TPA3255 - all about DIY, Discussion, Design etc

Jump to Latest
A

Audio Mo

Member
Joined 2019
Hi everyone

I'm still researching (still learning here) how to power 6 x TPA3255 in one box so far I narrowed it down to 2 options:

1. SMPS600RS which I'm currently using to power 1 board and it's really good. However if I put 6 of them, current demand will exceed IEC320 C14 limit of 10A @230v, which means I need to use 2 power inputs to feed the SMPSs in two groups, which I prefer to avoid. I don't think they can be parallel connected (current sharing) either.

2. HRPG-1000-48 x 2 in current sharing setup to power 5 boards + 1 SMPS600RS to power the 6th. The issue is those bad boys are big and have built in fans that will be audible. Data sheet does not mention fan noise levels.

Amp boards will be in Master/ slave setup, will be installing quite fans to get the heat out.

Target here is to power 7.2.4 8Ohm speakers (6 of them are floor standing that I accumulated over the years)

2nd option is an expensive one and will blow my budget for this project.

Should I change the setup and use fewer amps? Would appreciate if someone can advise or suggest a better solution.



Thanks
 
W

wealas

Member
Joined 2017
I don't know what speakers you have and at what volume levels you plan on using them but you should probably hook up a couple of speakers and measure the power draw from the SMPS for few types of music and volume levels. I won't be surprised if a single power supply for all of the satellite speakers, one for the subwoofers and one for the front speakers would leave you with plenty of headroom to spare.
 
A

Audio Mo

Member
Joined 2019
wealas and Chris thanks, this is very informative.

So in real world application, so to speak, whether it's a home theater or music one would only need 1125w continuous SMPS (or say 1200w) @ 48 volts to power 12 channel amp using 6 x tpa3255 with an output of 150w per channel @ 8Ohm. Is that correct?

1200w input can produce 1800w output? Just sounds too good to be true for a noob like me
 
C

chermann

Member
Joined 2017
Audio Mo said:
wealas and Chris thanks, this is very informative.

So in real world application, so to speak, whether it's a home theater or music one would only need 1125w continuous SMPS (or say 1200w) @ 48 volts to power 12 channel amp using 6 x tpa3255 with an output of 150w per channel @ 8Ohm. Is that correct?

1200w input can produce 1800w output? Just sounds too good to be true for a noob like me
Click to expand...


hi

i dont think you need 1200W music power over your speakers... some 20W- 50W pure music is f** loud !:eek:


if you write
1200w input can produce 1800w output...it looks like we found here a magic source of power;):D its physic => input - losses = output
but for music listening and look at the crest factor....yes

from the link i posted:

The classical way assuming test with a constant amplitude sine-wave: The maximum output power from the amplifier is divided by 0.8 (80%) to take amplifier losses into account. Using the values from the example above, 72W in 8 Ohm requires a power supply of 72W / 0.8 = 90W. In 4 Ohm, 144W / 0.8 = 180W.
The realistic way assuming use with music: The power supply capacity (Watt) calculated above is divided by two. Why? Because music has got a “crest-factor” meaning that no music is demanding full power for long. Actually, most music has long passages with rather quiet sound and short passages with powerful sound. Therefore, it is seen as unrealistic to design the power supply for maximum sound level all the time. Clever and competent people estimate the steady-state power supply need to be half of what is normally needed for the maximum power the amplifier can output. The power line decoupling capacitors are handling the need for short but powerful surge currents.
Using the example above with 72W in 8 Ohm will then require only a 45W power supply (half of 90W)

chris
 
xrk971

xrk971

Founder of XSA-Labs
Joined 2012
etsy.com/shop/XRKAudio
Paid Member
Kokanee said:
Hi X,

2 other threads may have info in them: "TI TPA3255EVM by Kay Pirinha", page 3 in this forum, and "Whats wrong with TPA3255 by diytoon", page 6 in this forum. The first one is shorter and I believe talks about resets and relays used in construction.

Regards
Click to expand...

Thanks, Kokanee. I will look at these other threads. I think a conversation with Dr Mord, the resident TPA3255 expert is also in order!
 
C

chermann

Member
Joined 2017
xrk971 said:
I have a question about the TPA3255’s auto protection and going to fault mode:
I’ve been testing a generic (“Nobsound”) TI reference design of the TPA3255 and noticed that it is very sensitive to input pops and clicks from source material. It will go into fault mode and require power cycle to reset. That’s a problem for me - is it simply the reference design or inherent in the TPA3255 chip itself? I realize it is there to protect speakers - but there has to be a better way to reset (maybe automatically?). The pops were small and came from my DAC switching tracks or changing but rate etc. not ideal but no other amp I have locks me out.

