TPA3255 - all about DIY, Discussion, Design etc

Jump to Latest
A

audiogue

Member
Joined 2010
feket663 said:
The output filter Approximate match the design. The inductors are 2 x 22 uH, paralelled, = 11 uH as opposed to the original 10 uH recommendation. Unfortunately, better quality, 10 uH Coilcraft or Würth inductor doesn't fit.
Click to expand...
Hi feket663, the inductors are not so bad as it seems. I have this module and tested the cores for a Aiyima A07. Attached the distortion of the Aiyima at 4W with 35V supply and 8.2 Ohm load. The dc resistance of the dual core paralleled to 11uH is only 11 mOhm and they seem to be flat copper wire type.
If you want to upgrade the sound:
1. put a Wima MKP4 33nF 250V parallel to the 10uF Electrolytic coupling capacitors,
2. exchange the ceramic 1uF (used on this module for the LC filter) to a MKS or better MKP type of film capacitor. See Attachments.
 

Attachments

  • Core A07.JPG
    Core A07.JPG
    66.4 KB · Views: 171
  • core TPA dual 22uH paralled.JPG
    core TPA dual 22uH paralled.JPG
    67.6 KB · Views: 158
  • Würth.JPG
    Würth.JPG
    67.8 KB · Views: 154
  • 1uF MKS.JPG
    1uF MKS.JPG
    69.3 KB · Views: 154
  • Original 1uF Ceramic.JPG
    Original 1uF Ceramic.JPG
    64.3 KB · Views: 165
Last edited:
  • Like
Reactions: 1 users
shadders

shadders

Member
Joined 2010
audiogue said:
The inductor is the core assembled on the mentioned board on which I reply. Look at the post 4784, this is the board.
Click to expand...
Hi,
Thanks. I checked the Wurth coil referenced in your measurements, and compared to the Coilcraft used by Texas Instrument in their PFFB document. Your measurements show THD much lower for the Wurth output inductor than the TI implementation of the Coilcraft.

After checking Audioscience review the THD is as expected, so perhaps your measurements are showing the THD lower than it should be ?

Regards,
Shadders.
 
A

audiogue

Member
Joined 2010
Hi shadders,
I have PFFb implemented. This modification reduce the overall distortion. The output resistance changed from 130mOhm to 25mOhm. This modification gives also more control over the woofer and hence a lot more clarity to the sound.
My PFFb implementation is:
- Input resistor between Opamp and coupling capacitor: 3kOhm
- Feedback only 22kOhm parallel with 100pF (no t-network)
- Output Zobel network with 10 Ohm and 1uF MKS (TI solution: 1 Ohm and 220nF)
 
Last edited:
  • Like
Reactions: 2 users
A

audiogue

Member
Joined 2010
Hi shadders,
all the distortion info in the TI documentation is related to the highest supply voltage (51V).
From my measurements Distortion vs. Voltage is best at about 35V.
And the heat dissipation is low at this low voltage, and about 80W vs 8 Ohm is enough output power.
 
Last edited:
  • Thank You
  • Like
Reactions: 1 users
T

trungdtmc

Member
Joined 2022
audiogue said:
Hi feket663, the inductors are not so bad as it seems. I have this module and tested the cores for a Aiyima A07. Attached the distortion of the Aiyima at 4W with 35V supply and 8.2 Ohm load. The dc resistance of the dual core paralleled to 11uH is only 11 mOhm and they seem to be flat copper wire type.
If you want to upgrade the sound:
1. put a Wima MKP4 33nF 250V parallel to the 10uF Electrolytic coupling capacitors,
2. exchange the ceramic 1uF (used on this module for the LC filter) to a MKS or better MKP type of film capacitor. See Attachments.
Click to expand...
but there inductors bulid quality not so good
 
A

audiogue

Member
Joined 2010
Hi trunkdtmc,
I forgot to mention that the inductor saturation current is at about 10A in the 11uH configuration (Würth indictor 22A). This is also fully akzeptable.
My post was addressed to people which have this board and plan to exchange the cores. In my opinion this is not necessary due to the electrical good data and measurements. The sound was improving with these cores in Aiyima A07 and has not notable changed (or further improved) with the Würth cores).
 
  • Like
Reactions: 1 user
T

trungdtmc

Member
Joined 2022
audiogue said:
Hi trunkdtmc,
I forgot to mention that the inductor saturation current is at about 10A in the 11uH configuration (Würth indictor 22A). This is also fully akzeptable.
My post was addressed to people which have this board and plan to exchange the cores. In my opinion this is not necessary due to the electrical good data and measurements. The sound was improving with these cores in Aiyima A07 and has not notable changed (or further improved) with the Würth cores).
Click to expand...
yes, i am also using these boards. Overall sound quality is good, except for issues with heatsinks and slightly poor finish. The 12vdc source circuit for the opamp is quite clean, even better than the original a07
 
Last edited:
P

Puffin

Member
Joined 2006
Mr Baldingfast wrote :
According to this post, minimum distortion occurs with a supply voltage of 35V (whether this is audible is another matter):

https://www.diyaudio.com/community/...diy-discussion-design-etc.287470/post-665202

I made up a linear PS today by nicking bits from another dormant amp. The supply gives 42vdc. I have only run it for an hour or so and can't really say if it makes any difference yet. It may be that the presentation is a bit more beefy but the overall presentation which I like is still there. Turning it up is a hoot as there is obviously plenty of power and within sensible limits it all holds together.
 
You must log in or register to reply here.
Top Bottom