TPA3250 somebody is listening?

,,,I was able to understand the cause of the discomfort I felt by looking at the measurement report of Amirm.
Especially I think that FX-502SPRO is completely miss design. When the volume is raised the protection circuit works and the error lamp lights up, it is an incredible operation anymore!



i cannot agree......


if some of these guys read our thread here....they will try something i guess...;):D


chris
 
,,,I was able to understand the cause of the discomfort I felt by looking at the measurement report of Amirm.
Especially I think that FX-502SPRO is completely miss design. When the volume is raised the protection circuit works and the error lamp lights up, it is an incredible operation anymore!



i cannot agree......


if some of these guys read our thread here....they will try something i guess...;):D


chris

@chermann + 10
 
'm satisfied with the unit for music, not so much for TV where I get occasional harshness /sibilance on some voices, I don't know why.

Edit: I need to amend the above. I had an epiphany realizing the TV Box volume is digital and should perhaps be treated as the volume control on a computer music player - i.e. resolution should be better as max volume. And lo and behold, 90% of the harshness/sibiliance problem went away, and it doesn't bother me anymore. The remaining 10 % means, IMHO, that the FX502S PRO leans slightly on the bright side. Though I've read here and there that it is a balanced amp, my ears don't agree. I'm no expert but can't help relate this to the amp's very poor test results on ASR.

It's still useable, and preferable for me to both a class AB amp and an AVR I had many times its cost and size. But given the 3520 is a reference chip, it is disappointing as it could have been a home run.

I think I have a good idea now of the amp's sound will try the op amp "upgrade" and will be able to hear if there is a difference or not. I remain skeptical: why would FxAudio skimp on a few cents and use crappy or even fake, as it was claimed, op amps? What if it is an engineering, and not a parts problem? The worst that can happen is I'll lose $20.
 
Edit: I need to amend the above. I had an epiphany realizing the TV Box volume is digital and should perhaps be treated as the volume control on a computer music player - i.e. resolution should be better as max volume. And lo and behold, 90% of the harshness/sibiliance problem went away, and it doesn't bother me anymore. The remaining 10 % means, IMHO, that the FX502S PRO leans slightly on the bright side. Though I've read here and there that it is a balanced amp, my ears don't agree. I'm no expert but can't help relate this to the amp's very poor test results on ASR.

It's still useable, and preferable for me to both a class AB amp and an AVR I had many times its cost and size. But given the 3520 is a reference chip, it is disappointing as it could have been a home run.

I think I have a good idea now of the amp's sound will try the op amp "upgrade" and will be able to hear if there is a difference or not. I remain skeptical: why would FxAudio skimp on a few cents and use crappy or even fake, as it was claimed, op amps? What if it is an engineering, and not a parts problem? The worst that can happen is I'll lose $20.


please try the opamp upgrade with a LME49720NA it cost about 6 euro with shipping = 2x 6 = 12 euros
a red WIMA cap under the board in parallel to the decoupling caps would be help too
 
please try the opamp upgrade with a LME49720NA it cost about 6 euro with shipping = 2x 6 = 12 euros
a red WIMA cap under the board in parallel to the decoupling caps would be help too

In Switzerland the cheapest provider, Conrad, charges an extra 8 CHF "small quantity upcharge" so the cost will be ~$20. Changing caps is beyond my ability. If that is necessary, then the op amp upgrade is perhaps not worth it.
 
changing caps is not need now.just add a caps in parallel.


pic 2..the smal red capshttps://www.diyaudio.com/forums/class-d/315681-tpa3250-listening-53.html#post5423051

Hadn't seen that pic, thx a lot for your trouble.
The soldering is straightforward right? Can you use a regular iron or do you need a small, electronics one?

As it's the weekend and I had a little time on my hands, I retried my class AB amp... and contrary to my initial impressions, liked the sound better.
In particular, all harshness was gone. Will do some more comparing later.
The reasons for the above may have been :
-inflated expectations (you warned me against this IIRC)
- new toy syndrome
-lack of proper level matching

Never mind, it's been a fun and instructive experiment. My thanks to you and other members in this thread.
 
Hi,
My first post in this thread.

