| Electronics > Projects, Designs, and Technical Stuff |
| Headphone Amp front end design. |
| << < (4/4) |
| SiliconWizard:
--- Quote from: NiHaoMike on November 01, 2018, 03:02:29 am --- --- Quote from: SiliconWizard on October 31, 2018, 05:25:25 pm ---I was first a bit wary of the 39.2 ohm series resistors at the heaphone outputs, but apparently this is what they recommend on the example applications in the TPA6120A2's datasheet. Still think that may be a bit on the high side, dissipating unnecessary power and giving too much attenuation for lowish impedance headphones (such as 16 ohm which they claim this chip supports). This will certainly work, but you may want to test this and maybe lower their value a little depending on the headphones you're going to use. --- End quote --- The power into a pair of headphones at usual listening volume is well below -20dBm. In fact, it can often be helpful to increase the resistance to get the usable SNR up, especially when using mass market parts. --- End quote --- Didn't I say that it would essentially depend on the headphones impedance you use? With 16 ohm impedance headphones, which are common in the higher-end range of headphones, it's almost guaranteed that the max sound level with those series resistors' value will be way too low. They are just giving values that are OK for an average use and that would limit the output power to safe levels for average headphones. Given that the op's project rather looks like an "high-end" headphones amplifier (XLR inputs, metal chassis, ...), he's bound to use high-end headphones with it IMO. So I'm just pointing this out. High-end headphones amplifier can deliver up to 1W RMS of power to adequately drive some high-end headphones. |
| Jbliss:
--- Quote from: Jbliss on October 31, 2018, 11:58:40 pm --- --- Quote from: SiliconWizard on October 31, 2018, 05:25:25 pm ---First off - your schematic is rather neatly laid out. Maybe I would have flipped (mirrored) the XLR symbols so as to avoid crossing the + and - lines, but that's a detail. I was first a bit wary of the 39.2 ohm series resistors at the heaphone outputs, but apparently this is what they recommend on the example applications in the TPA6120A2's datasheet. Still think that may be a bit on the high side, dissipating unnecessary power and giving too much attenuation for lowish impedance headphones (such as 16 ohm which they claim this chip supports). This will certainly work, but you may want to test this and maybe lower their value a little depending on the headphones you're going to use. Another point is about the separate grounds. I don't know how you're going to do the PCB layout, but this split ground thing is a recurrent topic of discussion here. I personally would just use two big ground planes (on each side, going for a 2-layer board) and avoid splitting the grounds and joining them in one point, which is usually a bad idea in spite of still being regularly suggested. Obviously others may have a different take on this. Lastly, I would have used the opamps (op1611) as non-inverting summing amplifiers instead of inverting. Not that it will matter a whole lot, but you would avoid an unnecessary phase reversal and possibly a marginally higher noise (although it would probably again not matter here). It would load the volume pots less as well. Just a thought. Other than that, I haven't seen anything problematic so far. --- End quote --- Thank you! Great idea regarding the OP1611 as non inverting! On the topic of grounding I was basing the grounding system off this https://www.rane.com/note151.html I suppose my idea was to run a trace from C3, C23, C33, C43 back to there respective Pin 1's. The pin 1's would be tied to chassis Ground at a star point where both grounds would meet. This could be a terrible idea? any opinions on the best way to do this ? Thanks Again JB --- End quote --- Anyone have any thoughts on my plan for the grounding system ? Starting to layout board would be good to get some opinions ? |
| technix:
--- Quote from: Jbliss on November 04, 2018, 02:30:49 am --- --- Quote from: Jbliss on October 31, 2018, 11:58:40 pm --- --- Quote from: SiliconWizard on October 31, 2018, 05:25:25 pm ---First off - your schematic is rather neatly laid out. Maybe I would have flipped (mirrored) the XLR symbols so as to avoid crossing the + and - lines, but that's a detail. I was first a bit wary of the 39.2 ohm series resistors at the heaphone outputs, but apparently this is what they recommend on the example applications in the TPA6120A2's datasheet. Still think that may be a bit on the high side, dissipating unnecessary power and giving too much attenuation for lowish impedance headphones (such as 16 ohm which they claim this chip supports). This will certainly work, but you may want to test this and maybe lower their value a little depending on the headphones you're going to use. Another point is about the separate grounds. I don't know how you're going to do the PCB layout, but this split ground thing is a recurrent topic of discussion here. I personally would just use two big ground planes (on each side, going for a 2-layer board) and avoid splitting the grounds and joining them in one point, which is usually a bad idea in spite of still being regularly suggested. Obviously others may have a different take on this. Lastly, I would have used the opamps (op1611) as non-inverting summing amplifiers instead of inverting. Not that it will matter a whole lot, but you would avoid an unnecessary phase reversal and possibly a marginally higher noise (although it would probably again not matter here). It would load the volume pots less as well. Just a thought. Other than that, I haven't seen anything problematic so far. --- End quote --- Thank you! Great idea regarding the OP1611 as non inverting! On the topic of grounding I was basing the grounding system off this https://www.rane.com/note151.html I suppose my idea was to run a trace from C3, C23, C33, C43 back to there respective Pin 1's. The pin 1's would be tied to chassis Ground at a star point where both grounds would meet. This could be a terrible idea? any opinions on the best way to do this ? Thanks Again JB --- End quote --- Anyone have any thoughts on my plan for the grounding system ? Starting to layout board would be good to get some opinions ? --- End quote --- If your board have 4 layers or more, having a continuous ground plane in one of the middle layers can be a better idea than star grounding, and it puts less stress on the layout. |
| Jbliss:
There is a ground plane, this is just for the seperate chassis ground. |
| Navigation |
| Message Index |
| Previous page |