Hi everyone!
I recently bought a small and cheap amply with TPA3116D2, an XH-M576, like this one in the link: CNKESIN XH M567 DC12 24V 80 W x 2 Digital Audio TPA3116D2 Alti Bassi Regolazione Preset Pre Amplificatore Consiglio in CNKESIN XH-M567 DC12-24V 80 W x 2 Digital Audio TPA3116D2 Alti Bassi Regolazione Preset Pre Amplificatore Consiglioda Circuiti integrati su AliExpress.com | Gruppo Alibaba.
The amply output an annoying and persistent hiss, even without input connected, and it goes louder with the increase of the volume.
I tried the 4k7 resistors solution on R and L input to ground and I also removed the 50k volume pot and tried a 10k pot, but nothing changed.
If I touch with bare hands the metal rod of volume or bass or trebble pot further noise add up to the hiss. I also have a "nice" pop when I remove power (19 volt dc dell laptop psu)!
20190429-142354177-i-OS — imgbb.com
Do you have any suggestion or I can trash it directly?
I recently bought a small and cheap amply with TPA3116D2, an XH-M576, like this one in the link: CNKESIN XH M567 DC12 24V 80 W x 2 Digital Audio TPA3116D2 Alti Bassi Regolazione Preset Pre Amplificatore Consiglio in CNKESIN XH-M567 DC12-24V 80 W x 2 Digital Audio TPA3116D2 Alti Bassi Regolazione Preset Pre Amplificatore Consiglioda Circuiti integrati su AliExpress.com | Gruppo Alibaba.
The amply output an annoying and persistent hiss, even without input connected, and it goes louder with the increase of the volume.
I tried the 4k7 resistors solution on R and L input to ground and I also removed the 50k volume pot and tried a 10k pot, but nothing changed.
If I touch with bare hands the metal rod of volume or bass or trebble pot further noise add up to the hiss. I also have a "nice" pop when I remove power (19 volt dc dell laptop psu)!
An externally hosted image should be here but it was not working when we last tested it.
20190429-142354177-i-OS — imgbb.com
Do you have any suggestion or I can trash it directly?
Last edited:
The challenge is to make it work. Trashing is only for those who do not like DIY and just want a cheap amplifier.
The 4K7 and 10K-pot tricks are only for amplifiers with the volume potentiometer connected at the TPA3116 signal coupling capacitors.
I see another chip (OP-AMP I assume) that is used for the bass/treble and includes the volume potentiometer somewhere.
The way to get rid of the persistent hiss is to ensure that the impedance just before the signal coupling capacitors is below 5K. It may also be noise from the bass/treble circuit with the OP-AMP.
The outputs of the OP-AMP are most likely pin 1 and pin 7. Which OP-AMP is it? Are you ready to draw-up the connections between pins 1+7 of the OP-AMP and the TPA3116? Without knowing these connections we work in the dark.
The 4K7 and 10K-pot tricks are only for amplifiers with the volume potentiometer connected at the TPA3116 signal coupling capacitors.
I see another chip (OP-AMP I assume) that is used for the bass/treble and includes the volume potentiometer somewhere.
The way to get rid of the persistent hiss is to ensure that the impedance just before the signal coupling capacitors is below 5K. It may also be noise from the bass/treble circuit with the OP-AMP.
The outputs of the OP-AMP are most likely pin 1 and pin 7. Which OP-AMP is it? Are you ready to draw-up the connections between pins 1+7 of the OP-AMP and the TPA3116? Without knowing these connections we work in the dark.
I assume pin 1 of the chip is in the upper right corner.
Input pins 4 and 5 are then going to the two brownish capacitors in the first row just above the chip, components number 3 and 2 counted from the right (the first is a black resistor). If you connect the sides of the capacitors turned away from the TPA3116 chip to ground through 4K7 the hiss in one channel should stop.
Alternatively, if you remove the black resistor in the first row just above the chip, number 5 counted from the right and number 4 counted from the left, the hiss should disappear. This, because you then reduce the gain to 20dB (minimum). That should work on both channels as the gain for both channels is lowered.
I must agree with "dzwer" that a TPA3116 design without output filter coils is not a good design. The datasheet mentions this possibility but the TPA3116 is too powerful to operate without a proper output filter and will generate a lot of emitted noise.