I bought the FX502spro and trying to find out how much sound I can squeeze out of it.
Changed all electrolytics for better Q and 50V. Opamps 4562. filter the noise on the LM317. exchanged the 33uH for 150mH after the buck regulator. Exchanged the air coils for shielded types (lot of work as in these where glued...) and took filter caps out and smaller value polyprop. 470 uF on AVDD.
Bought 32V adapter, so its running on 32 V now. (much better than 24V)

The sound has improved dramatically. And it is really interesting to see what can be squeezed out. Generally I like the sound of this class-D, it is still not perfect. Sound stage is wide and high, but lacking real depth.

After exchanging the 33uH after the Buckregulator the hum with volume turned down has gone. With my ear at the speaker I hear still some noise.

But the strange thing is, when I turn up the volume, at a certain point the hum is back. But turning it further (last quarter or so) the hum disappears again. Ofcourse the hum is in the range I would listen to:(.
Someone any idea how this can happen? And how to resolve?

Many thanks.
 
Hi,
My first post in this thread.

I bought the FX502spro and trying to find out how much sound I can squeeze out of it.
Changed all electrolytics for better Q and 50V. Opamps 4562. filter the noise on the LM317. exchanged the 33uH for 150mH after the buck regulator. Exchanged the air coils for shielded types (lot of work as in these where glued...) and took filter caps out and smaller value polyprop. 470 uF on AVDD.
Bought 32V adapter, so its running on 32 V now. (much better than 24V)

The sound has improved dramatically. And it is really interesting to see what can be squeezed out. Generally I like the sound of this class-D, it is still not perfect. Sound stage is wide and high, but lacking real depth.

After exchanging the 33uH after the Buckregulator the hum with volume turned down has gone. With my ear at the speaker I hear still some noise.

But the strange thing is, when I turn up the volume, at a certain point the hum is back. But turning it further (last quarter or so) the hum disappears again. Ofcourse the hum is in the range I would listen to:(.
Someone any idea how this can happen? And how to resolve?

Many thanks.

Hi

oh....nice that you did a lot of changes...do you have some pics for us.

@hum
hum like 50Hz-100Hz...or e.g. pink noise over the speakers?


lg
 
Hi,
My first post in this thread.

I bought the FX502spro and trying to find out how much sound I can squeeze out of it.
Changed all electrolytics for better Q and 50V. Opamps 4562. filter the noise on the LM317. ........

Many thanks.

Do you have no thermal problems with the LM317. originally it is set for 24V output...but if you push 32V you should have at least 1 Watt Dissipation.
e.g.

8V different , 160mA drawn by all elements (post #222)= 1.28Watt
 
No, not noise. But indeed like 50..100Hz. Only higher, I think.
 

Attachments

  • WP_20181203_15_29_10_Pro.jpg
    WP_20181203_15_29_10_Pro.jpg
    986.2 KB · Views: 489
10% safety factor for capacitor, Oh~Ye - thks fr confirm, now i have peace of mind supply FX502spro with meanwell configured 31V.

You have a point on heat problem, my thoughts of using 31V beyond 24V:
~ TPA3250 Power output depends on voltage -- was hoping 31V provides more for any transient music peak.
~ 24V from LM317~1.5A if undergood dissipation
Rough estimate: This 24V supplies NE5332 Op-Amp (datasheet 16mA) , to LM2596HVS gives 12V for TPA3250 Logic/Gate control (datasheet 40mA), to AMS1117 to provide 3.3V for 8bit MCU ST S003F3P6 (datasheet 100mA) ~ total max approx 0.16A
LDO LM317 ~ 7V drop gives 1.2W heat.

What else did i missed?

2 items unsure:
~ International Rectifer IRF9530N -- what function on FX amp?
~ Does FX TPA3250 back-stage driven by whatever voltage from DC Supply voltage (OR being internally regulated by LM317 - i highly doubt, but not sure what jigglebones is implying)?

from post#222

I checked again, but the lm2596 is fed directly from powersupply, not from LM317. So only the 2 opamps are behind the LM317.

Rectifiers are there to prevent wrong polarization of power.
 
Oops, forgot to address this question -

Yes, in PBTL mode, as I listed above ...
- but you would need to study the datasheet carefully to work out what wiring changes are necessary, and I suspect it may be tricky to execute these changes on the FX502SPRO circuit board,
- and here's the kicker - PBTL mode is only suitable for speakers in 2 - 4 ohm range.

Again, if you want more power - go for the TPA3251 or 3255.

I think for PBTL you need 4-8 ohm speakers, because 8 ohm speaker become 4 ohm at PBTL