Alto TS subwoofer amplifier

[JP-Stereo]

When measuring at D100/D101 anodes:
The first time the circuit boot the raise is lnear fron 0 to 4.47 vdc in about 3 seconds.
At turn off, it swing to negative 2 vdc for a second then rest at 0 vdc.

The folowing time it's turned on in a short period, voltage raise and drop to zero five or seven time before finnally staying at 4.47 vcd.

There is also a little noise visible at this DC scale ort around 400mv in AC.

[JP-Stereo]

The problem occur at power on/off and randomly too.

The behavior of the SD line let me think the problem is there...

Is it possible to isolate or disconnect the DC Offset, the OTP or the LM393 to see if the problem go ?

[x-pro]

.... and randomly too.

To check the stability of this point during operation, you can try: temporarily unsolder R118, check C101, C104, C144. Even if it is SMD ceramics.
(and please remind me of the nature of the defect.)

[JP-Stereo]

We have load pops and clicks at power on/off and some random background noise while on.
Scoping the SD line show that the muting signal is not stable.

The system don't let me upload a .mov, so i rename it with a .zip extension.
Just download the file and rename it with a .mov extension.
The yellow line is the SD signal and the blue is the speaker out.
Look at the behavior of SD at power on and off.
It doesn't look optimal for me...


The SD line is definitly suspect, the first condition is not filled, the signal is badly swinging :



image widget

[JP-Stereo]

C104 test OK at 1000pf.
C144 and U12 are not on the board at all.
R118 removed the scope stay the same, a stairway up and down before stabilising...

[JP-Stereo]

According to the U2 (IRS2092) manufacturer (International Rectifier) application notes, the CSD (SD) line require only a capacitor for basic operation.
The DC Offset, the OVP, the OTP and U14 are only optionnal protection circuits.




U12 and D120 are not implemented in this subwoofer only version of the amp.

The U14 will sink current on the SD line until E rise over +5vdc, i have to verify it's reference.

The DC Offset is not really a DC offset control, it's more of a DC detector.
It's monitoring the Speaker output and if +VO2 get higher than +0.6vdc at the + input of U3, it drain the SD line.
So this is also a suspect.

The Over Voltage Protection and Over Temperature Protection are two other comparators.
It look like U4A handle the OTP, do not fully understand OVP at U4B, but it is monitoring -VCC.

I think i is safe to simply remove R109 or D3, in order to test U2 on it's own and verify the operation of the previous comparators.

I think it is safe but, any other consideration ?

[JP-Stereo]

Things have evolved...

I decided to remove D100, D101 & D136 and eventually R403.
Starting from U4 to find D101, i meet this:


U4 has sparked between pin 1 & pin 2.
Once this IC is removed, no more stairway in the SD line at power on or off.
(A new IC is ordered)
But the SD line is resetting at an interval of about 3 seconds, cycling back on and off.

D100, D101, D136 and R403 were all removed and reinstalled without any change.
So it's coming from U4 itself or else but not from the SD line.

This IRS2092 have it's own Over Current Protection and it's looking like the LO or HO is overloaded and it's resetting in cycle.
I am reading around +/- 70 vdc on the +VCC -VCC, this is also present on Vss input, it should be +/- 60vdc.

U4 has sparked between pin 1 & pin 2.
Once this IC is removed, no more stairway in the SD line at power on or off.
But the SD line is resetting at an interval of about 3 seconds, cycling back on and off.

D100, D101, D136 and R403 were all removed and reinstalled without any change.
So it's coming from U4 itself.
This IRS2092 have it's own Over Current Protection and it's looking like the LO or HO is overloaded and it's resetting in cycle.

IRS2092.pdf

AppNotes_irs2092s.pdf

How to know what is loading U4 or if it's an over-voltage ?

[JP-Stereo]

All the pictures i've posted on the last days are not visible anymore ?

Problem: Loud Pop and cracking sound at power on and off, white noise and hiss once open.
The SD line that is supposed to mute the amp at power on or off was erratic and pulsing.

Here is a recap, investigating the SD line, i've found a sparked U4 (pin 1 to 2).



I upload a MOV with a fake .ZIP extention, just replace the .ZIP with .MOV and it will play.
The yellow line is the SD line between R403 and D3, the blue line is the U2 VSS at pin 6.

Removing the damaged U4 solve the erratic power on /off, but now the SD line is pulsing on and off by itself.
To verify, i have remove, D100, D101, D136 and even R403, that isolate completly the IC CSD line from the SD bus.
The application notes talk about an over current/voltage protection internal to the IRS2092S (U2) that is monitoring both Mosfets.
And it's CSD is cycling as long as a fault is detected on the Mosfet, over current or over voltage. (App Notes p10 to 12)

These Mosfets are new, so it may be some support components around too.
I will double check everything around these Mosfets...

[JP-Stereo]

I just found that DS2 is pulsing a flash at the same rate as the CSD is resetting.
VAA pin 1 = +4.82vdc and VSS pin 6 = -5.07vdc very low ripple.

Here is the .MOV of the SD buss.

The +VCC is at +64vdc and the -VCC is at -80vdc, this is not symetric.
The positive stay high for very long, the negative is getting to zero v fast after power off.

Not shure how to test U7, i measure -6.5vdc between pin 2 and 3 and it's not stable.

[x-pro]

If the power supply has been causing the problem, be sure to check all the electrolytes in the power supply.
Next, once again check all the supply voltages in the power supply, namely +/-15, +/-5, and +/-1.8, +/-0.6. They should be stable. Especially the small values. The values of Vcc, both (+) and (-) must be equal and appear simultaneously on input U4. This chip disables the amplifier when the supply voltage is unbalanced +/-60(70)V.
The +12V supply from U7 is also important. Note that this 12V is relative to the minus of Vcc.
With D135 present at the same time as D136, U14 has no effect on SD, only on the "Protect" line. Where it goes is not clear.

[x-pro]

Not shure how to test U7, i measure -6.5vdc between pin 2 and 3 and it's not stable.

Solder it out and test it separately from a +(15-20)V voltage source by loading it with a current of about 100 mA.
Check or replace C147.

[x-pro]

The +VCC is at +64vdc and the -VCC is at -80vdc, this is not symetric.

That's not good. You need to look for a skew problem. Be careful with the IRS2092, it is capricious.

[JP-Stereo]

Hi, thanks everyone for help on this project.

I've learn a lot on the Class D and i feel i was closed to the problem...
But the time has run out for this one and the project is cancelled.

From this point i think Class D have some advantages for performers and public events but are not durable in a daily daylong use.
When a problem occur it's often catastrophic...