TPA3116 D2 mono amplifier as a stereo system?

Hi,
I was also searching for a new Class D, because I got tired of the heat and the inconvenience of the JLH1969. :)
I ordered the same 2 boards and the power line decoupling module. The PS will be the Meanwell RSP7524 or the 20024 model, I don't know yet. I have both lying around.
Input selection will be through this:
3P4T 3 Pole 4 Position Single Wafer Band Selector Rotary Switch w Knob-in Switches from Lights & Lighting on Aliexpress.com | Alibaba Group
Because I have one at home. I wonder, however, if a board like this would be better in any way:
Audio Input Signal Selector Relay Board Signal Switching Amplifier Board RCA For Speakers-in Amplifier from Consumer Electronics on Aliexpress.com | Alibaba Group


The volume will be regulated through a 50K Alps pot.

An I doing any obvious misakes in my plans?


@AudioMech73, did you recieve any of the Chinese goodies yet? Tried them?
 
I have never seen the Signal Selector Relay Board before. I doubt you can hear any difference whether you use the simple selector switch or the more sophisticated board. For the selector switch, you should probably use 22KOhm resistors to ground for all sources.

I only see one possible issue: the 50KOhm Alps. 50KOhm at the input of a TPA3118 with high gain may leave hiss. I use 10KOhm Alps. Or, you can lower the TPA3118 gain.

I used to have the JHL1969 back in the days where it was quite unique in sound. But, it was bulky and heated a lot. It was surprisingly simple and for that period very performance competitive. Today, it is a reminiscence from the past. They can be fun to try but personally I find that other modern and less impractical constructions can do as well. I took mine apart many years ago but kept the important parts, like the massive heatsinks, for new challenges.

NB: How much capacitance a Meanwell will accept at the output from the decoupuling module, I do not know. I have no Meanwell myself. Nigel warned that this could be an issue.
 
Last edited:
I have never seen the Signal Selector Relay Board before. I doubt you can hear any difference whether you use the simple selector switch or the more sophisticated board. For the selector switch, you should probably use 22KOhm resistors to ground for all sources.

I only see one possible issue: the 50KOhm Alps. 50KOhm at the input of a TPA3118 with high gain may leave hiss. I use 10KOhm Alps. Or, you can lower the TPA3118 gain.

I used to have the JHL1969 back in the days where it was quite unique in sound. But, it was bulky and heated a lot. It was surprisingly simple and for that period very performance competitive. Today, it is a reminiscence from the past. They can be fun to try but personally I find that other modern and less impractical constructions can do as well. I took mine apart many years ago but kept the important parts, like the massive heatsinks, for new challenges.

NB: How much capacitance a Meanwell will accept at the output from the decoupuling module, I do not know. I have no Meanwell myself. Nigel warned that this could be an issue.


22k resistors to ground? What do you mean? Have you got a schematic? What do they do?


If the smps' won't take the capacitance, I could always just use a 230/24VAC trafo and a Graetz, right? Bulkier and less sexy:(
 
22k resistors to ground? What do you mean? Have you got a schematic? What do they do?

If the smps' won't take the capacitance, I could always just use a 230/24VAC trafo and a Graetz, right? Bulkier and less sexy:(

The 22KOhm resistors serve to bias the output capacitors in the sources correctly (so you don't have a "plop" in the speakers when you change source) if the sources have not already solved that problem.

A 230V/24Vac will give you more than 32V when rectified! If you need to use a transformer, it is rather 230V/15Vac you need.
 

Attachments

  • InputSelector.jpg
    InputSelector.jpg
    23.8 KB · Views: 472
Last edited:
Disabled Account
Joined 2018
@peroz I have not received the parts yet.

Once I receive them and the whole system has been assembled, I will post some pictures. :)

I first thought about building a JLH1969 system for the unique sound, but the heat sink requirements are a bit crazy and you don't get much output power.

Maybe at some point...
 
Last edited:
The guy in the YT review mentions he uses a common ground for his speakers (big no no for class D and running in BTL) and at first he did not ground half of the differential input...... doing it did not even change the sound; LOL
I have used this board for a subwoofer and it sounds fine!
\\ lastest (?) version of this board has a 3300uF 25V 'nichicon' cap
\\\ I would not spend much money on the power supply: 24V does not give much more then 19V as 3dB down from max volume allready uses half the power. I like the 19V 3.14A HP laptop adapters best and they seem to be powerful enough even for my 2.1 dual chip board with 2Ohm (pbtl) subs and 5Ohm tops.

I received my TPA3116D2 board today. Only one (mono) for test. Nothing to complain about.
Started with 12V supply and cheap test-speaker - worked. Then, 19V supply (laptop brick) and cheap test-speaker - worked as well. Finally 19V supply and a good speaker (only mono) - absolutely nothing to complain about with the input tied correctly.

As my source is not differential signal, I tied one of the inputs ("-") to ground. No hiss and no hum. Then, for curiosity, I untied the negative input from ground and got noise. I did not try for long because playing with an open input is to tempt Murphy's Law. It indeed needs to be tied if the source is is not differential.

