Aiyima A07 Max Amplifier Review

[trungdtmc]

I also noticed this early. I think AIYIMA thought that if you changed the main decupring capacitor to 63V, you would not notice this small electrolytic capacitor. Maybe the designer knew this matter. Aiyima's response is completely wrong.

I think some 1206 63-100V x7r smd caps are better

 

[MCH]

I also noticed this early. I think AIYIMA thought that if you changed the main decupring capacitor to 63V, you would not notice this small electrolytic capacitor. Maybe the designer knew this matter. Aiyima's response is completely wrong.

Has been discussed with them:

https://audiosciencereview.com/foru...a-a07-max-amplifier-review.49920/post-1792613

I am considering changing mine, but 0.1uf >50V in that smd package are very difficult to find. Maybe that's one of the reasons why.

 

[trungdtmc]

i replaced with x7r 100v

 

[AIYIMA:GUO]

我正在寻找一款价格便宜、可以在 2 欧姆负载下运行的 2 通道（或双单声道）高 SQ 放大器。在单声道中运行 2 A07 Max 似乎符合要求。
不幸的是，首选的 Fosi ZA3 额定电阻并未降至 2 欧姆。

还有什么我应该考虑的吗？

推荐使用DC48V电源

 

[Abstracted]

推荐使用DC48V电源

In response to me asking about the A07 Max in mono they said:

In English: "It is recommended to use the DC 48V power supply".

 

[MCH]

i replaced with x7r 100vView attachment 352566

Have you taken into account the capacitance drop at 48V?

 

[Toku]

i replaced with x7r 100vView attachment 352566

Due to the X7R characteristics of multilayer ceramic capacitors, the capacitance decreases significantly when DC bias is applied. Therefore, it is not suitable for use in power supply circuits.

 

[trungdtmc]

Due to the X7R characteristics of multilayer ceramic capacitors, the capacitance decreases significantly when DC bias is applied. Therefore, it is not suitable for use in power supply circuits.

So TI wrong?

 

[Toku]

So TI wrong?View attachment 352589

MLCC types have excellent characteristics against temperature changes, but products with high dielectric constants have a significant decrease in capacitance due to DC bias (DC voltage application). Therefore, consideration must be given to reducing capacitance depending on the circuit voltage used.

 

[trungdtmc]

MLCC types have excellent characteristics against temperature changes, but products with high dielectric constants have a significant decrease in capacitance due to DC bias (DC voltage application). Therefore, consideration must be given to reducing capacitance depending on the circuit voltage used.

My quote is from the documentation of the LMR16030, it is the main 12v step-down component used in the A07 Max, the maximum input voltage is 60V. Those capacitors are for decoupling and filtering high frequencies, so the reduction in capacitance is not as important as the effectiveness in reducing pulse noise.
I also noticed a few things that were not good enough in the design with this regulator in the A07 max

 

[Toku]

The attached circuit diagram is a sample circuit for BTL connection described in the TPA3255 manual. The capacitor C33 C36 C39 C20 that was replaced in the modification photo is probably the capacitor surrounded by the red frame in the circuit diagram. The power supply voltage is applied directly to these.

 

[MCH]

. The power supply voltage is applied directly to these.

I can confirm this. And to a fifth one 0.1uF near the DC input.

 

[trungdtmc]

And those are decoupling caps

If you try measuring voltage noise at those 2 capacitor points, you will get results of about 5-600mVpp. After I replaced it with x7r capacitors and modified the input of the LRM16030, the result was only about 2xxmVpp

 

[dr_mick51]

And those are decoupling capsView attachment 352892

If you try measuring voltage noise at those 2 capacitor points, you will get results of about 5-600mVpp. After I replaced it with x7r capacitors and modified the input of the LRM16030, the result was only about 2xxmVpp

Hello Trung, I'm a fan of your mods. I do mods myself as well.
The decoupling capacitors you are referring from the documentation are these:

Those SMD Eletrolytics you replaced do not seem to be part of TI original design. They don't exist on any other TPA3255board I've seen.

They normally lie under the heatsink and right next to the TPA3255/51 chip and serve to decouple/bypass the big bulk power capacitors.:

In the case of A07 Max, and Fosi V3 they are placed on the opposite side to the chip(because of thermal design improvements on those amps) but still very closed to the chip and connected thought vias.


I have not fully analyzed the design of the A07 Max, but it looks like those 4 (2x 1uF and 2x 0.1uF) SMD Electrolytic capacitors are on the output audio path, or used by the relay circuit. So they don't withstand the Power supply voltage, only the output voltage.
If that's the case then replacing them with better capacitors might improve the sound quality of the amp. You replaced them with ceramic capacitors and noticed lower noise. What if replaced with MLCC or film caps? I think you can manage to solder them to those SMD pads.

I'll open up my A07 Max this weekend and try to figure out what those caps are doing and where they are connected (in reference to the TI schematics)

 

[MCH]

.


I have not fully analyzed the design of the A07 Max, but it looks like those 4 (2x 1uF and 2x 0.1uF) SMD Electrolytic capacitors are on the output audio path, or used by the relay circuit. So they don't withstand the Power supply voltage, only the output voltage.
If that's the case then replacing them with better capacitors might improve the sound quality of the amp. You replaced them with ceramic capacitors and noticed lower noise. What if replaced with MLCC or film caps? I think you can manage to solder them to those SMD pads.

