TPA3116D2 protection

Hi,

I am designer of audio amplifiers for PA companies.

We got TPA3116D2EVM from MOUSER and operate it on 20V.

The problem described above happen to us also. (works fine at 8OHM but goes in to limit easily at 4OHM or less).

Maybe there is an initial problem in the entire design of this IC.

If it works for somebody as should, please let us know.

Hello, Dror and Manuel. Sorry you're running into trouble with lower load impedances!

When the TPA3116D2 drives a load less than 8 Ω in value, sometimes thermal issues arise. Dror, on your board, have you used a similar heat sink to the one we recommend? Have you applied thermal paste between the pad of the IC and the heat sink? Manuel, you're using the EVM, so that shouldn't be the issue, but can you confirm the heat sink is securely fastened?

Do you have a way to measure the temperature at the top of your heat sink? If it approaches anywhere near 100°C, that's a definite sign of thermal shutdown.

It's very helpful for me to know the numbers for output levels that are causing the amplifier to mute or shut down. Do you have a way to measure the power driven to your load when the amplifier shuts down? Do you have a way to measure the current coming out of the amplifier? (A current probe on the output wire should give us some insight.) What do you measure at the output for "strong" music?

Can you please use an oscilloscope on the FAULT pin to see if it goes low (normally logic level high) when this problem occurs?

What gain setting are you using in your individual setups?

On your power supplies, what is the current limit set to? Sometimes if the peak power required by the amplifier can't be met by the supply, the sound won't come out as expected.

Best Regards,
Matt

Thanks for answering Matt,

It will be a big surprise if it is thermal problem:

We use 2 units, cooled together with one large heat sink, 4.8 times better W/Celsius than the one recommended in the Data Sheet. Thermal coupling is perfectly done with the right Silicon Paste ( We are experienced Amplifier manufacturers for over 40 years... We have amplifiers with TAS 5611/13 which works perfectly even with LIne transformers loading, just they are for 24V and we look now for 12V operation).  The problem accrues the minute the amplifier is turned on.

After we checking with the 2 IC's, we operate now just one, so, cooling is even better per one unit.

Our item works with 20dB gain. We tried different gains - it did not help. We tried the values as in EVM , did not help either.

As I thought it could not be that the IC's which should work at 2OHM  that cannot operate at even 4OHM, the power supply became suspicious. Instead of using the internal switching power supply, we had connected to well stabilized 20V Lab type Power Supply - At peak of 1.8A the sudden mute when loud passages are played, or just shouting a bit into microphone, happened again. This is at 4OHM  so the momentary current into load should be some 3A, calculated at 100% theoretical efficiency.

At first we thought it may belong to use of long wires or complex loads, however, even with 10cm cables and pure resistors the problem is there.

As it is late in the evening here already, we can check the FAULT pin reaction just tomorrow , other than that, I believe I gave all possible information.

Kind regards, Dror

Thank you for following up, Dror. It's a relief to know it isn't a thermal issue.

What is the input amplitude of the loud passages when the muting occurs?

Along with checking the FAULT pin, please collect some measurements of the actual power driven to the load at these loud audio passages, as well as the actual current measurement if you are able to. For power, you can look at instantaneous voltage delivered at those peak times if you don't have a dedicated power measuring system (like an Audio Precision). Please capture any oscilloscope plots and post them here.

It looks like you're pretty new to the TI E2E Community. Welcome! If you have trouble with posting scope plots, you can do so in the reply window. Click on the square, green icon in the bottom row of the editing window. It will allow you to attach a picture. The paperclip next to that icon will allow you to attach any PDFs or documents you'd like me to see.

Best Regards,
Matt

Yes Matt, I am new here as it is first time for me to face such a problem after almost 40 years of experience in audio design.

Unfortunately, even it is my hobby to develop such products, I own large companies that do not leave me a lot of time to deal with such matters.

The voltage of signal that caused the protecting mute was 0.72V peak tompeak at input of power amplifier section.

We have AP2700 as well as other equipment and we will take measurements tomorrow even we had started today already to develop the power amplifier section around the NXP TDF8599A. We do not face any problems with this IC ( sound quality is a bit inferior, but we need it for Background Music, EVacuation and such application where reliability is the main issue and not sound quality).

Kind regards, Dror

We just got a new batch of 3116D. They work just fine with hIgh end sound quality. So, there was problematic batch and new ones are very good! Dror

Hi, Dror. How peculiar that the old batch of ICs had the problem, but the new ones didn't. Thank you for following up to let me know! Please let me know if you need any other help in your design.

Best Regards,
Matt

Halo Mat, 

Design had been finished few weeks ago, just there was the problem with lower impedance (we had 4 units in hand, all had the same problem). I had contacted Yan from Spain that faced similar problem and his EVM had the IC from the same batch. Now all is fine for both of us. We just got the first reel of 2,000 pcs. from MOUSER in China and will check that they work well too.

Item is perfect now and we run 2 types in PBTL: Battery operated mobile system with ODM 2 Ohm 10" speaker and Hi Fi type but for our use in PA, with or without out put line transformer, 2 - 6 channels. I have recommendation for you: do not use Electrolyte Polarized  Capacitors in inputs, not here and not for the 5613 / 5630! We use  MKP 2.2u 100V for high end and Bi Polar for PA. Better sonic results both at hi and low ends, more transparent sound. The sound quality is much better than NXP.

Kind regards, Dror