Hi Jens, great project. I'm using the same ADC & DAC with a project of my own, was just wondering a few things:
- do you bias the ADC input buffer from the ADC VREF, or just from a standard resistor divider as in the datasheet/EVM?
- do you sum the DAC differential outputs to SE and then make them differential again later (as some designers prefer due to improved CMRR) or just go differential all the way from DAC to output?
- Any experience with DC-coupled OPA1632D straight at the DAC output? Initial test results on a preceding AK4396 in a space-conscious design showed no big difference when simply DC-coupling the DAC output to a OPA1632D buffer than via 100uF caps.
- do you bias the ADC input buffer from the ADC VREF, or just from a standard resistor divider as in the datasheet/EVM?
- do you sum the DAC differential outputs to SE and then make them differential again later (as some designers prefer due to improved CMRR) or just go differential all the way from DAC to output?
- Any experience with DC-coupled OPA1632D straight at the DAC output? Initial test results on a preceding AK4396 in a space-conscious design showed no big difference when simply DC-coupling the DAC output to a OPA1632D buffer than via 100uF caps.
I have seen a big difference in THD with DC and AC coupling of 1632, working with PCM1804 and AK4394.
I tend to agree with Demian 
Some time ago I got access to a relatively well-known Audio Tester. I have attached a comparison of the generator in that one and the generator in my own Audio Analyzer.
The noise floor of my generator is relatively high. This is because the DAC is used at -20dBFS in this case, loosing some of the dynamic range. But improving the distortion a lot as well.
@PWatts
Yes, the bias circuit is a resistor divider with decoupling.
I use differential design all the way through. DC-coupled by the way.
I have not tried the OPA1632 together with the AK4490. I used it in one of the first test boards together with AK4399 and AK5394A. On the DAC side I used LME49710 directly on the DAC outputs and the OPA1632 was used as the output driver. It seemed to be too sensitive to the load on the output, so I changed it to a different design.
Some time ago I got access to a relatively well-known Audio Tester. I have attached a comparison of the generator in that one and the generator in my own Audio Analyzer.
The noise floor of my generator is relatively high. This is because the DAC is used at -20dBFS in this case, loosing some of the dynamic range. But improving the distortion a lot as well.
@PWatts
Yes, the bias circuit is a resistor divider with decoupling.
I use differential design all the way through. DC-coupled by the way.
I have not tried the OPA1632 together with the AK4490. I used it in one of the first test boards together with AK4399 and AK5394A. On the DAC side I used LME49710 directly on the DAC outputs and the OPA1632 was used as the output driver. It seemed to be too sensitive to the load on the output, so I changed it to a different design.
Very true. I still prefer it because the part is available through Digikey and the datasheet doesn't require an NDA. What a stupid practice.
Not to derail, but do you have contact info for an applications guy there? I was wondering if the "heavy load" drive mode had any downsides, but the datasheet is really short on information.
Not to derail, but do you have contact info for an applications guy there? I was wondering if the "heavy load" drive mode had any downsides, but the datasheet is really short on information.
taking way to long to launch this product!
stop fooling about, get the product done already!
at least get some prototypes out there so you can fix/debug it get feedback etc.
now with all that said I would gladly be a beta tester for you
stop fooling about, get the product done already!
at least get some prototypes out there so you can fix/debug it get feedback etc.
now with all that said I would gladly be a beta tester for you