Alto TS subwoofer amplifier

[JP-Stereo]

Is someone has ever been able to repair the SEIKAKU amplifier bloc used in the TS12su sub and also in some other models.


It's labeled DM-350w+65w, a Class D amplifier module with components for the 65w A/B amplifier removed on the HB03189 board version for this subwoofer.
I already change 2 ic on the separate preamp board MS15SA-PRE, now the preamp is fully operationnal but the amp module behave.

On the output, there is a hissing noise and loud cracking at power off.
I have sound on some instances.
I've changed the 4 biggest capacitors because they were looking "imflated".

But it's the first Class D i look at and i don't really know where to start.
It also seem that some crazy voltages are there.

Thanks for any hint.

I am including a kind of generic schematic, Seikaku is manufacturing customised version with small differences for different brand.
They have different HB0xxxx version number and more or less components or empty space on the board.

[JP-Stereo]

I was able to get the preamp/crossover/limiter circuit to work fine.
An OpAmp and a few caps replaced.

When plugging the amp board to the preamp, i have a hissing noise when on along with a big thumb at power off.
- Replaced the 4 big PS filters caps.
- Power supplies check OK.
- The two IRFB4020 were not measuring the same, so i changed both.
- Two more ceramic caps.

It seem to work well but still have some background noises and it look prone to oscillation.
Also have a strong click sound at poweroff.

Is there a bias as in older class of amp ?

Thanks.

[x-pro]

...Also have a strong click sound at poweroff.

Is there a bias as in older class of amp ?

C11 is responsible for the click (schematic Sheet 3).
There is usually no offset adjustment in class D amplifiers. Only the oscillation frequency and the dead time between pulses can be adjusted.

[JP-Stereo]

Hi, happy see you here...

I remove it and it test 36 uf offboard, the schematic show a 10 uf but "it identify itself" as a 33 10 b.
It may be a 33 uf testing 36 uf..?
it's a polarised surface mount capacitor.
I want to try a 33uf 16v electrolytic cap to see.

[g1]

I remove it and it test 36 uf offboard, the schematic show a 10 uf but "it identify itself" as a 33 10 b.
It may be a 33 uf testing 36 uf..?

What kind of cap tester? Some inexpensive ones (like found in DMM) will give a high reading if the cap is leaky. But I don't think that 36 is high if it is actually a 33uF cap.

[x-pro]

replace with 10uF

[JP-Stereo]

Hi this is the surface mount cap.
On these caps the marking indicate the (+) side, since the pad on the marked side conduct to U2 pin 5.



And this is my DMM, i think it's a middle class one.
Agilent U1552b, it give pretty consisttent results exept in very high values.


I will try a standard 10uf electrolytic cap.
I figure the 10 stand for 10 volts, so a 16v should be fine.

[x-pro]

Your measuring device is very good! You can put a 33 µf capacitor as well. 36 is something wrong.

Since the whole amplifier is digital, it is advisable to check the capacitors in the power supply - C140, C141, C142, C143, C150, C151. Especially since the larger ratings were under suspicion.
Carefully check with an oscilloscope the noise level on pins 3 U8, U10. Also check the noise on (+)C150 and (-)C151 relative to the common. Check the noise signal with the probe input closed.
For the IRS2092, the absence of any of the +/-5 V is unacceptable!

[JP-Stereo]

I already put a 10uf in place of C11, but is a 30uf preferable ?

C124, C125, C133, C135 were the biggest cap, they are already replaced by new ones.
I just replace C140, C141, C142, C143, some were a little off and i had them on hand.

For C150, C151, since they are just filters for the linear regulators, can i substitute two 33uf 50v i have on hand too ?
Instead of the original 22uf...

"Check the noise signal with the probe input closed.", please explain ?

For the +/-5v if it's they are present at U2 pin 1 & 6 is it enough ?

[x-pro]

stitute two 33uf 50v i have on hand too ?
Instead of the original 22uf...

"Check the noise signal with the probe input closed.", please explain ?

For the +/-5v if it's they are present at U2 pin 1 & 6 is it enough ?

If 22 uf is bad, you can put 33 uf and 47 uf.
The +/-5 V voltage on the U2 pins is good, but it can be with interference and noise.
This can be checked with an oscilloscope. Input Y in AC mode ("closed input")

[JP-Stereo]

Hi and thank you again.

In order to take the measurments you suggest, i need to run the board off the aluminum case/backplane.
It imply that the board doesn't rely on the chassis for any ground network...
Do you think temporary heat sinks may be required ?

OK by the AC mode you must mean something like "AC Coupling" for vertical deflection ?
In order to zoom on the noise instead of being distracted/blinded by any DC offset, do i get it ?

[x-pro]

In order to take the measurments you suggest, i need to run the board off the aluminum case/backplane.
It imply that the board doesn't rely on the chassis for any ground network...
Do you think temporary heat sinks may be required ?

OK by the AC mode you must mean something like "AC Coupling" for vertical deflection ?
In order to zoom on the noise instead of being distracted/blinded by any DC offset, do i get it ?

1. This is not necessary. Perhaps there are other elements with this voltage where you can install a dipstick.
2. Yes. ("OK by the AC mode you must mean something like "AC Coupling" for vertical deflection ?")

[JP-Stereo]

OK, the specified caps are replaced.
About the U8, there is an error on the schematic but physically it was installed correctly (in reverse).
As per the 78L05 specifications the output is pin 1:



The scoped voltages are +5.02 and - 5.02 volts DC at ouputs, very little ripple at this scale.
In AC i see about 175mV of noise on the negative rail, the positive is flat as seen at 200mv/line.
At power down, the positive is swinging a few time before getting down.
The negative is linear to 0 vdc
They are not behaving in syncro.
I am measuring about 2vdc on the speaker output with a DMM.

[JP-Stereo]

Here are some measurment the DC on the left and the AC coupled on the right:

Input of U6 & U9



Output of U6 & U9

The top trace noise is constant and the bottom one is intermittent.

Input of U8 & U10


Output of U8 & U10


I measure -4.9v at R4 and +4.7v at R8 with around 0.003v of ac noise with the DMM.
C148 & C153 are testing good off circuit.

[JP-Stereo]

Is the above too much noise for such supply lines or it is normal ?
Ounce stabilised, the voltages are not looking so bad.

Here is the VCC-VEE bus:



But at power up and at power down crazy things happen.
This is the VCC-VEE bus at power off: