• State TI Thinks Resolved
  • Locked Locked
  • Replies 18 replies
  • Answers 2 answers
  • Subscribers 30 subscribers
  • Views 16309 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPA3116D2: Input DC offset and difference with TPA3118

B2B
Prodigy 110 points
Part Number: TPA3116D2
Other Parts Discussed in Thread: TPA3130D2, TPA3118D2, , TPA3251, TPA3221

I have few questions about this Class-D amplifier chip, which are not clear and can't be found in the datasheet.

First of all, what is the DC voltage offset on the input pins (4,5,10 & 11) .

Is that just 1/2 PVCC?
Although the datasheet also states that the maximum input voltage can only be 6.3V max.

Seconds question, what is the real difference between the TPA3116 and TPA3118 (and even the TPA3130) ?

According to the datasheet, the only difference between the 3116 and 3118 is the position of the thermal pad.

However, that would mean that if you would make the cooling adequate enough, the TPA3118 could also deliver 2x50W?

The datasheet also only mentions one Rds(on) value, so that would suggest that the TPA3130 is the same as the other two.

The only clear difference is in the thermal information. Although, ones again the TPA3118 has similar or even better results compared to the TPA3116.

What are the physical differences between the three?

Last thing, is there any information about oscillation frequency and performance?

  • 0
    TI__Mastermind 24835 points

    Hi Bart,

    The pins(4,5,10 &11) are the input pins of the device, and they are biased at around 3V. They can sustain -0.3V ~ 6.3V input voltage.

    The OC THR for TPA3116D2 and TPA3118D2 are 7.5A, and TPA3130D2 is 4.5A. It means that TPA3130D2 can only support lower output power application. While for TPA3116D2 and TPA3118D2, the only difference is the thermal pad-up or pad-down. TPA3116D2 is a pad-up device, so it supports a heatsink, then it has the ability of outputing higher power with a heatsink, comparing to TPA3118D2. They have the same Rdson, but different output power ability: 3116>3118>3130.

    TPA3116/3118/3130D2 has oscillator circuit amploys a multiple frequency option(400kHz, 500kHz, 600kHz, 1000kHz, 1200kHz are available).

    Best regards,

    Shawn Zheng

  • 0 in reply to Shawn Zheng
    Prodigy 110 points

    Hi Shawn,

    Thank you for your reply.

    It only doesn't completely my answer.

    If the difference is only the Power PAD up or down, that also means that if you have an adequate thermal design, they should be able to handle the same amount of power?

    Also, the packaging of the 3118 and 3130 are identical. So why should there be a difference between to two?


    What tolerances are with the 3V bias?

    (this is an important parameter for a design I am making)

    Are there any performance graphs (THD+N vs power vs freq) for different oscillation frequencies?

    Is post filter feedback (PFFB) also available with these chips?

    This option was mentioned in the TPA3251 datasheet, if I compare the functional diagrams, I don't see any differences between the two.

  • 0 in reply to B2B
    TI__Mastermind 24835 points

    Hi Bart,
    Please find my answers below:
    If the difference is only the Power PAD up or down, that also means that if you have an adequate thermal design, they should be able to handle the same amount of power? [Shawn] The pad-up device basically uses a heatsink to dissipate heat, while the pad-down device mainly depends on the PCB which the bottom thermal pad is soldered on. The different packages produce the different thermal performance. Generally TPA3116 has better thermal performance than TPA3118. I fully agree with you that theoretically they could handle the same heat dissipation as soon as TPA3118 has the adequate thermal design. But in reality the thermal performance for PCB is far worse than a heatsink.  

    Also, the packaging of the 3118 and 3130 are identical. So why should there be a difference between to two? [Shawn] I guess the 'difference' which you mentioned is for the output power ability, right? Package is not the only factor for the output power ability. A lot of our products use HTSSOP28 package but have the different thermal performance. Many factors(e.g. different Rdson, different OC THR...) could affect the output power.

    What tolerances are with the 3V bias? [Shawn] 3V DC bias is from the INN/INP, you don't need add a bias externally.

    Are there any performance graphs (THD+N vs power vs freq) for different oscillation frequencies? [Shawn] Sorry, we don't have this data. The data shown in the datasheet is based on 400kHz frequency, but the THD performance should be very similar for different oscillation frequencies.

