• State Resolved
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 29 subscribers
  • Views 3264 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.

Please help to design a servo for lm4780.

vishnu gopalakrishnan
Intellectual 320 points
Other Parts Discussed in Thread: LM3886, LM1875

can anyone help me design a servo feedback for lm4870 chip.i read out that removing cap from feedback path can really improve sonic performance.

in An1192 a sample circuit is given for servo,but the exact calculation of circuit is not provided.

exact calculation is required for getting minimum offset right

then An1192 is also applicable for lm4780 right?

  • 0
    TI__Expert 3840 points

    Hi Vishnu,

    You are correct that the servo principle is applicable to the LM4780 or any power amplifier for that matter. You are also correct that removing signal-chain capacitors can improve sonics based on reducing non-linearities. Many times signal-chain capacitor values are selected to be large so that they don't negatively impact signal frequencies, thru attenuation or phasing. This is the most cost effective solution as many times, adding servo circuits add PCB area, components and cost. For the most discriminating listener where cost is not a factor, you are correct, the best option is to remove the components completely.

    I don't have an exact calculation for the servo circuit; it is essentially an integrator circuit with a lowpass filter at the input or an RC charge-up circuit. The key concepts for the circuit are:

    1. to provide DC feedback from the output to the input to cancel out the offset from amplifier gain when no DC blocking capacitors are present.

    2. the circuit should have a very long time constant as it is affecting DC only, we don't want it to operate on audio signals.

    3. the circuit should look transparent to the amplifier; ie it shouldn't load the amplifier circuit down in any way. This is why in the text of the application note, it explains that the impedances are scaled to be about 10x the values in the amplifier circuit. If you use the scaling approximations on your circuit values, then I presume that you won't have any troubles. The resistor and capacitor values in the circuit are approximately on the order of what you would find for an audio servo circuit.

    I hope this helps.

    Best Regards,

    JD

  • 0 in reply to John De Celles
    Intellectual 320 points

    Thanks for your immediate replay john,

                                                                         well i know that when we remove the cap from the feedback path we have to provide proper offset correction using a servo mechanism,and as far as i understood the servo feedback actually averages the output dc value from the output of amp and feedback it to inverting terminal so as to attain a 0v offset at output.

                                         But still i can get to design one.can you please conform on this,i have a well designed servo feedback for LM3886 chip,can i just copy and place the feedback for lm4870,

    OR

    can i just just use the feedback circuit from application note.

  • 0 in reply to vishnu gopalakrishnan
    TI__Expert 3840 points

    Hi Vishnu,

    If you have a working servo circuit for the LM3886, I don't see any reason why it wouldn't work for the LM4780.

    Also, you can apply the servo circuit in the application note to the LM4780 as well.

    Either of these two options should work.

    Thanks and Best Regards,

    JD

  • 0 in reply to John De Celles
    Intellectual 320 points

    Hi

        is there some way i can limit the power output from the chip?

    i am designing a 5.1 system and my front and rears are rated for 30W only....

    also how much is the current requirement per channel for 8ohm at 38v supply..please provide a equation

    will it be better to go with lm1875..please also provide the current requirement......do you have any method my which i can bring my +/-36v TO required level for 1875

  • 0 in reply to vishnu gopalakrishnan
    TI__Expert 3840 points

    Hi Vishnu,

    Probably the best and easiest thing to do is to set the amplifier gain according to what you desire for a maximum output power. Then you can use your preamplifier to adjust the volume level accordingly. 30W @ 6.8 ohms is 14.3Vrms and with an input sensitivity of about 0.75Vrms, gain should be about 19.

    Personally, I'd rather have the amplifier more capable than the speakers so that I don't run a clipped output waveform into the speaker. I let my ears do the judging on whether the speakers can handle the power.

    Speaker impedances are not constant over frequency; typically amplifiers are designed for about 6.8 ohms for a typical 8 ohms speaker. For a single 38V single power supply voltage, you will have 38V/2 = 19Vpk not including amplifier saturation, which is probably around 2-3V, so 16Vpk of output swing. This translates into 2.35Arms of output current; 16Vpk/6.8 ohms.

    Checking with the power rating of your speakers: 30W @ 6.8 ohms is approx 2.1Arms from P = I^2R

    The equations for power are: P = VI = V^2/R = I^2R

    If you want to use a lower output power amplifier, the LM1875 would work as well. If you have a +/-36V supply, you would have to regulate the supply down underneath the maximum supply voltage rating of the IC. TI offers many options for this function on the power webpage. You could try experimenting with Webench to come up with the most appropriate design.

    I hope this gets you started.

    JD

     

  • 0 in reply to John De Celles
    Intellectual 320 points

    Hi,

    so i have decided to go with lm3876 for my 5 channels ie fronts surrounds and center.i have uploaded the schematic to get reviewed.i got this topology for diy forum.It has its feed back form output to non-inverting terminal of opamp which is unity gain inverter.the advantage is the ability to use good quality polypropelene cap for that feedback.is the design okay? can some please explain to the the essence of that feedback.

    i will wait for you sugestion...mean while continue with sub amp with lm3886

  • 0 in reply to vishnu gopalakrishnan
    TI__Expert 3840 points

    Hi Vishnu,

    I believe you were trying to utilize a DC servo to null out output DC offsets so that you wouldn't have capacitors in the signal path. The 1M ohm and 1uF cap accomplish this and are the same time constant that I originally suggested from AN1192. This is a standard set of component values for this function.

    I'm not sure that the circuit you have proposed will accomplish your intended goals. If you really want to use this circuit, I would suggest that you either use TINA to simulate it and/or try breadboarding it up to see if it meets your needs. If you are looking for a known good working circuit, the one proposed in AN1192 is a standard way of accomplishing the function and has been verified.


    The essence of the desired feedback is to ensure that DC offsets created by the op amps are nulled out by an amplifier with a slow time constant (ie 1M and 1uF) by connecting the output back to one of the inverting inputs, so that an opposite polarity DC voltage is essentially zeroed out.

    I hope this helps.

    JD

     

  • 0 in reply to John De Celles
    Intellectual 320 points

    Thanks john,

                           yes my aim was to attain dc servo feedback only.if the circuit in AN1192 is tried and verified by you definitely i will go with that design.

    please give me some time.i will complete the schematic as soon as possible and post it over here.i prefer using inverting configuration of power op-amp because i had read it gives much lower distortion i that configuration.I believe its necessary this inverting configuration is preceded by and inverter and a buffer stage