• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 1135 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.

TLE2142 Power AMP Gap and Speed

Po-Chen Lin
Prodigy 150 points
Other Parts Discussed in Thread: DAC5652, TLE2142

Hi,

  I use the 10-bit DAC DAC5652 to generate a trapezoid waveform,

the rising time and the falling time are  2us, the hold time is almost 2us.

The following is the waveform I measured from the input  pin(PIN 2 +)  of TLE2142.

Q1 But I found a gap at the rising edge, it is almost 1V in this case as following,

what is this? and how can I tackle  this?

Q2 The slew-rate of TLE2142 + Buffer is extremely slow,

in the spec the min slew-rate of TLE2142 is 27-V/us and the buffer is 75V/us.  How can I improve it.?

The following is the sch of power amp, but I only keep the buffer stage of the power amp.

The probe point of above figure is TP68, There is no load on the point.

The TLE2142 is powered by +36 and 0V

  • 0
    TI__Guru* 93105 points

    Hi,

    You are using a single, ground referenced +36 V supply to power the TLE2142 and LT power buffer. The input signal is ground referenced as well. The region where input suddenly turn on may be due to an input, or output, swing limitation of one/both amplifiers. The TLE2142 does have a common-mode  input range (Vicr) that extends to the negative supply rail. However, I can't quite tell from the buffer specifications if it has a lower Vicr that extends to the negative rail.

    The outputs are the other circucit points where the swing to the negative rail is limited. It appears the TLE2142 will swing to within about 200, or 300 mV, of the negative supply rail. It is about the same for the LT buffer. Both are linear amplifiers that aren't intended to be driven into a supply rail and doing so can have an negative effect on the swing range. You will note that the waveform appears to swing closer to the rail when the transistion is from the linear reagion, to saturation, as seen for the falling edge.

    The sure way to eliminate swinging into and out of a rail is to use dual supplies. You don't have to go all the way to +/-18 V supplies. It could be an assymetrical supply setup like +30 V/ -6 V, for example. If a negative supply isn't readily available, then a charge-pump dc-to-dc converter could be used to generate the negative supply from a positive supply. See TI's Power management products:

    http://www.ti.com/lsds/ti/analog/powermanagement/power_portal.page

    The TLE2142 is our fastest slewing precision, HV (44 V max) amplifier.

    Regards, Thomas

    PA - Linear Applications Engineering

  • 0 in reply to Thomas Kuehl
    Prodigy 150 points

    Hi,
    I remove the buffer as following figure, the problem seems like what you described.


    Maybe I will apply +36V and -5V in revision, but there is a question.
    In my application a output voltage range of 30V~0V will be needed,
    and there will be 16-CH of pre-AMP and PA.
    The backend of this PA is a power BJT pair in following figure,
    and the loading of this AMP is almost a 0.33uF capacitive load.
    The question is that does the big discharge current will sink to -5V power in the falling time?
    In the other word, should I decide a big current capability of -5V power some as the +36V does?

    I didn't meant the slew-rate of TLE2142 in spec is poor,
    it acts poor with the buffer when the feedback loop resistor is 6.2K/1.2K
    (now I change to 1K/200R to improve slew-rate).
    But when I left the buffer unconnected, the TLE2142 performs good.


    May I ask a extra question, it is about my PA, what's the PA type of this BJT backend?
    It doesn't like darlington pair, actually I know not much about PA.

  • 0 in reply to Po-Chen Lin
    TI__Guru* 93105 points

    Hello Po-Chen,

    The question is that does the big discharge current will sink to -5V power in the falling time?
    In the other word, should I decide a big current capability of -5V power some as the +36V does?

    If the Q27 emitter is connected to ground any current throught the transistor will return to the supply ground return. That would be the Vp24 V supply shown in your schematic. However, if the Q27 emitter is connected to the -5 V supply instead, its emiiter current will flow back to it. Therefore, the -5 V supply would have to be able to sink the high return current. 

    The output stage is a complementary push-pull stage. Transistor pairs Q76/Q28 and Q75/Q27 are compound connected in what is referred to as a Sziklai Pair. The upper pair uses NPN/PNP compound-connected transistors, while the lower pair uses PNP/NPN compound-connected transistors. The compound connection provides beta (HFE) multiplication like a Darlington pair, but only requires about 0.6 to 0.7 V to turn on the pair. A Darlington pair requires twice that voltage because the base-emitter junctions of the two transistors are in series. You can learn more about this transistor configuration at:

    http://en.wikipedia.org/wiki/Sziklai_pair

    Regards, Thomas

    PA - Linear Applications Engineering