Other Parts Discussed in Thread: TAS5825M,
Hello,
I reached out to TI support and they directed me here. The following indented area is the body of the email I sent. Below that describes the testing I did after sending this support request, and some conclusions I would appreciate confirmation on.
Hello,
My company is using two TAS5828M in bi-amplification configuration in a pair of desktop speakers we are currently developing. In the documentation it states to go to the TAS5825M process flow PDF to get details on the process flows available for configurations. There does not seem to be an exact process flow that corresponds with what we see in Pure Path Console 3, but the best fit is process flow 9 (Base/Pro, 48kHz, 1.1), but our workflow does not have the phase optimizer, DPEQ, clipper, and is 96kHz instead of 48kHz. This is also the only process flow available when in the 1.1 configuration.
Background information out of the way, our question is with regards to the thermal protection that is present while in this process flow. There are no customizable automatic gain limiting, or over-temperature foldback settings for this configuration/process flow like there are in other configurations (for example Base/Pro 1.0 48k). I understand that with more demanding configurations/process flows that handle higher sample rates or process two signals have to cut down on features, but I am wondering if the features are still running and are just not customizable.
To summarize my questions:
In Base/Pro, 96kHz 1.1, is there automatic gain limiting or over-temperature foldback systems running, even if not customizable?
If no, is there any thermal protection present? To what extent?
Will there be a process flow document specific to TAS5828M available in the future?
Testing done after sending support ticket:
- In our current PPC3 configuration, I played pink noise at above rated power through our speakers to try and activate an overtemperature shutdown state. While doing this, I watched the register map for changes in 0x72 and 0x73. While 0x73's first four bits went from 0 to 1 in order, the output power I was measuring did not change (indicating that it was not doing over-temperature foldback). Shortly after 0x73 bit 3 flipped to 1, 0x72 bit 1 (over temperature shut down fault) flipped, and the unit went into thermal shutdown mode.
- Conclusion: in 1.1 configuration, process flow Base/Pro 1.1, 96kHz, the only thermal protection is the over-temperature shut down.
- I switched our speaker to work in Base/Pro (2.0, 48k) so that I could use the PVDD AGL/OTFB module. I assigned the tweeter and woofer channels to be left or right, then used un-ganged EQ bands to set up a simple diplex filter (it wasn't ideal, but it worked). When the temperature foldback was enabled, I found that while repeating the previous test, the output power decreased as 0x73 bits 0-3 flipped to 1. After bits 0 and 1 were flipped, it stably operated at ~80% of rated power for as long as the test continued. I also stopped playback for a while after this to see if the warning bits would return to 0, but they wouldn't unless I manually set 0x80 bit 7 (write clear bit) to 1, then pressed 'Read All Registers'. I also found that even though some of 0x73's warning bits may have been flipped to 1, if the speaker was given time to cool down, the speaker would return to rated power.
- Conclusions:
- The temperature foldback settings in 2.0 configuration limited output power to prevent the system from overheating. In Bi-Amp configuration, it does not seem like this feature is available.
- The operation of the temperature foldback feature is independent of the values of 0x73.
- Warning and fault bits are not reset when the condition is no longer there (i.e.: when temperature drops below 134C, 0x73 will not go from '00000111' to '00000011') unless 0x80 bit 7 is flipped.
- Conclusions:
Thank you for taking the time to read my question. Please let me know if there is something to add, or an aspect I have overlooked.
