Hi, I'm posting first time here.
I bought a Sanwu HF183 board with a TPA3116 amp chip and an unbranded BT, FM, MP3 chip. I immediately heard some distortion that's nothing like I have heard before. Not a clipping type distortion, instead it's kind of excited by upper mid frequency content of the music, triggered by distortion guitars or crash cymbals. No matter what's the volume level, the distortion remains the same. Has anyone noticed similar behavior with this board or any other TPA3116 based amps? Or may the BT chip cause this kind of distortion? Other issue with this board is quite loud power up pop.
I bought a Sanwu HF183 board with a TPA3116 amp chip and an unbranded BT, FM, MP3 chip. I immediately heard some distortion that's nothing like I have heard before. Not a clipping type distortion, instead it's kind of excited by upper mid frequency content of the music, triggered by distortion guitars or crash cymbals. No matter what's the volume level, the distortion remains the same. Has anyone noticed similar behavior with this board or any other TPA3116 based amps? Or may the BT chip cause this kind of distortion? Other issue with this board is quite loud power up pop.
Yes the hiss in both channels
I replaced ne5532p to LME49860NA but this hissing is still there
the board:
Aliexpress.com : Buy Breeze Audio BA100 HiFi Class D Audio Digital Power Amplifier tpa3116d2 TPA3116 Advanced 2*100W Mini Home Aluminum Enclosure amp from Reliable tpa3116 advanced suppliers on HIFI College Store
Thanks for the photo. It seems like a nice amplifier. It is a double-chip class D amplifier stage with a pre-amplifier stage (NE5532 based). The power line decoupling and output filter implementations seem much better than on the cheap TPA3116D2 board I have, but my board doesn't make any significant hiss. Just in front of my speakers, I am not sure if I hear any hiss and then it is fine. That should be possible for your board as well. On your board the heatsink can easily be removed – nice.
The NE5532 should not be the reason in itself. The 50K/5K potentiometer may be a part of the pre-amplifier circuit and the common problem with hiss using a 50K potentiometer (straight on the input of the TPA3116D2) may not be applicable for this amplifier.
In general, TPA3116D2 is known to generate hiss 1) If the TPA3116D2 amplifier input looks into a high impedance (>5K) from the input source, and 2) If the TPA3116D2 is set up for high gain.
First, it is important to separate the two active blocks (TPA3116D2 / NE5532) from one another and work backward from the output in order to identify the noise source. The NE5532 is in a socket such that it can easily be removed. By removing the NE5532, the pre-amplifier block is eliminated. But, we also need to ensure that the amplifier inputs are grounded such that they are not just floating inputs.
Between the input pins (pins 4 and 5) of each TPA3116D2 and the preamplifier there should be two 1uF signal coupling capacitors. On one side, the two 1uF capacitors are connected to pins 4 and 5 (of the TPA3116D2) respectively. On the other side, one of these 1uF capacitors is connected (directly?) to an output of the NE5532 (pin 1 or pin 7) and the other capacitor to ground (power supply “-”). I cannot see such details from the photos but I guess that the 1uF capacitors may be two of the small electrolytic capacitors located along the side of the heatsink. Can you please verify that?
IF, the 1uF signal coupling capacitors are of the electrolytic type, of a standard grade and not audio grade, we may already have found the source of the hiss. Standard grade electrolytic capacitors in the signal path is very often a source of noise. On my board they use ceramic SMD capacitors which is not good but at least somewhat better than standard electrolytic capacitors.
Will you please verify if the two 1uF capacitors are of the electrolytic type?
The NE5532 should not be the reason in itself. The 50K/5K potentiometer may be a part of the pre-amplifier circuit and the common problem with hiss using a 50K potentiometer (straight on the input of the TPA3116D2) may not be applicable for this amplifier.
In general, TPA3116D2 is known to generate hiss 1) If the TPA3116D2 amplifier input looks into a high impedance (>5K) from the input source, and 2) If the TPA3116D2 is set up for high gain.