I only have one small jumper on the board I am at present not sure what is doing. It is marked with "M" and "GND".
All-in-all a fine board at a very attractive price. Good heatsink that can be removed. Assembly standard is average.
 
Last edited:
@FauxFrench please check my video. It's 3.5mm jack.
YouTube

Thanks for the video. As far as I can see, you touch with your fingers the two input signal lines of the amplifier. The hum your body collects is transmitted to the amplifier input for amplification. It is fully normal you have hum at the output when you do so.
When you plug the jack into your smartphone and touch around the jack when inserted in the smartphone, do you then have buzzing (noise)?
 
Thanks for the video. As far as I can see, you touch with your fingers the two input signal lines of the amplifier. The hum your body collects is transmitted to the amplifier input for amplification. It is fully normal you have hum at the output when you do so.
When you plug the jack into your smartphone and touch around the jack when inserted in the smartphone, do you then have buzzing (noise)?

It's the mono amp so just one input signal and gnd. If i touch input (+) it makes buzz sound, but if touch gnd it dont make any buzz. Everytime while i plugging jack i can hear hiss and loud buzz. After music starts the buzz disappears.
Maybe it's possible to add some kind of capacitor in signal line?
 
It's the mono amp so just one input signal and gnd. If i touch input (+) it makes buzz sound, but if touch gnd it dont make any buzz. Everytime while i plugging jack i can hear hiss and loud buzz. After music starts the buzz disappears.
Maybe it's possible to add some kind of capacitor in signal line?

It seems your smartphone output is high impedance when no music is played. That is clever when you want to economize power. When you have a high impedance at the input of a TPA3116 amplifier, it tends to make hiss and be sensitive to hum.

Your TPA3116 amplifier board uses green screw terminal connectors, also for input. At the input terminals where your signal input wires are connected, you also connect a 4K7 resistor between the two terminals (signal input and ground). This ensures an impedance of no more than 4K7 at the TPA3116 input. That should keep the amplifier quiet.
 
Last edited:
Disabled Account
Joined 2018
I received my TPA3116D2 board today. Only one (mono) for test. Nothing to complain about.
Started with 12V supply and cheap test-speaker - worked. Then, 19V supply (laptop brick) and cheap test-speaker - worked as well. Finally 19V supply and a good speaker (only mono) - absolutely nothing to complain about with the input tied correctly.

That sounds promising. Can't wait to receive my parts... :)
 
Issues with the read mono TPA3116D2 PTBL amplifier board.

Recently I had to test a 24V SMPS for continued operation after repair. I used my red mono TPA3116D2 board (PBTL) as load. The 24V supply worked well but I noticed some important issues with the TPA3116 board.

After operation for about an hour, the output filter chokes became pretty hot. That is unusual compared to more other TPA3116 boards I have, also running PBTL.
I did not know the 10uH filter chokes so I changed the chokes into a type I know, assuming the heating could be due to excessive core losses. The change only improved the situation very little.
I put my oscilloscope on expecting that the operational frequency would be higher than normal (400KHz). Actually, with only 120KHz the operational frequency was much lower than normal. Realizing that the operational frequency is set by three chip input pins (AM0-AM2; pins 13-15) only and they were all grounded for 400KHz, I concluded it had to be a chip error. I changed the TPA3116 chip but even with the new TPA3116 chip, the frequency remained 120KHz.
Then, I studied the layout and concluded that pin 16, “SYNC”, was grounded. In “master” configuration, the SYNC pin is an output. Strange.
Following, I looked at the operational settings for master/slave and gain (two resistors described in Table 1 of the datasheet). With in-circuit measurements of the resistances, it seemed the configuration was “slave” and 36dB gain. I swapped the two resitors (R2 and R3) around on the board. Then, I had a the correct 400KHz operational frequency.

It seems that a number of these boards have a PCB error such that the SYNC output is shorted to ground while it should be left open. And, at least for the batch of which I have an item from, the two setting-resitors (R2 and R3) have been chosen wrong such that the amplifier is operating at an idle frequency of 120KHz and configured as “slave”.

If your chokes on that 3116 board tend to get hot, check if you also have a 120KHz operational frequency. If so:

  • Cut pin 16 on the chip such that there is no longer a connection to gound and pin 16 is open,
  • Swap the two resistors R2 and R3 around or remove R2. (R2 on the board corresponds to R2 in Table 1 (datasheet) and R3 on the board to R1 in Table 1.
Else, I find it a fine board with balanced input option, film input capacitors, correct 10uH filter chokes, a good heatsink that can be removed and good power decoupling (3300uF). A pity that such mistakes are implemented after a good start.

The board has now been tested for more than an hour at 24V and all appears normal like my other TPA3116 boards.
 

Attachments

  • TPA3116Board1.JPG
    TPA3116Board1.JPG
    791.5 KB · Views: 453
  • TPA3116Board2.JPG
    TPA3116Board2.JPG
    642.9 KB · Views: 490
Last edited: