Great video, Hans,
but if you go through the work of making such a nice film, please take some more time and show all of your installation, like wires, power source, batteries etc.
Just the whole desk.
Very often the problem lies in some secondary detail no one thinks about.
Der Teufel lauert im Detail sagen wir...
I may think the power source is somehow slow, as the clipping rises a bit, which is unusual to me. Batteries are not very fast, but I have never noticed such an rising in voltage under load, sinking would be more logical. Can you put the voltage of the power supply on the second channel?
Could it be some protection feature of the board? Some D-amps have programabel limiters you can activate
e in hardware, but I´m not aware the 3116 has any?
Impedance of the speakers is of great interest too, maybe some miss match. Has anyone else such a board?
but if you go through the work of making such a nice film, please take some more time and show all of your installation, like wires, power source, batteries etc.
Just the whole desk.
Very often the problem lies in some secondary detail no one thinks about.
Der Teufel lauert im Detail sagen wir...
I may think the power source is somehow slow, as the clipping rises a bit, which is unusual to me. Batteries are not very fast, but I have never noticed such an rising in voltage under load, sinking would be more logical. Can you put the voltage of the power supply on the second channel?
Could it be some protection feature of the board? Some D-amps have programabel limiters you can activate
e in hardware, but I´m not aware the 3116 has any?
Impedance of the speakers is of great interest too, maybe some miss match. Has anyone else such a board?
Last edited:
Is 100Hz, one scope-probe on one output channel (AC-coupled) and the other probe on the supply voltage, possible? With one output, we can see when the output signal starts clipping and then it is interesting what the supply voltage does.
Ifisrt mesure 100Hz sinus with channel 1 and then unplug ground and take channel 2 to input.
Amp gets really hot. Never feel so hot.
The big noise on channel 2 is with saturation of output.
Amp gets really hot. Never feel so hot.
The big noise on channel 2 is with saturation of output.
Attachments
Around 2.5Vpp ripple. That should explain why the clipped tops are not horizontal. The voltage is simply dropping.
Did you measure the impedance of the speaker on that channel with an Ohm-meter? What is the DC impedance of the speaker?
@FauxFrench: Speaker is 2ohm. But I hade same Problem with 4ohm and 8ohm speaker to.
I reconnect fist TPA3116 dual mono from ali with 33uH chokes and it seems to work. I the past I have problem with this amps with high level output. But fast test yesterday was working great to max volume
@Turbowatch2: Internalresistance of battery is 100mOhm
I reconnect fist TPA3116 dual mono from ali with 33uH chokes and it seems to work. I the past I have problem with this amps with high level output. But fast test yesterday was working great to max volume
@Turbowatch2: Internalresistance of battery is 100mOhm
Attachments
One good thing is that something starts working.
With a 2 Ohm speaker and sine-wave peaks of 16V as seen before, the peak currents in the speaker is 8A and the rms current 5.6A. The 19V laptop supply can hardly handle 5.6A rms as it is probably a 90W supply such that the voltage is sagging and that can cause the amplifier to cut-off.
For the battery, for which Turbowatch2 is more competent than I, it seems that there are two different impedances: a static impedance of about 100mOhm (as you measure) and a dynamic impedance (from reaction delay) which is higher. It is very normal that a voltage source has a dynamic impedance higher than the static impedance.
When you use a 2 Ohm speaker for the sub-channel and play at full power, you already have an rms current of 5.6A that may exceed the capacity of the power supply and 8A that is at the limit of what the TPA3116 chip can handle.
When you then use the two stereo channels at the same time, the current for these channels add to the sub-channel current such that the total current further overload the supply. Luckily the amplifier seems to use 2 TPA3116 chips such that the two may handle the current but the supply not.
If you want the amplifier not to cut-off, you should for a start use 8 Ohm speakers on all three outputs and see if the amplifier then cuts-off. Then, you may use 4 Ohm for the sub-channel and see if that works without cut-off. Then 4 Ohm on all channels etc. It is important that you start from something that works and then lower the speaker impedance until it does not work. The other way around, using low impedance speakers right away, realize the amplifier cuts-off at a certain level and ask around what is the problem is more difficult to handle.
I believe you use your amplifier at a higher power than most of us. Therefore, you run into current problems with low impedance loads.
