A couple of TPA6120 chips arrived in the post the other day, I was anxious to hear one in operation, so I scrambled together a PCB.
The first problem was the pad in the middle of the bottom of the chip. This needs a solder connection to the ground plane. I solved this by drilling a 3mm hole right through the PCB that I could fill with a solder puddle. I’m pretty sure there’s a connection to the copper on the topside of the board, I made no effort to clean up the rough metal edge left by the drill, and I believe the chip is stood off the surface by a few thou. Anyway the chip is not getting hot and has survived a few hours under drive by now.
The next problem was components. I have, like most enthusiasts, a reasonable selection of caps and resistors on hand, but not everything I needed. There is a Maplin’s not far from me, but they don’t carry the range they used to. A quick check on the net showed that they had positive and negative 12V regulators, electrolytics and polyester layer caps available though. I had a 0-15V 0-15V Myrra transformer and some SMT 16V tants, so with the 1206 resistors and 0805 caps in hand and a couple of SMT LME49710s I figured I was OK for the first try. The bulk capacitances were decided more on the basis of component size and availability than on calculation and this has resulted in the mix of SMT and thru hole components.
The datasheets are full of dire warnings about feedback and layout so I decided to pretty much follow the recommendations, establish a baseline for future reference, keep things simple and try to get a working amplifier at the first attempt.
I followed the datasheet recommendations using the non-inverting configuration and used a 1k feedback resistor with a 330R for a modest gain of 4. The TPA was preceeded by a LME49710 unity gain buffer. DC offset at the output is 0.1mV one channel and 0.2mV at the other. The amplifier itself seems to be close to ambient according to my finger, although the transformer, which is 15VA is warm (not very, but perceptibly).
How does it sound? I have a pair of Behringer studio-type cans (not wonderful, but OK) and some Shure E2C in-ear buds (wonderful). In both, however, the sound is solid and reassuring. You don’t feel anything is missing.
w
The first problem was the pad in the middle of the bottom of the chip. This needs a solder connection to the ground plane. I solved this by drilling a 3mm hole right through the PCB that I could fill with a solder puddle. I’m pretty sure there’s a connection to the copper on the topside of the board, I made no effort to clean up the rough metal edge left by the drill, and I believe the chip is stood off the surface by a few thou. Anyway the chip is not getting hot and has survived a few hours under drive by now.
The next problem was components. I have, like most enthusiasts, a reasonable selection of caps and resistors on hand, but not everything I needed. There is a Maplin’s not far from me, but they don’t carry the range they used to. A quick check on the net showed that they had positive and negative 12V regulators, electrolytics and polyester layer caps available though. I had a 0-15V 0-15V Myrra transformer and some SMT 16V tants, so with the 1206 resistors and 0805 caps in hand and a couple of SMT LME49710s I figured I was OK for the first try. The bulk capacitances were decided more on the basis of component size and availability than on calculation and this has resulted in the mix of SMT and thru hole components.
The datasheets are full of dire warnings about feedback and layout so I decided to pretty much follow the recommendations, establish a baseline for future reference, keep things simple and try to get a working amplifier at the first attempt.
I followed the datasheet recommendations using the non-inverting configuration and used a 1k feedback resistor with a 330R for a modest gain of 4. The TPA was preceeded by a LME49710 unity gain buffer. DC offset at the output is 0.1mV one channel and 0.2mV at the other. The amplifier itself seems to be close to ambient according to my finger, although the transformer, which is 15VA is warm (not very, but perceptibly).
How does it sound? I have a pair of Behringer studio-type cans (not wonderful, but OK) and some Shure E2C in-ear buds (wonderful). In both, however, the sound is solid and reassuring. You don’t feel anything is missing.
w
Last edited:
There were a few problems with the build, the clearances on the PCB are small, and I had to clean up a few solder strings where I dragged the iron inadvertently. I have increased the clearances in the layout to make this less of a problem in future. All vias are oversize from standard to ease hand drilling. The copper fill at the top of the board has been left to economise on etching.
I have been meaning to buy a toaster oven for reflow to replace the one I used to have, but I haven’t got around to it yet. Producing solder stencils is a bit of a problem as I no longer have access to a laser cutter, I have thought about buying some spray-on ultraviolet resist and making them in brass sheet but although I’ve got the brass I still haven’t got the spray.
I still haven’t solved the issue of PTH and vias. I tried conductive paint but didn’t get far, perhaps it’s a question of technique.
I have read some people’s concerns regarding output impedance. I am therefore looking for a small 1206 inductor with which to replace the 10R resistor at the output. I have also redrawn the circuit with some global feedback, the values shown are not set in stone, and although rework is not fun on a board such as this, the new layout leaves room for some variations in the actual build.
View attachment TPA6120A_headphone amp - CADCAM.ZIP
I’ve included a set of Gerbers for anybody who wants to have a play. You can import these into Eagle using Falk Stricker's 'Eagle PCB Power Tools' if you want to try isolation routing or you can print them from GraphiCode's GC-Prevue.
w
I have been meaning to buy a toaster oven for reflow to replace the one I used to have, but I haven’t got around to it yet. Producing solder stencils is a bit of a problem as I no longer have access to a laser cutter, I have thought about buying some spray-on ultraviolet resist and making them in brass sheet but although I’ve got the brass I still haven’t got the spray.
I still haven’t solved the issue of PTH and vias. I tried conductive paint but didn’t get far, perhaps it’s a question of technique.
I have read some people’s concerns regarding output impedance. I am therefore looking for a small 1206 inductor with which to replace the 10R resistor at the output. I have also redrawn the circuit with some global feedback, the values shown are not set in stone, and although rework is not fun on a board such as this, the new layout leaves room for some variations in the actual build.
View attachment TPA6120A_headphone amp - CADCAM.ZIP
I’ve included a set of Gerbers for anybody who wants to have a play. You can import these into Eagle using Falk Stricker's 'Eagle PCB Power Tools' if you want to try isolation routing or you can print them from GraphiCode's GC-Prevue.
w
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.