I'll open up my A07 Max this weekend and try to figure out what those caps are doing and where they are connected (in reference to the TI schematics)

The 4 smd electrolytic capacitors + the one near the power input are connected to the input power. The ceramic ones circled in your picture are not (or not directly at least. There is no continuity)

 

[trungdtmc]

Hello Trung, I'm a fan of your mods. I do mods myself as well.
The decoupling capacitors you are referring from the documentation are these:
View attachment 352916
Those SMD Eletrolytics you replaced do not seem to be part of TI original design. They don't exist on any other TPA3255board I've seen.

They normally lie under the heatsink and right next to the TPA3255/51 chip and serve to decouple/bypass the big bulk power capacitors.:
View attachment 352917
View attachment 352919
View attachment 352920
View attachment 352921
In the case of A07 Max, and Fosi V3 they are placed on the opposite side to the chip(because of thermal design improvements on those amps) but still very closed to the chip and connected thought vias.


I have not fully analyzed the design of the A07 Max, but it looks like those 4 (2x 1uF and 2x 0.1uF) SMD Electrolytic capacitors are on the output audio path, or used by the relay circuit. So they don't withstand the Power supply voltage, only the output voltage.
If that's the case then replacing them with better capacitors might improve the sound quality of the amp. You replaced them with ceramic capacitors and noticed lower noise. What if replaced with MLCC or film caps? I think you can manage to solder them to those SMD pads.

I'll open up my A07 Max this weekend and try to figure out what those caps are doing and where they are connected (in reference to the TI schematics)

Hi Mick, thanks remember me. The capacitors you circled in red are the capacitors with the numerical value 33n in this diagram, no more caps on botom side
Not like A07, Aiyima set A07 Max operating at 450khz pwm

 

[dr_mick51]

Hi Mick, thanks remember me. The capacitors you circled in red are the capacitors with the numerical value 33n in this diagram, no more caps on botom side
Not like A07, Aiyima set A07 Max operating at 450khz pwm
View attachment 352928

I'm sorry to differ, please reconsider.

On all TPA3255 design there are 4 big ceramic capacitors next to the chip. These are the decoupling capacitors. Please see the schematics and compare against this closeup pic:

The bootstrap capacitors are the ones in yellow, the other 4 big ceramic capacitors are the decoupling capacitors.
I won't comment on the Aiyima A07 Max 4 SMD electrolytics yet, because I have not personally checked them. It is possible they also are part of the coupling caps, or maybe not.

I'll come back after I check them and try to identify where in the schematics they fit. The only thing I know is those caps do not exist on any other TPA3255 amp that I've seen so far. What role do they play? let's find out.

 

[trungdtmc]

I'm sorry to differ, please reconsider.

On all TPA3255 design there are 4 big ceramic capacitors next to the chip. These are the decoupling capacitors. Please see the schematics and compare against this closeup pic:
View attachment 353008

The bootstrap capacitors are the ones in yellow, the other 4 big ceramic capacitors are the decoupling capacitors.
I won't comment on the Aiyima A07 Max 4 SMD electrolytics yet, because I have not personally checked them. It is possible they also are part of the coupling caps, or maybe not.

I'll come back after I check them and try to identify where in the schematics they fit. The only thing I know is those caps do not exist on any other TPA3255 amp that I've seen so far. What role do they play? let's find out.

Okay, you can be skeptical. It will be easy to check on the actual board, but I give you this picture, you can see the decoupling capacitors are connected directly to the power supply, but the boostrap capacitors are not
Anw, I added 2 film capacitors in that position and checked the noise level again, there was no significant improvement, still at 230-240mVpp

 

[dr_mick51]

Okay, you can be skeptical. It will be easy to check on the actual board, but I give you this picture, you can see the decoupling capacitors are connected directly to the power supply, but the boostrap capacitors are not
Anw, I added 2 film capacitors in that position and checked the noise level again, there was no significant improvement, still at 230-240mVpp
View attachment 353098

Hello Trung. I could not wait until the weekend, haha. So, I just opened the amp and measured continuity on all those caps, and...... I can confirm that on the A07 Max those caps in the yellow box are effectively the bootstrap capacitors, and the SMD Electrolytic (normally these are X7R ceramics on other amps) capacitors are the decoupling caps for the 2 bulk PS capacitors.
The TI design says to use 4x 1uF X7R ceramic capacitors, but the used 4x (1uF Electrolytic plus a 0.1uF Electrolytic bypass). Can you provide the specs for those 4 ceramic capacitors you replaced first? I'm interested in reducing noise on the supply voltage; even though I am already using an external SMPS DC filter,

 

[MCH]

Thanks guys for discussing this. I might go as well with X7R caps, I had not realised how well they fit in the electrolytics footprints and I have some laying around.
TI eval board uses 4 1uF. These ones GRM31CR72A105KA01L

But I would love @AIYIMA:GUO or @AIYIMA the reason they have chosen to go with the electrolytics.

And what do you guys plan to do with C6? It is also at PSU voltage