TPA3255 - all about DIY, Discussion, Design etc

Hi, mates again. Still looking for more loud from the tpa3255. Cant say it is no too much loud but small edifier r2700 produce more bass than canton sc-l with my tpa3255. Consider rather small main capacitor cant produce enough peaks. What do you think? I see original board contrain two big capacitor, my have only one with 1000mF/50v noname CpXon .


i might be, that the GAIn of the edifier active speaker is quite high or higher then the 27dB of the TPA3255 board.
i had 8 boards (some damaged :rolleyes::D) and i have never the problem that is not to loud !! PC--> DAC-AMP-speakers with 85-89dB

as Think wrote...check your settings from the quipment before the amp if volume settings are powerd up..

cap:

its 1000µF not 1000mF ;) big difference...in listening room you should not have trouble with bass. yes you should add as closed as possible to the TPA chip a low ESR (the original are more or less not low ESR caps) about 1500µF on both rails.

...but sorry without the pic without heatsink its not easy to say where.
i personally recommend to read the Texas Instrument datasheet of the TPA3255 chip and the TPA3255EVM board to learn more about the class D world
(too large to attach it here...)


chris
 
The TPA3255 chip on my evm board just died with a bang (no smell?); pins 29-31 and 36-38, PVDD_CD and PVDD_AB are gone.


This happened after about a week listening at 24V and about 15 minutes after raising the voltage from 24V to 48V with a voltage booster.
To run in new drivers, it was playing a 25Hz sine wave a minute before it died, with 16.2V on the outputs with 8ohm speakers connected so that was about 2 times 33W and I was wondering why I saw only 0.940A ~45W going into the board.
While it died, there was only 1.6V on the outputs as I just had switched sources and had not turned the new source up.

I'm puzzled and disappointed, will make pictures with daylight and clean it up after that.

The heatsink has some burnmark on the side and C33 has a little bevel like the heatsink has been toughing it. It and the 3 similar caps next to the chip measure ok.

//edit: The board was as delivered, no mods, not removed the heatsink before, just listened to it and liked it very much; especially after waiting so long for it. :(

//edit2: the voltage booster now seems to have a problem as well as it does not put out any voltage any more and the used 24V power adapter led fades when I connect it. The adapter seems to be ok.
 
Last edited:
I use a transformer from an old receive to power my TPA3255 with 40,000uf. No load voltage is 56VDC. Full load is around 50VDC. I have tested the amplifier both channels driven well into clipping with 4 ohm resistive loads and no problems. I have had it for over a year and quite often its on for days at a time, no problems so far. Ocassionaly when I have had too much wine things can get pretty loud. My hifi buddy says I am just showing off.

Maybe you had some bad luck with yours.
 
Thx, bad luck for sure... :p

But it looks like 'only' the 2x 3 legs to the chip are fried. The 2x3 pins are paralleled and all have a piece left going into the chip; they are the power pins going in to the chip.
The chip might still work.... and I might be damn lucky and be able to fix it with 2 blobs of solder, replacing the 2x3 pins. Looks at least a lot easier then just re-soldering 1 pin as they are so tiny.

The voltage booster reads 2.5Ohm at it's + and - inputs ..... that died and seems to have shorted itself in the proces.

But what happened?
1) Could it be that the voltage booster put out a much higher voltage while in the process of dying and only fried the power legs to the chip?
2) Where did the 'bang' (baby firecracker?) sound come from? Was that just a voltage spike frying the 6 pins?
The voltage booster does not have any visual sign or smell of damage and it's 15A fuse at the input is still fine. All caps seems to be fine.
All ideas welcome.

//edit: 3) there are 2 solder puddles left at pins 29-31 and 36-38 (job well done :D), should I just reconnect the 2x3 tiny ends sticking out of the chip, with a blob of solder?

TypAppPBTL.gif
 
Last edited:
After a long time following this thread, I decided to order a TPA3255 board, and it just came yesterday :) Very nice!
Thank you all for sharing your knowledge and passion!

BTW, I'm going to add a potentiometer to control volume, so which value to chose? Is a 50k Ohm one will work? Thanks!
 

Attachments

  • 20190221_210220.jpg
    20190221_210220.jpg
    766.6 KB · Views: 794
Last edited:
Thx, that makes sense.
Offtopic; probably the same season I could not get the picture thing right even though I used to make websites in the early days.... it should be simpler now... somehow or is it only me who got simpler as time goes by? :?
Tinypic was down, so I thought let's try something else.
 
So, in the end they found together, TPA3255 2.1 100x100mm system in direct contact with the 1200W 4-phase LTC3784 boost-converter board, making it possible to go the distance with even only having 10V at the input to put out 600Wrms. As this asks for 1200W power into the amp and an realistic efficiency of 92% from the boost converter, this setup is asking for 130A at the 10V input. (Things get much easier when using a 36V supply) For the AFE module and bulk caps there is a mounting holder within half of the height to make good support for stability while driving with the bike. The whole thing measures 110x110x70mm, i'd call this a power-brick.

attachment.php


attachment.php


attachment.php


attachment.php


Bike? The bike with the actual setup, TPA3255 + 3x12V SLA system...

Youtube: YouTube
 

Attachments

  • IMG_20190226_134652.jpg
    IMG_20190226_134652.jpg
    147 KB · Views: 2,786
  • IMG_20190226_134754.jpg
    IMG_20190226_134754.jpg
    186.1 KB · Views: 2,781
  • IMG_20190226_134907.jpg
    IMG_20190226_134907.jpg
    143 KB · Views: 2,739
  • IMG_20190226_134943.jpg
    IMG_20190226_134943.jpg
    140.5 KB · Views: 2,758
Last edited:
Very nice!

1) Did you do some measurement of maximum continues power draw?
2) Is passive cooling OK with 36V?
3) How many Ah are the SLA's and do you have an indication of run time?

To power a 600W AB car amp at high volume, a laptop and a tiny usb powered mixer for 8 hours I needed to bring 2x 12V 105Ah deep cycle batteries (2x20+kg) .... and could warm my hands on the amp. :) As I could only draw them to 50%, that is only a bit more then 1kwh.

I intend to use a 36V 8.4AH (0,3kwh and 3kg) ebike battery which was given to me as defective but is fine with only a new fuse on the charging port. :)

I estimate it will run more then 8 hours.



The Hooverboard /Oxboard (2 wheel kids balance toy) batteries might be interesting to use. They are 2p10s 18650's, 36V 4.4AH often Samsung cells and sell for €40. I know an address to pick up defective packs for €5 of which most prob. need a new bms as they are tortured (rigorously tested :) ) by the 2x 250w motors going back and forth.
 
Short answer first:

1: What measurements are you looking for, amp or boost? i have plenty of documents/measurements done

2: yes, even in an closed enclosure

3: I use 3x5Ah deep cycle cells, so about 60Wh which will do about 5-6hours at early full power. (The speakers are better than 96dB. With the all new 36V 10s4p 11.8Ah pack which is about 420Wh, the amp theoretically will run 42 hours.. :D but i want to use them in a 30-80% manner, so charging only to 4.00V and discharge only to 3.3V per cell at normal.

The 10s2p packs might struggle on peak power performance, if not the cells, then the BMS. I have one of these packs here, with fake Samsung cells and rusty BMS, sold as brand new...