Testing it here with a 12v (13.8v nominal) lead acid gel cell powering a 600w DC-DC step up, followed by a high current CLC with flat wire CoilCraft.
Click to expand...


post 1120

you are mearure without diff scope probe?
 
F

FauxFrench

Member
Joined 2017
Hi Kay,

There are better experts on this forum than I.
I have the board you show. Mine has been modified somewhat.

First the price, I paid close to 46 EUR for my board.

The board sounds well in the configuration as sold. That does not prevent DIY's from finding details they believe could be improved.
I will list my findings. Then, the experts can add their findings afterwards.

The amplifier board is stated to accept up to 50V supply. Chris noticed that an LM2575 was used instead of an LM2575HV. The LM2575 can only handle 40V while the LM2575HV 60V. This means the board in reality can handle up to 40V supply and not 50V unless they have changed the LM2575 on later production lots.

The LM2575 steps the voltage down to some 15V and a linear post regulator takes it further down to 12V. The 12V is used for a part of the TPA3255 chip and for the input balanced signal pre-amplifier. As the input balanced signal pre-amplifier is based on TL072, 12V does not allow the TL072 to achieve the best THD values. The coupling capacitors to the TPA3255 are ordinary electrolytic types. The input circuit I do not see as up to level of the TPA3255 chip being about the top of monolithic class D amplifiers.

The two decoupling capacitors are marked "ELNA, 3900uF/100V". Why use a 100V rated capacitor in an amplifier stated to accept max. 50V? 3900uF (times two) is not excessive for an amplifier board rated for more hundred watts. Chris mentioned they are probably "fake" and not really low ESR. I replaced them and measured their value - it is indeed 3900uF.

I changed the whole input circuit and power regulator system. I use LM4562 supplied with 24V and linear regulators. Sounds really fine. How much better these modifications sound compared to initial I cannot say.

And now for the experts.............

In short, my conclusion is a TPA3255 amplifier board at a very attractive price but not without certain less good design details.
The TI evaluation board I do not have or know.
 
C

chermann

Member
Joined 2017
Hi FF
Hi kay


jes that´s true what FF wrote.
in the thread "what is wrong with TPA3255" i had with the help of other members a lot of modification except the input circuit as FF.


the Ti board and the 50% discount is close since some month and the 3e board group buy is close too. this chinese board was a very nice/cheap starting point to get a good powerful (not as Ti EVM boards) amp to play and learn.
all polymere caps and other caps are fake and the opamps are not always real T072---change it to opa1602, the coil after the regulator is wrong (221--->220) and the reset/mute function is missiong. all LED indication are other way round , the output filter coils are about 13µH and the grey caps i changed...


so if you want to play around...;)


chris
 
C

chermann

Member
Joined 2017
A

a3cervo

Member
Joined 2012
Hi X,

I have the TI 3255EVM board and my DAC also generates an audible "clicking" sound when changing bitrate but it doesn't trigger the amp to go into fault mode. Also, the amp is not picking up the "clicking" sound from the DAC as I don't hear it from the speakers at all. Just wanted to share this info with you.



xrk971 said:
I have a question about the TPA3255’s auto protection and going to fault mode:
I’ve been testing a generic (“Nobsound”) TI reference design of the TPA3255 and noticed that it is very sensitive to input pops and clicks from source material. It will go into fault mode and require power cycle to reset. That’s a problem for me - is it simply the reference design or inherent in the TPA3255 chip itself? I realize it is there to protect speakers - but there has to be a better way to reset (maybe automatically?). The pops were small and came from my DAC switching tracks or changing but rate etc. not ideal but no other amp I have locks me out.

Testing it here with a 12v (13.8v nominal) lead acid gel cell powering a 600w DC-DC step up, followed by a high current CLC with flat wire CoilCraft.
Click to expand...
 
Last edited:
DarkOne

DarkOne

Member
Joined 2001
Kay Pirinha said:
Does anyone have experience with this ebay offer and seller, and this Amazon offer, resp.? They appear to be more economic than TI's evaluation module even at discount, but may not be as versatile, which wouldn't mind.


Best regards!
Click to expand...
I have this one for 2 weeks and its not worth 50$. Its too noisy just like all other TPA325x amps I tried. Swapped 3 power supplies, noise is the same.
I also replaced opamps and capacitors, but I haven't tried it yet.
 
You must log in or register to reply here.
Top Bottom