TAS2563: High THD % on certain input levels only - Odd peak at 60Hz

I tried again with just fresh new .ppc3 configuration, nothing changed from default settings, and measured MLS with -9dBFS, -26dBFS and -27dBFS. Peak is there from 0dBFS down to -26dBFS and then vanishes. What algorithm switches off at -27dBFS?

Hi Mikko,

This 60Hz sine tone is the pilot tone used for IV-sense real-time measurement. At very low input signal the tone is disabled as the protection won't be required and it may cause with low amplitude playback content.

If you test in ROM mode, this 60Hz tone will not show. You can select ROM mode in Device Control panel, top-right corner drop-list. Note that if you go back to Tuning and Audio Processing panel the device will go back to Tuning mode and the 60Hz tone will be back.

Best regards,
-Ivan Salazar
Applications Engineer

Ivan,

As always, thank you for very professional and quick reply!

Ok, Perhaps I am missing here something and made a rookie mistake.
I thought ROM mode does not contain "Tuning and Audio Processing" -modules active.
So if go through End System Integration process while in ROM mode, Smart bass, EQ and other features are disabled.
Am I mistaken?


Second, this IV sense is a fundamental feature of Smart Bass Protection, right? And Smart Bass can not be disabled?

Thirdly, Exactly how high this 60Hz Pilot tone is? On default setting, with only change being Smart Bass High-pass filter tuned to about 20Hz, I can see that Pilot tone is higher than fundamental signal. You can see how lower part of the MLS response turns into a mush, which usually indicates high THD.

In this case, it seems not so, but we can see some anomalies on the measurement for the low end.
 

However, I did also measure THD+N . I do not usually measure that, as in Audiomatica CLIO 14X I use, this measurement is awfully slow, over 15 minutes /sweep.. We can see large amount of non-harmonic distortion. Measurement below is done with -24dBFS input signal. I had to change the THD % scale to 100% to fit the THD+N (red) into screen.

All in all, our ODM manufacturer is worried about high electrical THD in low-mid volumes, and it's effect on speaker driver on long term. is there a way to get rid of it, while retaining audio tuning algorithms active?

I did also measure THD+N from -12dBFS input signal. Looking at THD+N % it is around 8-9%, which is a poor result.

Looking at THD+N as dBV, there is pretty much static noise on both measurements. In this, Red/Orange is -12dBFS, and Blue/Purple is -24dBFS input

Hi Mikko,

The processing features such as speaker protection, EQ, DRC, etc are not enabled in ROM, that is correct.
During end system integration, you still have to select either Tuning mode or ROM mode, regardless of the mode you were set before accessing End System Integration panel.

Smart Bass feature is not present in TAS2563, I assume you mean speaker protection instead? Or is it PBE (Psychoacoustic Bass Enhancement)?
The pilot tone is required for the speaker protection feature. If you don't need the speaker protection and still want to use the rest of the features (EQ, DRC, PBE, DEQ) the protection algorithm can be set to feedforward mode so that pilot tone is no longer used. Otherwise, if you consider that a lower frequency of the pilot tone would benefit, the pilot tone can be adjusted down to 16Hz. For both of these adjustments we'll have to make the changes within TI and send you back a PPC3 file. Let me know and I can help with this.

The pilot tone has an amplitude of 187mV.

In general, the electrical performance is tested using ROM mode, as this is the pure analog performance of the amplifier.
Adding speaker protection and any other processing such as EQ or DRC would affect distortion measurements, as the output is no longer following the input signal.

Best regards,
-Ivan Salazar
Applications Engineer

Hi Ivan,

As always, very professional answer!



Oh, sorry for my confusion. You are right, I mean speaker protection. I was confused by the naming of the menu "Smart Amp".

Pilot tone amplitude of 187mV sounds interesting data, as in my sweep based measurements, it looks MUCH higher. Converted to dBV, 187mV would be -14.6dBV .

To test this, I created new .PPC3, with default settings except under "Smart Amp" menu, I adjusted high-pass filter to 450Hz.
As an input signal I used 4kHz Sinewave with -10dBFS and utilized TrueRMS meter function in CLIO. -10dBFS was selected as it seems to correlate to about 2Vrms in Output.

When measuring with 200Hz lowpass -mode, I can find the pilot tone and level is very similar than what you said. I tried also applying White Noise, Multitones and MLS, measurement result was very consistent. 


I must admit that I am left a bit clueless, why Sweep or MLS based measurements show so much higher result.

Ivan Salazar said:

Otherwise, if you consider that a lower frequency of the pilot tone would benefit, the pilot tone can be adjusted down to 16Hz. For both of these adjustments we'll have to make the changes within TI and send you back a PPC3 file. Let me know and I can help with this.

Yes, I think this is a great idea!

I would like to try Feedforward mode to see how it behaves. I know I am not pushing our speaker driver to extremes and can always make a bit more conservative tuning for the low end, to keep things in better control.

After sending this message, I will try to reach you through private messages so I can share our .ppc3 file to you.

Thanks for helping!

Hi Mikko,

Sounds good, I'll follow up with you over private message for the PPC3 file editing.

Best regards,
-Ivan Salazar
Applications Engineer