First, it is important to separate the two active blocks (TPA3116D2 / NE5532) from one another and work backward from the output in order to identify the noise source. The NE5532 is in a socket such that it can easily be removed. By removing the NE5532, the pre-amplifier block is eliminated. But, we also need to ensure that the amplifier inputs are grounded such that they are not just floating inputs.
Between the input pins (pins 4 and 5) of each TPA3116D2 and the preamplifier there should be two 1uF signal coupling capacitors. On one side, the two 1uF capacitors are connected to pins 4 and 5 (of the TPA3116D2) respectively. On the other side, one of these 1uF capacitors is connected (directly?) to an output of the NE5532 (pin 1 or pin 7) and the other capacitor to ground (power supply “-”). I cannot see such details from the photos but I guess that the 1uF capacitors may be two of the small electrolytic capacitors located along the side of the heatsink. Can you please verify that?
IF, the 1uF signal coupling capacitors are of the electrolytic type, of a standard grade and not audio grade, we may already have found the source of the hiss. Standard grade electrolytic capacitors in the signal path is very often a source of noise. On my board they use ceramic SMD capacitors which is not good but at least somewhat better than standard electrolytic capacitors.
Will you please verify if the two 1uF capacitors are of the electrolytic type?
Last edited:
My TPA3116 board sounds good in my subjective opinion, but at least, it is rock solid in the sound. No such hint of a design-weakness.
You write that you hear this distortion at any sound level. Your board has only got 1300uF of power decoupling, such that, at slightly higher power levels your supply voltage may drop and that may be the source of distortion.
If the distortion is really at any sound level, it is likely to be the BT part. Is it the same with BT and USB-stick MP3?
The white ones should be foil 1uF/63V and fine. The two others are probably 2.2uF and electrolytic - grade I cannot see.
Remove the NE5532; make a ground connection to the terminals of the white capacitor turned AWAY from the TPA3116D2 chips and facing the NE5532 socket. Check that the two white capacitors are really grounded on one side and not on the other side (connected to the TPA3116D2) with an Ohm-meter. Try if you still have hiss using a 12V supply if possible (then no heatsink is needed at least for a shorter time).
Remove the NE5532; make a ground connection to the terminals of the white capacitor turned AWAY from the TPA3116D2 chips and facing the NE5532 socket. Check that the two white capacitors are really grounded on one side and not on the other side (connected to the TPA3116D2) with an Ohm-meter. Try if you still have hiss using a 12V supply if possible (then no heatsink is needed at least for a shorter time).
Last edited:
When I just remove NE5532 and turn on amplifier, noise is still there even at 0 volume and no input signal connection
When I remove NE5532 and set white capacitors to the ground the hiss much less.
But amplifier does not work w/o NE5532, no sound.
Maybe I should replace all 4 another electrolytic capacitors to film?
I do not know why manufacturer install only 2 film but not all 6.
Or need change gain somehow. I saw noise of one chip version of this amplifier (and 2.1 double chip board) can be fixed by changing gain. But no info how to fix noise of this double chip version.
When I remove NE5532 and set white capacitors to the ground the hiss much less.
But amplifier does not work w/o NE5532, no sound.
Maybe I should replace all 4 another electrolytic capacitors to film?
I do not know why manufacturer install only 2 film but not all 6.
Or need change gain somehow. I saw noise of one chip version of this amplifier (and 2.1 double chip board) can be fixed by changing gain. But no info how to fix noise of this double chip version.
Last edited:
It's the same with USB-MP3. What comes to mind after more listening it might also be a sharp spike in frequency response @4-6kHz. Not a surprise when speaking of noname decoder chip.
Don't get me wrong, it's not wery bad. The sound quality is good enough for example for background music. IMO it's just not HI-FI.
If you just remove the NE5532, the TPA3116D2 input is looking into a high impedance toward the signal source and from experience that results in hiss.
No, without a NE5532 the amplifier does not work from the inputs.
IF, the hiss had completely disappeared when you grounded the two white capacitors it would be most likely that the hiss originated from the pre-amplifier stage. Actually, the hiss was less (may be explained by that the source impedance is very low with grounding) but some remained. It is very likely that the remaining hiss is caused by the electrolytic capacitor(s) on the other input.