Input pins 4 and 5 are then going to the two brownish capacitors in the first row just above the chip, components number 3 and 2 counted from the right (the first is a black resistor). If you connect the sides of the capacitors turned away from the TPA3116 chip to ground through 4K7 the hiss in one channel should stop.
Alternatively, if you remove the black resistor in the first row just above the chip, number 5 counted from the right and number 4 counted from the left, the hiss should disappear. This, because you then reduce the gain to 20dB (minimum). That should work on both channels as the gain for both channels is lowered.
I must agree with "dzwer" that a TPA3116 design without output filter coils is not a good design. The datasheet mentions this possibility but the TPA3116 is too powerful to operate without a proper output filter and will generate a lot of emitted noise.
Sorry for the delay, thanks for the reply and suggestions.
What I wanted to imply is that if I have to spend 20/30 euros to make it works "right" maybe I can buy a better model, but even if I'm no expert I like to tinker with this type of things.
I bought this model because it fitted the holes that I already have in a pair of active speaker shelf.
The OP-AMP is a N5532, I circled in magenta pin 1 of the OP and the TPA.
What I wanted to imply is that if I have to spend 20/30 euros to make it works "right" maybe I can buy a better model, but even if I'm no expert I like to tinker with this type of things.
I bought this model because it fitted the holes that I already have in a pair of active speaker shelf.
The OP-AMP is a N5532, I circled in magenta pin 1 of the OP and the TPA.
An externally hosted image should be here but it was not working when we last tested it.
Many thanks. Then, what I wrote before should apply.
The hiss with TPA3116 often relates to too high gain-setting of the TPA3116 chip and too high impedance at the input. There I have suggested two tests/improvements. With this board the noise may also arise from the NE5532 tone-control circuit, in particular if high impedances are used. NE5532 in itself should not be the problem.
Hi,
I added more caption to the photo.
If I'm not wrong I could ground the circled end of the capacitors with 4k7 resistors for pin 4-5 (YELLOW, right channel) and 10-11 (RED, left channel) or I alternatively can simply remove the GREEN circled resistor: should I also put a jumper instead of the removed resistor?
Should I do something for the POP when I remove power?
Could I try to add some sort of output filters?
Thanks!
I added more caption to the photo.
If I'm not wrong I could ground the circled end of the capacitors with 4k7 resistors for pin 4-5 (YELLOW, right channel) and 10-11 (RED, left channel) or I alternatively can simply remove the GREEN circled resistor: should I also put a jumper instead of the removed resistor?
Should I do something for the POP when I remove power?
Could I try to add some sort of output filters?
Thanks!
An externally hosted image should be here but it was not working when we last tested it.
If I'm not wrong I could ground the circled end of the capacitors with 4k7 resistors for pin 4-5 (YELLOW, right channel) and 10-11 (RED, left channel) or I alternatively can simply remove the GREEN circled resistor: should I also put a jumper instead of the removed resistor?
Fully correct, but no jumper instead of the resistor. Page 14 of the datasheet http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf ,Table 1 - first line "Master 20dB": R2 OPEN should leave 20dB gain. In principle, R1 should be 5K6 but replacing an SMD resistor is much more difficult than removing one. Whatever R1 is at present, it may be sufficient to pull pin 8 down and give us 20dB gain.
Should I do something for the POP when I remove power?
Next problem to be solved when the hiss is no longer a problem. Still too much hiss and you may discard the board anyway.
Could I try to add some sort of output filters?
That's rather easy but again, first we solve the hiss problem.
NB: for the time being keep your speaker cables as short as possible (within reason). We don't want radiated noise from the speaker cables to influence the NE5532 part of the circuit.
Fully correct, but no jumper instead of the resistor. Page 14 of the datasheet http://www.ti.com/lit/ds/symlink/tpa3116d2.pdf ,Table 1 - first line "Master 20dB": R2 OPEN should leave 20dB gain. In principle, R1 should be 5K6 but replacing an SMD resistor is much more difficult than removing one. Whatever R1 is at present, it may be sufficient to pull pin 8 down and give us 20dB gain.
Should I do something for the POP when I remove power?
Next problem to be solved when the hiss is no longer a problem. Still too much hiss and you may discard the board anyway.
Could I try to add some sort of output filters?