    Is post filter feedback (PFFB) also available with these chips? [Shawn] TPA3251 is a high power device, I'm not familiar with it. Technically, TPA3116D2 can work with PFFB circuit, but very few customers use it in their design. One thing to be noted is that please design/tune the filter carefully to make sure the loop is stable/robust.

    Best regards,

    Shaw Zheng

  • 0 in reply to Shawn Zheng
    Prodigy 110 points
    I understand that there are 'some differences' in different packages.
    My questions is what kind of differences are these exactly?
    So in other words, why is the 3130 rated for only have the load (twice the minimal resistance)?
    There is absolutely nothing in the datasheet that explains this, but can be vital in board design.

    I understand that there is no external BIAS needed (otherwise you wouldn't use a blocking capacitor normally).
    I am asking what the tolerance is of this BIAS voltage.
    Is this related to in maximum input voltage for example?

    About the PFFB, the TPA3221 is a very similar device and will perform very similar with the same rail voltage (say 24V).
    Also in this datasheet the PFFB option is being mentioned., but there is no notification about tuning and stability.
    As well as in the associated application note.
    Unfortunately the TPA3221 also doesn't have any measuring data in the datasheet at all.
  • 0 in reply to B2B
    TI__Mastermind 24835 points

    Hi Bart,

    For your first question, are you asking why TPA3130D2 has 8ohm NORM load while TPA3118D2 has 4Ohm in BTL mode? Even they have totally the same package, the OC THR is different. As I mentioned above, the OC THR on TPA3130D2 is 4.5A and TPA3118D2 is 7.5A, becasue TPA3130D2 is only for lower output power application. TPA3118D2 can support lower load resistance wihout OC FAULT becasue higher OC THR. Please let us know your application, e.g. PVCC, Load resistance, Output power... so that we could recommend the appropiate device for you.

    The DC bias voltage is fixed and not related to input voltage. The maximum ratings on INN/INP pin is -0.3~6.3V. There is no minus voltage inside the device, and the input signal is biased at around 3V.

    About the PFFB, we don't have resommendation application for TPA3116D2/TPA3118D2. Please find this app note http://www.ti.com/lit/an/slaa702/slaa702.pdf it's for high-power devices, but mid-power should be very similar.

    Best regards,

    Shawn Zheng    

      

  • 0 in reply to Shawn Zheng
    Prodigy 110 points
    So what is the reason why the TPA3116 has a lower over current threshold?

    How much is the tolerance around the 3V input bias?
    Is there another reference pin with this 3V bias voltage?
  • 0 in reply to B2B
    TI__Mastermind 24835 points
    Hi Bart,
    I can't understand why you think TPA3116D2 has a lower OC THR. TPA3116D2 has 7.5A OC THR and TPA3130D2 has 4.5A OC THR, because TAP3130D2 is for the lower output power application. You could find this info in the datasheet.
    Could you please let me know why you need the tolerance of input DC bias? Generally the input signal amplitue peak value should be less than 2.6V to avoid clipping, because the min gain is 20dB and the max power supply voltage is 26V. I believe the tolerence on the input DC bias voltage is far less than 0.4V. I don't think there is any reference pin for the bias voltage.
    Best regards,
    Shawn Zheng
  • 0 in reply to Shawn Zheng
    Prodigy 110 points
    Obviously this was a typo, I meant the TPA3130.
    It's also very clear to me that the 3130 is for lower power applications and has a tighter over current threshold.
    We have been discussing that now over and over again.
    My question was and still isn't answered , what are the differences between the 3118 and the 3130?
    Because there has to be a reason that the 3130 uses a lower over current threshold?
    Or are these actually just 3118 chips which didn't meet the specs?

    This customer wants to avoid (expensive) capacitors.
    So the input bias voltage could be used to also bias the analog (filter) circuitry.

    It only really baffles me why all this standard information isn't simply in the datasheets to be very honest.
  • 0 in reply to B2B
    TI__Mastermind 24835 points

    Hi Bart,

    I think I explained this previously, the ONLY difference between TPA3118D2 and TPA3130D2 is the OC THR. They are trimmed with different OC THR in the factory. All of the devices have to meet their own spec requirement, else they are scrapped.