Let's try from the psychological angle: The advertisement of the board shows no white coupling capacitors, only electrolytic capacitors. Your board has got two foil capacitors, but inserted in a place marked for electrolytic capacitors. A hypothesis - the manufacturer realized that he had a problem with the electrolytic capacitors and decided to use foil capacitors instead. Probably due to a noise problem. He could not fit other than a 1uF foil in the place available. But, the manufacturer did not take account of that it is a differential input such that also the two other electrolytic capacitors should have been replaced.
Again, it is most likely that also the two last electrolytic capacitors should be replaced with 1uF foil capacitors (1uF is recommended by TI in the datasheet).
Let's see how much two more foil capacitors help.
No, without a NE5532 the amplifier does not work from the inputs.
IF, the hiss had completely disappeared when you grounded the two white capacitors it would be most likely that the hiss originated from the pre-amplifier stage. Actually, the hiss was less (may be explained by that the source impedance is very low with grounding) but some remained. It is very likely that the remaining hiss is caused by the electrolytic capacitor(s) on the other input.
Let's try from the psychological angle: The advertisement of the board shows no white coupling capacitors, only electrolytic capacitors. Your board has got two foil capacitors, but inserted in a place marked for electrolytic capacitors. A hypothesis - the manufacturer realized that he had a problem with the electrolytic capacitors and decided to use foil capacitors instead. Probably due to a noise problem. He could not fit other than a 1uF foil in the place available. But, the manufacturer did not take account of that it is a differential input such that also the two other electrolytic capacitors should have been replaced.
Again, it is most likely that also the two last electrolytic capacitors should be replaced with 1uF foil capacitors (1uF is recommended by TI in the datasheet).
Let's see how much two more foil capacitors help.
Last edited:
You may try the TPA3116D2 part out with a better signal source if you can find a suitable input point.
We always have to distinguish between what is clearly a flaw in a construction and what is performance we like to improve. A flaw may be repaired by resolving one fault, general performance improvement often requires more things to be changed.
Thanks FauxFrench. I will try change remaining 4 capacitors, But I do not have them yet 
It's need to buy them and replace...
Interesting, why onboard marked 1u but manufacturer installed 2.2µF
Would be acceptable replace all 6 capacitors to this?
WIMA 1uF
10pcs WIMA 1uF/100V 10% MKS2 Film Capacitor Hi Fi Audio 105K-in Inductors from Home Improvement on Aliexpress.com | Alibaba Group
It's need to buy them and replace...Interesting, why onboard marked 1u but manufacturer installed 2.2µF
Would be acceptable replace all 6 capacitors to this?
WIMA 1uF
10pcs WIMA 1uF/100V 10% MKS2 Film Capacitor Hi Fi Audio 105K-in Inductors from Home Improvement on Aliexpress.com | Alibaba Group
Last edited:
Fine Wima capacitors that seem to have the same physical size as the white ones used at present. A good choice.
BUT, we are talking about the 4 capacitors that are connected to either pin 4 or pin 5 on the TPA3116D2. The two last (electrolytic) capacitors are used for a different purpose and are likely not to affect the hiss. They may be used for the purpose shown in Fig.37 as C39 of the datasheet. It looks like a decoupling of the power limit and gain settings. So, a (better) foil capacitor here will probably not increase performance. For a start I would leave the two last (electrolytic) capacitors as they are.
Never change too much in one go because if you afterwards have a new surprising result you do not know which change caused it.
BUT, we are talking about the 4 capacitors that are connected to either pin 4 or pin 5 on the TPA3116D2. The two last (electrolytic) capacitors are used for a different purpose and are likely not to affect the hiss. They may be used for the purpose shown in Fig.37 as C39 of the datasheet. It looks like a decoupling of the power limit and gain settings. So, a (better) foil capacitor here will probably not increase performance. For a start I would leave the two last (electrolytic) capacitors as they are.
Never change too much in one go because if you afterwards have a new surprising result you do not know which change caused it.
New mono board?
Hi,
I just saw this while looking for mono TPA3116 boards: Aliexpress.com: Acheter XH M545 high power automobile power amplifier board TPA3116D2 boost program DC12V single channel 150W de channel protection fiable fournisseurs sur Great vendor
Have you already seen this XH-M545 board? (the search on the forum didn't return anything).
