This thread has been locked.
If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.
Hi All,
“The CLIP signal indicates that the output is approaching clipping. The signal can be used to either an audio volume decrease or intelligent power supply controlling a low and a high rail...”
That’s all I found about /CLIP signal in the TAS5630 datasheet … I would like to use this output to control my DC/DC Converter supplying my TAS5630 based amplifier. The converter will switch the output voltage rails from 25V to 40V in case of amplifier goes to clip the output signal. Therefore I would like to ask some details about it:
- What is the /CLIP threshold? Is it defined as % of the output signal or % of the supply voltage?
- How the /CLIP output behaves – is active just for short time every time when the single signal period is going to be clipped or is detecting the high signal level (still below clipping) and stays for a time to assure power supply have a time to regulate high rail of the voltage?
- What is the /CLIP detect-response delay time?
Thanks for your support!
Cheers!
Tomasz
Tomasz,
The CLIP signal singals if the amplifier is skipping output pulses (clipping). Whenever the feedback tries to produce output pulses twice the duration of the propagation delay of the feedback loop, (or shorter) it starts to skip pulses (this happens at appr. 95% modulation index). Each time a pulse is skipped (the activity of the comparator output is monitored), the /CLIP signal asserts. The /CLIP signal is cleared upon first activity of the comparator output.
/CLIP activity starts outputting short pulses when the amplifier starts to saturate, which puts the operation into a semi clipping mode where one or more pulses are skipped (open loopo operation), then returning into normal operation (closed loop operation). If input signal is increased further, all pulses are skipped (full saturation), /CLIP stays low, and a pulse injector inserts small pulses at 1/3 fs to keep everything alive.
You can use the /CLIP signal to raise the PSU voltage, which should do a fast transition to the high voltage rail to allow unclipped output voltage. It is preferred to let the voltage rail be high for a little time before going back to the low supply. The high-low transition can be made by letting the amplifier drain the PSU bulk output capacitor down to the low voltage rail. We have good experience with this kind of system, power losses are reduced significantly, and it do not influence sound quality.
Tomasz, /CLIP behaves as follows.
- When feedback loop duty cycle exceeds about 95%, /CLIP asserts a low pulse which can be used for external control (like reducing the input signal, increasing the power supply voltage, etc.)
- Response time is less than one clock cycle, and so is /CLIP pulse width.
- So /CLIP does not stay low continuously, and power supply control must allow for this.
This means that, when /CLIP commands the power supply to switch to its high level, it must stay at the high level for several seconds and then fall to its normal level over several more seconds. The time periods must be selected for expected program material.
Our power supply design for TAS5630/31, SLOU293, does these things without audible artifacts. Details may be found in SLOU293.
Regards,
Steve.
Hi Søren, Steve,
Thanks for your detailed description - now everything is clear.
In my application I have connected the /CLIP signal directly to step-up converter and to the interrupt input of the main µC - to have fast response on /CLIP active state as well as defined time keeping the high voltage rail. I’m using a dual-phase step-up converter based on LTC3862 with effective switching frequency set around 900kHz, so two times faster than the TAS5630 is – it should work well without audible artifacts. I will test it as soon as the second board version is ready. Having the board running I will try to post some more information how does it works in my application. This configuration should definitely solve my problem with output filter core losses in the amplifier idle run.
Cheers,
Tomasz
HI All,
I’ve invested some time to measure the /CLIP output behavior in real. I measured how it works in my amplifier. I pulled-up the /CLIP output to 12V because my DC/DC converter needs more than 3V3 to activate high voltage rail.
To observe the clipping really fast, on low power level, I supplied the amplifier from around 25V unregulated power supply and connected 8 Ohm load. At the beginning I found the highest signal amplitude without any /CLIP activity:
As next I tried to observe how the /CLIP outputs reacts/behaves increasing the signal amplitude.
Soren/Steve – exactly how you wrote guys, at first /CLIP output starts outputting short pulses before audible clipping comes.
Ones going into real audible clipping the /CLIP signal stays low for the same time the output signal stays clipped.
Going out of the clipping is similar - first the /CLIP output starts back outputting short pulses and after the output signal decreases to initial unclipped level the /CLIP output goes back high.
You can see the pulses on the zoomed view below:
The /CLIP outputs reacts really nice! It starts inform really before the clipping come.
First I connected the output, through small high side switch to change the logic, to control input of my DC/DC converter but it doesn’t really work well – the /CLIP low periods were simple to short to boost the supply voltage.
I added simple small 10k 10µF RC low pass and now it works really nice. The interesting effect I got, is the DC/DC converter produces exact the voltage value I need to have unclipped signal on the output – looks like the optimal control feedback loop. I tried also to play single period of 100Hz, high amplitude sine wave to see if the loop controls/switches the voltage rail enough fast to react immediately – it looks it does work however I had no more time yesterday to perform exact measurement on it. The new PCB revision is on the way and I will test it definitely having the new PCB.
Tomasz,
Good to see that the device behaves as it is designed to.
Please note that the poen drain outputs are designed to accept pull up to 5V and not 12V as you are using.
Hi Søren
I know for you it is nothing new :) you know exactly how the chip works.
I hope the sceenshotts will be helpful to understand the /CLIP behavior for someone without so deep knowledge about TAS5630.
The measurement I performed definitel helped me to better develop the voltage rail swiching part of my DC/DC converter.
When I completely finish the DC/DC converter with rail switching I can post the whole information in this thread – including schgrematic diagram and PCB of my power supply.
Thanks for the infor about open drain voltage limitation – I have simple overseen it reading the datasheet.
I have definitely to cheange/redesign this part – no problem at all … the issue is, I have only the 25V/40V and 12V voltage available on the board and I do not want to add separate voltage source for it.
Every additional component makes the product more pricy :)