TPA3116D2 Amp

Account Closed
Joined 2010
If somebody is testing at bench one PBTL 3116 board, I’m asking him to hook an oscilloscope probe between ground and one of the two outputs and watch the waveform (with a sine wave at the input).
Doing this with two different PBTL boards, I saw, for each of them, a very distorted waveform, despite the differential output voltage was clean (distortion is more evident at frequency higher than 10kHz); it seems like the harmonics generated by one of the half-bridges were cancelled by the harmonics produced by the other half-bridge.
I’ve never seen something like this in BTL 3116 boards.
Did anyone confirm your findings Sir?
 
Account Closed
Joined 2010
It's silicone.......I got the heatsink off of one of the other boards no problem; it had a lot less silicone. Here's the dead board.
Is that chip glued tightly to the heatsink? I thought silicone grease isn't glueing...
That board has tl074 op-amps.What't that doing? Do you have a schematic? Now that you sacrificed that board , would you be kind to reverse engineer the op amp section? I have an identical board and I'd like to know what can be done to make it work better if possible.
 
I think they mean the audio from the bluetooth stops and you get audio from the AUX, into the amp/speakers, when you connect the AUX line. It automatically detects the AUX signal, but I don't think the bluetooth turns off.
the bluetooth always seems to stay on, regardless if the unit is switched off. Its drained my battery a few times. i was lead to believe that you can solder the two points below the pot to close this, but i dont know how true mr youtube is
 
These ae a very good implementation of the PAM8406 chip, which can run in D Class (default) or A/B Class.

I have used them to build active speakers and to convert vintage radios into BT speakers (there is a 'matching' BT board which works very well).

On a decent phone charger wall wart they are totally silent, and will run for hours on a 10000mAH battery bank.

You can get them for under $5/5€ on AliExpress if he Amazon pice is too high!

https://www.amazon.com/PAM8406-Ampl...sr_1_3?keywords=pam8406&qid=1636442960&sr=8-3

Edit: Here's he matching BT board:

https://www.amazon.com/Bluetooth-DR...47856&pd_rd_wg=Otr81&pd_rd_i=B07P94Z9XR&psc=1
i would suggest getting a QCC3003 bt board, or an equivlent in the CSR range. I use it for my 5v builds and get that magic 5w per channel when you use a buck boost converter from 3.7v to 6v. Sounds pretty good to for a low fi setup

https://techtalk.parts-express.com/...p-inspired-3d-printed-ghettoblaster-half-size

my latest build uses it and i get 10 hours run time from a 3000mah 18650 battery....its my own fault as well as i put 4 LED's running always which each draw 20mah, so could have got longer if i wasnt going for the fancy look.
 
  • Like
Reactions: 1 user
XH-M590 I just finished the modifications mentioned in the previous posting and the results were as expected. This version of the XH-M590 is set for a gain of 36dB using a 75K and a 47K resistor (master, 36dB). This setting is the same for both TPA3116 chips such that the two chips are not synchronized but run at close to 400KHz (carrier frequency) individually.
Changing the gain to 20dB (as master) just requires the two 75K resistors (marked with "753") to be removed. That requires the heatsink to be removed. Removing the heatsink I realized that no thermal paste was used between the TPA3116 ICs and the heatsink. If you remove the heatsink and find that no thermal paste is used, give it a little on both ICs before putting the heatsink back.
after removing these two resistors, ... should i leave those points open? thanks
 
i would suggest getting a QCC3003 bt board, or an equivlent in the CSR range. I use it for my 5v builds and get that magic 5w per channel when you use a buck boost converter from 3.7v to 6v. Sounds pretty good to for a low fi setup

https://techtalk.parts-express.com/...p-inspired-3d-printed-ghettoblaster-half-size

my latest build uses it and i get 10 hours run time from a 3000mah 18650 battery....its my own fault as well as i put 4 LED's running always which each draw 20mah, so could have got longer if i wasnt going for the fancy look.
pam8965 have built-in booster . This should meet your requirements
 
low budget audiopile
TPA3116D2 dual mono block :)
PreAmp Mose2 x2
Juma CM single psu
xFormer 18v
nice clean flat low noise
photo_2022-11-27_09-08-23.jpg
 
Last edited:
Fake TPA3116D2DADR board XH-M543 from Aliexpress

The seller is NK-Shopping Store

The heat sink was not firmly attached and seemed to be glued with some kind of silicone, so I pulled it. There was not need to do it forcefully and the heat sink came out to find beneath a SOIC16 chip labeled TPA3116D2DADR. Some photos of the board and PLEASE UPDATE # 1 with photos of good and fake pcbs
Does anyone knows what the real chip is this ? And maybe some datasheets

Elias

1673478772474.jpg
1673478772465.jpg
1673478772482.jpg
 
I have an identical kit that works...never thought opening it...It might be the tpa3118 relabeled .They are identical chips with different placement of the heatsink.I actually wondered how much heat can extract a good active heatsink like a peltier out of a 3118 chip so that i could use it at the same output power as 3116.
 
I have an identical kit that works...never thought opening it...It might be the tpa3118 relabeled .They are identical chips with different placement of the heatsink.I actually wondered how much heat can extract a good active heatsink like a peltier out of a 3118 chip so that i could use it at the same output power as 3116.
No, there is no TI SO-16 class-D-chip. These are chinese chips with a fake label on them. Sometimes two of these a below the heat sink of a "TPA3116" board. This has been shown here, maybe in another thread.
 
I have an identical kit that works...never thought opening it...It might be the tpa3118 relabeled .They are identical chips with different placement of the heatsink.I actually wondered how much heat can extract a good active heatsink like a peltier out of a 3118 chip so that i could use it at the same output power as 3116.
I think it is more likely a chipstar cs8673e... http://pdf1.alldatasheet.com/datasheet-pdf/view/1143318/YONGFUKANG/CS8673E.html

I have seen some supposed tpa3116s that were actually the chipstar chip and not labled with the BS DADR (which is not even a good way to distinquish it given that DADR is a valid designation for the real 3116 chip in a specific package and ordering format from TI)

I tested the same board above and got 13.3v RMS with a 24v DC supply before clipping with a 1k sine wave into an 8 ohm resistor so it was about 22W. I have no idea how long it could handle that load. I don't think I tested it with an 4 ohm load but I could do it sometime. The real tpa3116 can get just about 28w into 8 ohms IIRC.

Note that I believe the opamps on these are also fake tl072s. I think there was a way to test by sending a fast square wave through them. I will have to try that somtime.... :) I found it better to bypass them, they add WAY too much gain on the real tpa3115 board. I am not sure what the gain is on the fake one.
Get your money back unless they stated that it was the "domestically produced chip" Some suppliers are honest and will send you a real chip if you pay a few dollars more where others are just scamming.

-Rich
 
Last edited:
  • Thank You
Reactions: 1 user
Ok...never looked closely at the chip, only bought a few kits , supplied and played them...I only saw that tpa3118 has the heatsink underneath the chip on the pcb side unlike 3116.Can anyone confirm if wiith better active cooling tpa3118 can perform identically to 3116?
TPA3116 and TPA3118 are identical chips. With loads of 8 Ohms - there is no problem with the TPA3118 without dedicated heatsink. I use them this way for some time in my guitar amps.With 4 Ohms or even 2 Ohms load I would consider TPA3116 with a heatsink.