What do you think of this?
What are the 2 extra heatsinks used for?
Hi,
I just saw this while looking for mono TPA3116 boards: Aliexpress.com: Acheter XH M545 high power automobile power amplifier board TPA3116D2 boost program DC12V single channel 150W de channel protection fiable fournisseurs sur Great vendor
Have you already seen this XH-M545 board? (the search on the forum didn't return anything).
What do you think of this?
What are the 2 extra heatsinks used for?
Thanks FauxFrench. I will try change remaining 4 capacitors, But I do not have them yet
Interesting, why onboard marked 1u but manufacturer installed 2.2µF
Would be acceptable replace all 6 capacitors to this?
WIMA 1uF
10pcs WIMA 1uF/100V 10% MKS2 Film Capacitor Hi Fi Audio 105K-in Inductors from Home Improvement on Aliexpress.com | Alibaba Group
Can you perhaps show and describe your complete chain of components you are using in this setup. Why I ask is there may well be a single culprit which are giving you the bad sound result. I have had many of these types of amps and have yet to find one that is inherently noisy.
Biggest culprit is normally the power supply. My preference as to power supply is a 5V powerbank as long as the board can run off 5V - most can.
Also not sure why that board has an op-amp at all? There are no tone controls present, so why the op-amp?
If you have replaced the ne5532p with a LME49860NA as per your previous post, a 5V power supply should work 100% and enable you to test and possibly eliminate the power supply being the noise generator. A lot of AC/DC power supplies lets through enough AC to cause noise problems, even expensive laptop power supplies.
Good luck and let us know what you find.
Now I use 24v power supply and loudspeakers with 87 db sensitivity.
I have tryed different 12v power supplies and 12v from PC PSU, and 12v from batteries but hiss from tweeters of loudspeakers still exists.
As I wrote before, the hiss presents even when no input connection and when volume set to 0.
Anyway 5v is too low for this amplifier.
Yes I changed ne5532p with LME49860NA and maybe sound now a little bit better, do not sure, but this change not affected to the noise
I thought on the board set 50K potentiometer and wanted replace to 10K, but on this board installed 5K by manufacturer and no reason to change it I guess.
Now I ordered wima film capacitors and will replace capacitors marked 2.2µF 50v to those new
Btw as I understand, manufacturer installed 2.2µF but not 1µF as market on the board to improve bass, cause this amplifier has problem with low freq sound.
But all the same bass really very poor and the high frequencies are very sharp.
I will try install new film capacitors and look / hear changes. May be will install 2 capacitors in parallel to increase low frequencies. Because I ordered only 1µF capacitors, but you can always install them in two in parallel
I guess also make gain lower will make noise lower but I do not know how to change gain on this board.
If someone knows how to set gain lower on this board please write.
I have tryed different 12v power supplies and 12v from PC PSU, and 12v from batteries but hiss from tweeters of loudspeakers still exists.
As I wrote before, the hiss presents even when no input connection and when volume set to 0.
Anyway 5v is too low for this amplifier.
Yes I changed ne5532p with LME49860NA and maybe sound now a little bit better, do not sure, but this change not affected to the noise
I thought on the board set 50K potentiometer and wanted replace to 10K, but on this board installed 5K by manufacturer and no reason to change it I guess.
Now I ordered wima film capacitors and will replace capacitors marked 2.2µF 50v to those new
Btw as I understand, manufacturer installed 2.2µF but not 1µF as market on the board to improve bass, cause this amplifier has problem with low freq sound.
But all the same bass really very poor and the high frequencies are very sharp.
I will try install new film capacitors and look / hear changes. May be will install 2 capacitors in parallel to increase low frequencies. Because I ordered only 1µF capacitors, but you can always install them in two in parallel
I guess also make gain lower will make noise lower but I do not know how to change gain on this board.
If someone knows how to set gain lower on this board please write.
They use two XH6019 in parallel to boost the amps supply voltage. That's what the heatsinks are used for.
Thanks doctormod. Is this a good thing? Or should I stick to simpler board?