That's rather easy but again, first we solve the hiss problem.
NB: for the time being keep your speaker cables as short as possible (within reason). We don't want radiated noise from the speaker cables to influence the NE5532 part of the circuit.
Ok, I removed the resistor circled in green and I also resoldered the original 50k pot volume, the hiss is almost gone and now it doesn't gain loudness increasing the volume but it's costant. Should I also try something like 4k7 to grund the capacitors? It's a bit more difficult with a limited space, also I don't have 4k7, but 4k and 680ohm in series, could I try with only the 4k?
Good, you are on the right track.
4K will do fine but I'm not sure how much extra you will gain. If not too much trouble, give it a try. The NE5532 circuit will add some noise but exactly how much I cannot say.
NB: When you test, short-circuit the amplifier board inputs through around 1K (one for each channel) so the hiss contribution from the NE5532 circuit is at least at a minimum.
4K will do fine but I'm not sure how much extra you will gain. If not too much trouble, give it a try. The NE5532 circuit will add some noise but exactly how much I cannot say.
NB: When you test, short-circuit the amplifier board inputs through around 1K (one for each channel) so the hiss contribution from the NE5532 circuit is at least at a minimum.
If R1 is already 47 Ohm, it is as good as a jumper to ground and there is no purpose in adding anything.
Last edited:
Why not buying a proven board equipped with filters instead? I have quite good experience with the red sanwu board DC 12V 24V TPA3116 Dual Channel Stereo 2x50W BTL Mono 100W Audio Amplifier Board | eBay . Decent input caps, easy to use balanced inputs to avoid ground loops, decent output inductors, quite thick power-signal traces.
hello
i made today another portable musicbox.
An old sony homecinema subwoofer with LG speakers.
Amp is dual TPA3116. One for the bass and one for the highs.
Always same problem. When getting louder amp stops playing.
I made a video with the problem.
And some pictures from the finished box.
Each side has 2 mediums, 1 tweeter and a crossover
Total 13 drivers
i made today another portable musicbox.
An old sony homecinema subwoofer with LG speakers.
Amp is dual TPA3116. One for the bass and one for the highs.
Always same problem. When getting louder amp stops playing.
I made a video with the problem.
And some pictures from the finished box.
Each side has 2 mediums, 1 tweeter and a crossover
Total 13 drivers
Attachments
Amp is dual TPA3116. One for the bass and one for the highs.
Always same problem. When getting louder amp stops playing.
I made a video with the problem.
It seems your power supply, an SMPS "brick"?, is powering the board directly. The board has only two decoupling capacitors (1000uF?) and that means the supply voltage may swing a lot such that the UV or OV protections are activated and causes the amplifier to disable for a period (self-clearing).
And some pictures from the finished box.
Each side has 2 mediums, 1 tweeter and a crossover
Looks good.
Always same problem. When getting louder amp stops playing.
I made a video with the problem.
It seems your power supply, an SMPS "brick"?, is powering the board directly. The board has only two decoupling capacitors (1000uF?) and that means the supply voltage may swing a lot such that the UV or OV protections are activated and causes the amplifier to disable for a period (self-clearing).
And some pictures from the finished box.
Each side has 2 mediums, 1 tweeter and a crossover
Looks good.
You don't need that large of a heat sink.
I sliced these from a computer cpu heat sink.
Pix is of amp that I take to Toronto DIY get togethers. ( dual PBTL )
I use one of my DUG-1 boards ( PBTL ) for the bass of a WAW active crossover ghetto blaster.
Data sheet says 60W 4R and just gets warm after cranking it up in the back yard
Attachments
FauxFrench
I added two 3300uF decoupling caps. Change supply. 19V Notebook PSU and 18V9Ah battery pack. Always amp shuts down.
I have with all tpa3116 this problem.
I will test tda7498 or STA326/STA508. But I'm not sure if it would help.
On the other side I have some powerful IRS2092 but the need symmetrical supplying. For portable use not good.
I added two 3300uF decoupling caps. Change supply. 19V Notebook PSU and 18V9Ah battery pack. Always amp shuts down.
I have with all tpa3116 this problem.
I will test tda7498 or STA326/STA508. But I'm not sure if it would help.
On the other side I have some powerful IRS2092 but the need symmetrical supplying. For portable use not good.