datasheet does. The noise level spec'd for the TDA2030A is VERY low so maybe your TDA2030 IC is a fake or defective one.

 

[betwixt]

I have some in stock but they have been collecting dust for 30 years. The last time I saw one in a commercial product was probably in the mid 1980s, they were commonly used in low cost domestic "Hi Fi" units when cassette tapes were considered high quality and 8-track players were common. I would add that many of those "Hi Fi" units used to boast 20W per channel but the TDA2030s didn't even have heat sinks. I would think more like 5W per channel at 10% distortion but nobody checked in those days.

Brian.

 

[Zak28]

I have some in stock but they have been collecting dust for 30 years. The last time I saw one in a commercial product was probably in the mid 1980s, they were commonly used in low cost domestic "Hi Fi" units when cassette tapes were considered high quality and 8-track players were common. I would add that many of those "Hi Fi" units used to boast 20W per channel but the TDA2030s didn't even have heat sinks. I would think more like 5W per channel at 10% distortion but nobody checked in those days.

Brian.

Was enormous distortion due from improper layout and absence of noise reduction? It shouldn't have had much issue if it was battery powered.

Is there a drop in upgrade to lm1875 or at least a similar but newer IC?