Shui Yuan Audio TPA3255 Modification Thread

Jump to Latest
Dxvideo

Dxvideo

Member
Joined 2003
ozgurkolay.com.tr
jsiddall said:
I have one of the original Shui Yuan modules and when tested at higher frequencies (ex: 1 kHz) everything works pretty much as per the spec sheet. I can see about 225 W @ 4 ohm or 125 W @ 8 ohm. But at low frequencies of, for example, 100 Hz., the module goes into a latched fault (requires a manual reset) with 4 ohm loads at as little as 165 W output. At power levels just below the fault threshold there is some sporadic oscillation near peak power:

View attachment 1105364

I checked the supply voltage, and while it drops a few volts, perhaps from 48 V to 46 V as load increases, it doesn't seem to be doing anything crazy.

I confirmed OC_ADJ is 22k ohm, so per the datasheet, a latched fault with no CLIP_OTW indicates either UVP or OLP (CB3C>1.7ms) since the temperature is also low. The PVDD is well above the UVP threshold so unless something wonky is happening with GVDD it doesn't seem like an UVP issue. That leaves OCP, but I'm not sure how the FET current could be over 17A for an extended period given the peak output current at 43 V and 4 ohms is just over 10 A.

Since this is going to drive a 4 ohm subwoofer I really don't want to have to hard reset the amp whenever things get a bit loud. Has anyone else seen something like this or have any ideas where to start looking for the source of the problem?
Click to expand...
Did you use resistive dummy load or real speaker driver for this measurement?
 
R

Roland968

Member
Joined 2020
jsiddall said:
I have one of the original Shui Yuan modules and when tested at higher frequencies (ex: 1 kHz) everything works pretty much as per the spec sheet. I can see about 225 W @ 4 ohm or 125 W @ 8 ohm. But at low frequencies of, for example, 100 Hz., the module goes into a latched fault (requires a manual reset) with 4 ohm loads at as little as 165 W output. At power levels just below the fault threshold there is some sporadic oscillation near peak power:

View attachment 1105364

I checked the supply voltage, and while it drops a few volts, perhaps from 48 V to 46 V as load increases, it doesn't seem to be doing anything crazy.

I confirmed OC_ADJ is 22k ohm, so per the datasheet, a latched fault with no CLIP_OTW indicates either UVP or OLP (CB3C>1.7ms) since the temperature is also low. The PVDD is well above the UVP threshold so unless something wonky is happening with GVDD it doesn't seem like an UVP issue. That leaves OCP, but I'm not sure how the FET current could be over 17A for an extended period given the peak output current at 43 V and 4 ohms is just over 10 A.

Since this is going to drive a 4 ohm subwoofer I really don't want to have to hard reset the amp whenever things get a bit loud. Has anyone else seen something like this or have any ideas where to start looking for the source of the problem?
Click to expand...
I could imagine that this is due to the small double coils connected in parallel with 2 x 22uH (=11uH), which are much too small for the output power. The current of these coils is well below 10A. They probably going into saturation and putting the amp in an undefined state.
It's just a cost saving measure, they sell the complete boards for the price of 2-3 good 10uH coils, the built-in coils cost under $1 in China, in quantities even less than 50 cents.
 
  • Like
Reactions: 1 user
J

jsiddall

Member
Joined 2018
Roland968 said:
I could imagine that this is due to the small double coils connected in parallel with 2 x 22uH (=11uH), which are much too small for the output power. The current of these coils is well below 10A. They probably going into saturation and putting the amp in an undefined state.
It's just a cost saving measure, they sell the complete boards for the price of 2-3 good 10uH coils, the built-in coils cost under $1 in China, in quantities even less than 50 cents.
Click to expand...

Hmmm... not an easy fix. But if that were the cause these boards should all behave the same. And maybe they do. Has anyone else tested another one of these at low frequencies, low impedance and high power?

Dxvideo said:
It looks; the o/c protection of one side H bridges triggers (not latched). That's weird for this level :unsure: May be chip itself is faulty..
Click to expand...

Yeah, it's unclear what is going on there, but it is clearly something high frequency (the big blob of trace), which maybe could be causing something to trigger a fault condition. It is possible a faulty chip could do that too, but strange that it is not faulty at higher (but still pretty low) frequencies like 1 kHz.

If I can't find a better solution I might need to put a limiter on the LF crossover to make sure it never exceeds 165 W.
 
G

Gaspar74

Member
Joined 2021
I just got this version of the board it looks identical however doesnt say Shui-yuan audio on it. Says nothing. Came with one of the inductors chipped in shipping. NO PACKAGING aside from PCB in ESD bag and then in plastic bag. My question is, are the inductors in the Aiyima A07 PCB better than these? These are 10uh rated and so are the Aiyima however Aiyima are different. This is the Aiyima A07 stock ones. I have a scrap A07 board I can use as a donor of those are better quality.
IMG_7597.JPG
IMG_7574 2.JPG
 
Last edited:
D

dene001

Member
Joined 2010
I have a question regarding the input impedance of the board, which as I have read in one of the posts, is 10k. As I have a tube preamp with a rather high output impedance (>20k) I would like to increase the input impedance of the board. Would changing the input resistance from 10k to 47k have a negative influence? By changing, I mean change both 10k resistances of the 1st opamp in both the L and R channels. Any experience with such a modification?

Gaspar74 said:
I just got this version of the board it looks identical however doesnt say Shui-yuan audio on it. Says nothing. Came with one of the inductors chipped in shipping. NO PACKAGING aside from PCB in ESD bag and then in plastic bag. My question is, are the inductors in the Aiyima A07 PCB better than these? These are 10uh rated and so are the Aiyima however Aiyima are different. This is the Aiyima A07 stock ones. I have a scrap A07 board I can use as a donor of those are better quality. View attachment 1138514 View attachment 1138519
Click to expand...
Photos do now show the chipped inductor, but if it's a small chip and not a crack I tend to say it wouldn't have a hearable influence.
 
Dxvideo

Dxvideo

Member
Joined 2003
ozgurkolay.com.tr

alephotos said:
Hi, I'm trying an input selector with relay on the tpa3255, when I use the input where the cd player is connected it works perfectly, but if I move the selector to an input where there is nothing, it starts buzzing with a very annoying noise . How can I remove the noise? Thank you
Click to expand...
Dear Alephotos, this is normal because the input buffers are inverted design. So you may put about 10k resistor to empty input to gnd line for each channel.. If still exists then try a lower value.
 
  • Like
Reactions: 1 users
You must log in or register to reply here.
Top Bottom