TDA7293 capacitors

[muraj]

Hi,

I am planning to build a TDA7293 based amplifier. This is my first audio amplifier attempt. I think I have understood most of the TDA7293 datasheet. I couldn't understand which types of capacitors should be used? Based on availability in local markets I guessed the following

Power supply decoupling: electrolytic
Power supply noise suppression: disc
All others: polyster or polypropylene.

Polyster caps are very expensive. To my surprise I found that some of them actually cost more them the TDA7293 chip itself. Did I miss something basic?

Thanks in advance.

[danielwritesbac]

For example, instead of purchasing a 22uF polyester. . .you could use a 22uF electrolytic in parallel with a 0.01uF polyester.

I think that you might accidentally be shopping for speaker crossover size capacitors. You might not need 500v rated polyester where 100v models can work.

[muraj]

Thanks for this information. I was much smarter this weekend asked the right questions :-). I did miss the obvious earlier and asked too many questions at an electronics shop, which lead to him doubting whether I was going to buy anything at all.

Anyways I'll post more information as I proceed.

[muraj]

I got it working and find it's sound satisfactory. I find it's low end frequency response good enough, but I came across this website Avoiding TDA7293 pitfalls. I was wondering if this actually improves anything. TDA7293 datasheet says that increasing values of various signal path capacitors has no effect on frequency response.

[AndrewT]

Which part of the linked document says that changing audio path capacitors has no effect on the frequency response?
My copy runs to 14 useful pages.

[danielwritesbac]

I agree with AndrewT. He is proven correct about the caps--he fixed my TDA7294.

[muraj]

Page 10, section 4.1 applications suggestion says -

• increasing C1 (input DC coupling) has no effect on frequency response. Decreasing it will increase low frequency cut-off.
• inreasing C2 (feedback DC decoupling) has no effect on frequency response. Decreasing it will increase low frequency cut-off.

I used the values given in the datasheet and liked it's sound. I don't find any lack of low frequency response or high frequency roll-off. However the frequency response graph provided by BNICS shows low as well as high frequency roll-offs. Is that how it's expected or can be better. I found another article which says you can improve low frequency response by increasing all capacitors.

Could you help understand this?

Thanks.

Originally Posted by AndrewT

Which part of the linked document says that changing audio path capacitors has no effect on the frequency response?
My copy runs to 14 useful pages.

[danielwritesbac]

If you increase input cap without increasing NFB cap it will make boomy warm bass like a mass market retail amplifier.

Because of the onboard current limiter real time action of cutting the biggest current, which is the lowest notes, TDA7293/4/5/6 amplifiers require severe overdo on bass harmonic balance in order to achieve normal bass. That is not explained in the counterproductive datasheet examples--The datasheet example just throws big power caps at the problem, with an approach that levels the tone at the expense of the clarity.

The most important thing to remember about harmonic bass balance, is that its not the high pitched warm/boomy that you hear but rather it is the lowest pitches from source replayed incorrectly to higher pitches and thus the problem is easily solved by the final step of reducing the input cap size. . . after you have allowed/permitted clean amplification of bass by upsizing the NFB cap, a lot, such as 470uF//22uF//1nF. Anything big enough is fine as long as it is also clear. Using a larger NFB cap allows you to use a larger input cap, without incurring bass harmonic distortion.

Note that the feedback-shunt resistor could be reduced in value to make the NFB cap extremely large size, or the other way around--increase the feedback-shunt resistor value to facilitate more compact NFB caps, if you like. Some options include 220R to 3,300uF, 680R to 1,000uF, 1.5k to 470uF, or 2.7k to 250uF. . .

I don't mind a large cap(s) as long as it is clear and works. Any capacitor good enough for NFB cap is likewise good enough for power duties at the amplifier board. Yes, the quality demands can be that strong and quality is not related to cap price. For buying caps, just search to see what others are actually using and keep a tight grip on your wallet. Above all, try the right size part first.

Revision:
Bootstrap cap: smaller = supports tiny woofer, larger = normal bass
NFB cap: smaller = booming/warm sound, far larger = normal low bass
Input cap: smaller = less of the above, larger = more of the above
Predrive power pins: if you want clear sound, pair of 220uF with solo 2uF or just a pair of 330uF can work.
Output buffer power pins: unnecessarily high capacitance works unnecessarily well and probably enjoyable.
Mute does not require added cap; however standby is same as datasheet

[gloriel]

Originally Posted by danielwritesbac

Revision:
Bootstrap cap: smaller = supports tiny woofer, larger = normal bass
NFB cap: smaller = booming/warm sound, far larger = normal low bass
Input cap: smaller = less of the above, larger = more of the above
Predrive power pins: if you want clear sound, pair of 220uF with solo 2uF or just a pair of 330uF can work.
Output buffer power pins: unnecessarily high capacitance works unnecessarily well and probably enjoyable.
Mute does not require added cap; however standby is same as datasheet

Sir, what pin number of Output buffer pins? how do i pairs of the caps? Do you have complete schematic?

regards

Glo