TPA3255: Fault Shutdown

Hi Jim

As the fault pin and clip shown a overload or undervoltage error happen. It seems a fake overcurrent, you can increase the BST capacitor value. We recommend BST cap 0.033uF~0.22uF, the EVM only designed for typical application.

Interesting...I don't see this information in any of the documentation. I do see this value being used with the TPA3116D2.  When would I use a smaller cap and what criteria would make me increase the cap value?

Hi Jim

You can refer to this file to get more information about why fake OC is influenced by BST cap.

https://www.ti.com.cn/cn/lit/pdf/sloa259

The OC protection detection of our device is to sample the voltage across the MOS when it is turned on, and then divide it by RDS(on) to obtain the current flowing through the MOS. If the capacitance of the BST capacitor is not large enough, the MOS will not be fully turned on, resulting in a large voltage across the measured MOS triggering a false OC.

I have to admit, we are pushing this IC, trying to get all we can out of it. In the app note, it mentions going as high as 0.47uF if you are switching at 400kHz...we are switching at 450kHz. Can I make it that big or can this cause other problems?

From what I am seeing, there also seems to be a thermal element. If we leave them running with no signal for  an hour or so, then hit it with a large bass transient, it will easily trip the fault condition. At this point the heat sink is a little warm to the touch, but can be held with no trouble. Maybe 45C or so.

I tried increasing the BST caps to 200nF, same problem.

Hi Jim

As your describe, it'd hard to locate the problem, I have some conjectures and questions 

1. If this is is a fake OC issue, maybe the temperature influence the value of BST cap, you can try to increase the BST cap to 0.47uF, but a large BST capacitor will affect the startup of the device. 

2. Maybe it's a real OC issue. could you please share  the current wave while fault happen and your schematic? What the output power while the issue happen?

3. " In the app note, it mentions going as high as 0.47uF if you are switching at 400kHz.." could you please share me the link of this application note, I want to check it.

1. Ok

2. I don't think we are getting any where near the 17 Amp current ...experimenting today, it looks like our DSP limiter is keeping the output current below 12 Amps. I saw it trip well below the 12 Amp level.

3. It is in the app note you referenced me to, they mention it twice:

The app note is written for the TPA31xx series, so that could be specific to that series.

One thing I noticed today, the coupling caps going to the inputs are polarized electrolytics and they are reversed from how they should be installed. I replaced them with non-polarized caps on two of the modules and the fault has stopped. I need to test more before I declare that the issue, but it looks promising. Do you think this could be the root cause of the faults?