In such case, I would use the 19V/230W supply because it is very difficult to find faults in a circuit if more reasons for failure (sagging supply voltage) are present at the same time. I would also keep an eye on the supply voltage when you increase the power and use as many decoupling capacitors you can reasonably connect if the supply voltage shows important variations. With more than 8A current, the decoupling capacitors are working really hard and even the 230W power supply may have problems coping with the current variations.
Hi, mates. Here is my quick story about digital boards.
1.Aiyama tpa3255 complete unit with Bluetooth onboard at 50$
2. Noname tpa3116 double chip board only at 8$.
Loaded 8hom Canton SC-L 200w din power 89db sensitivity.
My review nr1 not worth to buy for 50$ while nr2 is available at 8$. Firstly nr1 died after 2month without reason, nr2 deliver more power and bass even with notebook adapter while tpa3255 power much more adaptor 50V/20A dc block.
Second step buy second tpa3116 board to get biamp. Thus in total biamp setup would cost 8$+8$+14$(two notebook adapter s)=40$. Perhaps aptx Bluetooth module would cost extra 20$. Thus in total 60$.
Meanwhile I still believe tpa3255 should beat tpa3116 , perhaps my case was unlucky circumstances. Will try reference board or close when they drop to 40$ per board.
Pictures for reference.
1.Aiyama tpa3255 complete unit with Bluetooth onboard at 50$
2. Noname tpa3116 double chip board only at 8$.
Loaded 8hom Canton SC-L 200w din power 89db sensitivity.
My review nr1 not worth to buy for 50$ while nr2 is available at 8$. Firstly nr1 died after 2month without reason, nr2 deliver more power and bass even with notebook adapter while tpa3255 power much more adaptor 50V/20A dc block.
Second step buy second tpa3116 board to get biamp. Thus in total biamp setup would cost 8$+8$+14$(two notebook adapter s)=40$. Perhaps aptx Bluetooth module would cost extra 20$. Thus in total 60$.
Meanwhile I still believe tpa3255 should beat tpa3116 , perhaps my case was unlucky circumstances. Will try reference board or close when they drop to 40$ per board.
Pictures for reference.
Attachments
When you evaluated the TPA3255, did you use the Bluetooth or did you use another input? Both should the Bluetooth protocol include compression, next does the bluetooth chip include a DAC that may be of more mediocre quality.
The part you test must be the weakest element in the chain for the most reliable result.
The part you test must be the weakest element in the chain for the most reliable result.
Is that a recommendation based on your experience?
Honestly, I do not like that board. The input pot is right next to the power supply terminals where all the large currents flow, the noisy outputs are traced around half of the board back to the inputs with quite thin traces, electrolytic (very likely polarized) input caps, etc.
Honestly, I do not like that board. The input pot is right next to the power supply terminals where all the large currents flow, the noisy outputs are traced around half of the board back to the inputs with quite thin traces, electrolytic (very likely polarized) input caps, etc.
Member
Joined 2018
Member
Joined 2018
hansueli-San,
I didn't change output inductors on this board, but I changed switching frequency from 400kHz to 1.2MHz instead of inductors change.
Additional modifications are...
+ Add voltage follower buffer in front of Chip input. (To reduce remain noise at the middle of volume position.)
+ Add active muting driver circuit.
+ Increase input coupling capacitance.
- Jump the Reverse Voltage protection Shotkey Diode .
I have another HX-M139 2.1ch TPA31116D2 Board. That board has terrible quality yellow toroidal core inductors (47uH@Low current)
That 47uH core saturates very easily then lost it's inductance. finally it reached to burn-out
Details information will be found following URL...
Destop 2.1 channel Amprefire Modification
Regards,
HILO
I didn't change output inductors on this board, but I changed switching frequency from 400kHz to 1.2MHz instead of inductors change.
Additional modifications are...
+ Add voltage follower buffer in front of Chip input. (To reduce remain noise at the middle of volume position.)
+ Add active muting driver circuit.
+ Increase input coupling capacitance.
- Jump the Reverse Voltage protection Shotkey Diode .
I have another HX-M139 2.1ch TPA31116D2 Board. That board has terrible quality yellow toroidal core inductors (47uH@Low current)
That 47uH core saturates very easily then lost it's inductance. finally it reached to burn-out
Details information will be found following URL...
Destop 2.1 channel Amprefire Modification
Regards,
HILO
