• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 33 subscribers
  • Views 4769 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

TPA312x Start up Failures

Andrew123
Prodigy 20 points
Other Parts Discussed in Thread: TPA3123D2, TPA3124D2, TPA3125D2, TPA3122D2, DRV134, NE5532, TL072

Hello,

I am attempting to create an amplifier with an output of 20-25w into an 8ohm load. The IC I am using to attempt this is the TPA3123D2 in BTL mode. I have built the circuit exactly as per the data sheet schematic. The issue is the amp will not come out of start up. I can see/hear a ticking happening about every 250ms. If I remove the output capacitors to ground the amp will start up but you cannot drive it very heard before it shuts down again. I am speculating the output short circuit protection is kicking in.

The unusual part is when I replace the TPA3123D2 with a TPA3124D2 in the same physical circuit it will work. I had the same experience with the TPA3122D2 in previous testing. The output will tick looking like the output short circuit protection is kicking in. Replace the TPA3122D2 with a TPA3125D2 in the same physical circuit and it works.

I have driven the inputs with a TL072, NE5532, and DRV134 based circuits. No change. I have try running from 12v to 24v.

What could be the differences between the TPA3123D2/TPA3124D2 and TPA3122D2/TPA3125D2 that would cause the output protection to trigger if that is what is happening ? How do I resolve this ?

  • 0
    TI__Guru**** 189399 points

    Hi, Andrew,

    Wow, that is VERY odd.. There shouldn't be a significant difference between those parts' short-circuit operation. As you guessed, you are indeed tripping the short-circuit protection.

    Do you have a tight PCB layout with low-inductance traces and great ceramic decoupling? Sometimes this can cause false OC trips.

    -d2

  • 0 in reply to Don Dapkus
    Prodigy 20 points

    Hi Don,

    I agree its odd which I why I am wondering if there was a change in the different ICs.

    Right now we are still evaluating different ICs so the core circuit is built on a temporary proto board (breadboard). This allows the different ICs to be dropped in for testing using the same circuit hence why there is some confusion as to why some ICs work while other don't and if there had been a change in their specs. The TPA312x are being connected through a SMD to DIP adapter board (pro quality with an exposed pad connection) except the TPA3122/25 as they are already DIP. The same SMD board is being used for all the different ICs. We have tried three or more of each IC to make sure it was not just an odd IC or bad soldering.

    Being on a breadboard it is difficult to control trace conditions. We are using ceramic caps for decoupling, low ESR electrolytic for larger values, and stack metal film for the output all being as close as possible to the IC.

    I should also mention this happens in both SE and BTL modes.

    Andrew

  • 0 in reply to Andrew123
    TI__Guru**** 189399 points

    Hi, Andrew,

    Any update on this? I hope you figured it out since I wasn't of much help! 

    -d2

  • 0 in reply to Don Dapkus
    Prodigy 25 points

    Hey! I know this is an old thread, but I had the same problem and just figured it out. 
    Promlem's hiding in Mute and Shutdown features. You shouldn't connect mute/shutdown pins directly to VCC/GND rails, connect them via an RC low-pass filter and everything will be fine. 
    I think that it should be be mentioned in next datasheet revision, not an obvious issue as for me.

    P.S.
    Sorry if my english isn't that good, i'm a little bit russian.

  • 0 in reply to Eugene Negrobov
    Prodigy 10 points

    Hello, I experienced the same with TPA3122: it clicks at the startup for al least one second and then it gets running but it doesn't support significant output power (about 2 or 3 watts) and it mutes at the first input signal peak, then comes the oscillating outputs. Even worst, I purchased 50 pieces of 3121D2 and I can not make them work yet. some of them do it but the most of them are dead by now: signal input pins reads 0-volt instead the bias voltage.

    I think Eugene Negrobov gave us a not  so obvious but useful aspirin but I really think we need a TI solution. Any idea about?

    P.S. Please, don't read me as angry, I'm just desperate :)

  • 0 in reply to Cesar Acosta
    Prodigy 190 points

    I have the same problem and if someone has a modified diagram that would fix this please post it. :http://e2e.ti.com/support/amplifiers/audio_amplifiers/f/6/p/366171/1286765.aspx#1286765

  • 0 in reply to Curt Gibson
    TI__Guru**** 189399 points

    Hi, Guys,

    From what has been posted, the problem is the Short-Circuit Protection is tripping. This can be due to one of a couple things:

    1. Incorrect output filter design - see app note on LC Filter Design.

    2. Poor decoupling of the power supply - either due to insufficient capacitance, or more commonly, poor PCB design. Try to copy the EVM as closely as possible to insure success.

    3. Leakage current on the BYPASS capacitor can cause start-up problems. If the BYPASS cap doesn't charge to its full value, the device won't start up. The voltage at this pin is nominally equal to AVCC/8.

    -d2

  • 0 in reply to Cesar Acosta
    Prodigy 25 points

    Hey, mate!
    I think your problem is caused by bad decoupling, put every capacitor you see on the the figure 28(page 15), caps on 1-20 and 10-11 pins must be as close as possible. Also, don't be greedy, put 1 or even 10 uF cercaps there, the bigger decoupling- the better.
    Also, as I mentioned, there must be RC filters on MUTE and SHUTDOWN pins, like that:

    Without them this thing will self-oscillate at the power up.
    This circuit must work without any problems even on a breadbord, powered by noisy switched-mode power supply, lol.
    :D

    P.S.
    Thanks for awesome IC, TI.
    Right now I'm listening music through my speakers with TPA3122 inside while writing this text.
    ;3