    No, input capacitors have to be used in the application. The DC bias on the INN and INP has to be the same, any unbalance could cause pop noise and even DC FAULT. Another potiential issue is in power-up sequency. If the external DC bias voltage is applied before the PVCC, current goes into the device input pins, which probably damage it. If the external DC bias is applied after PVCC, then the device is set to PBTL mode unexpectedly because the left channel inputs are connected to GND.1uF ceramic capacitors can be used for them, and I don't think they are quit expensive.

    There may be some typos in the datasheet, but all of the data in it was verified and is repeatable. Please feel free to correct us if you find anything incorrect.  

    Best regards,

    Shawn Zheng

  • 0 in reply to Shawn Zheng
    TI__Mastermind 24835 points

    Hi Bart,

    Please feel free to let us know if you have more question on this. Could you please help to close it if it's answered?

    Best regards,

    Shawn Zheng

  • 0 in reply to Shawn Zheng
    Prodigy 110 points

    It still doesn't make a lot of sense.
    The TPA3130 is a lot cheaper than the TPA3118, which is rather strange since you're saying that they're 100% identical EXCEPT for the over current threshold setting?

    I guess the point is that it's very confusing, even for people who are already in the audio field for many years.

    The only difference I see, is in the thermal Information, which suggests that the ICs are not identical, but do have some physical differences.

  • 0 in reply to B2B
    TI__Mastermind 24835 points

    If you read the datasheet, you could find that the OC trip point is very different between 3118 and 3130. For 3130, it's only 4.5A and it's 7.5A for 3118. this makes 3130 can be only used for lower output power application.

  • 0 in reply to Shawn Zheng
    Prodigy 110 points
    You are giving the same response over and over again, but it's still not an answer to the question.
    Yes, I have read the datasheet, yes it's very obvious that the OC trip point is different yes that (automatically) means the the 3130 can be used for low power applications.

    The question is, WHY?
    Especially because the 3130 is cheaper.
  • 0 in reply to B2B
    TI__Mastermind 24835 points

    Our products need to cover all of the mid-power output range(from 5W to more than 50W). 3118 and 3130 are defined for the different application requirement. The device with higher output power ability is more expensive of course. Is this clear for you please?

  • 0 in reply to Shawn Zheng
    Prodigy 110 points

    Well, I guess it's finally somewhat more clear, but it doesn't make a lot of sense.
    The OC threshold we are talking about is a worst case threshold, to prevent any excessive damage.
    Since the 3118 and 3130 are technically identical, I don't see a reason why you would change that value and as well as for the price difference

    (since it's basically the same chip, so no it's doesn't make sense that the higher power is more expensive)
    Maybe because a lower power system would have a smaller (or no) heatsink, but than again you have your over temperature protection to keep you safe.

    From a customer perspective you have to use the PLIMIT option.
    A pre-programmed current limiter to some kind of value isn't really going to be very useful.
    It really depends what type of driver you're using.

    If I might give a suggestion, for a featured product.
    It would be much more useful to have a pin to set a custom current limit with a resistor value.
    Similar like the PLIMIT.
    Although, I would suggest using a one resistor setting (so you need a tiny little current source inside the chip).
    Same for the PLIMIT.

  • 0 in reply to B2B
    TI__Mastermind 24835 points

    Hi Bart,

    Your understanding is right. Thanks for your suggestion, we will consider it seriously. Do you have any other question on this? Could you please help to close this question if not?

    Best regards,

    Shawn Zheng

  • 0 in reply to Shawn Zheng
    Prodigy 110 points
    Well, my other question about the PWM frequency is still not being answered.
    That being said, to give some more suggestions.

    I noticed that a lot of similar audio chips in your range all have different graphs and measurements.
    This makes it very difficult to compare them, as well as with competitors.
    One important parameter is the THD vs PWM frequency, and frequency response under different load conditions (4,8,16 ohm and infinite), which are missing.
    It would really be helpful to change this to a more standard type of format.
  • 0 in reply to B2B
    TI__Mastermind 24835 points

    Hi Bart,

    Usually we show the THD performance data with 400kHz PWM frequency in the datasheet, because there is no much difference in THD performance between different switching frequency basically, and 400kHz is a common setting. And for load resistance, we usually show 4, 8 and 2ohm cases. Because they are also the common cases for the real speakers. We will consider your suggestion seriously. Thank you!

    Best regards,

    Shawn Zheng