While writing a program to play with the THD compensation registers on my ES9038pro, I read the datasheet for other settings that would be nice to be able to change, so I could do it in an all-in one program.
That is when I noticed that in the latest online datasheet from ESS ES9038pro, now v3.7, the description of the registers Register 61-57 had changed from “Reserved” to “DAC Scaling”. Not (yet?) in the ES9028pro datasheet, though.
The registers are described as A 12-bit unsigned value that scales the input to the 2/3/4nd DAC from the 1st DAC in each of the 8 DAC channels. Each channel contains 4 DAC’s summed together.
Googling and searching forums, I couldn’t find any useful information on the workings of these registers. In their paper on the Hyperstream architecture, it is mentioned that each DAC is actually a balanced pair of the DACs. Maybe it has some relation to this?
Strangely, the default values are 0, which would indicate no signal is sent to DACs 2-4? To make things more confusing, the datasheet states that ‘A value of 12’4095 (0xFFF) will set the input at full scale, maximizing the signal to noise ratio’. But the default is 0?? Why??🤔
I cannot imagine, however, that the default value is non-optimal, but maybe there are other trade-offs, such as THD?
I don’t want to tinker with this setting without knowing what is does. Does anybody know more about this mysterious matter?
That is when I noticed that in the latest online datasheet from ESS ES9038pro, now v3.7, the description of the registers Register 61-57 had changed from “Reserved” to “DAC Scaling”. Not (yet?) in the ES9028pro datasheet, though.
The registers are described as A 12-bit unsigned value that scales the input to the 2/3/4nd DAC from the 1st DAC in each of the 8 DAC channels. Each channel contains 4 DAC’s summed together.
Googling and searching forums, I couldn’t find any useful information on the workings of these registers. In their paper on the Hyperstream architecture, it is mentioned that each DAC is actually a balanced pair of the DACs. Maybe it has some relation to this?
Strangely, the default values are 0, which would indicate no signal is sent to DACs 2-4? To make things more confusing, the datasheet states that ‘A value of 12’4095 (0xFFF) will set the input at full scale, maximizing the signal to noise ratio’. But the default is 0?? Why??🤔
I cannot imagine, however, that the default value is non-optimal, but maybe there are other trade-offs, such as THD?
I don’t want to tinker with this setting without knowing what is does. Does anybody know more about this mysterious matter?
That each DAC is a balanced pair of DACs probably just refers to the fact that each channel had an inverting and non-inverting output.
OTOH, ES90x8PRO chips have a feature such that when they are operated in stereo mode (two sets of four channels in parallel), the interpolation filter can be reconfigured to twice the standard number of taps, but it requires custom coefficients be loaded. Also, ESS dacs have scaling feature so that if using dacs in some configuration where gain matching is needed (say, perhaps, dual mono) then the DAC chip with the higher output can be attenuated with an internal register to match the lower output DAC chip. IIUC that's part if not all of what the internal ADC is for, measuring the actual output level. Therefore, maybe as another feature of operating the DAC chip in stereo mode, there is a more sophisticated level scaling system? Just a wild guess, nothing more.
OTOH, ES90x8PRO chips have a feature such that when they are operated in stereo mode (two sets of four channels in parallel), the interpolation filter can be reconfigured to twice the standard number of taps, but it requires custom coefficients be loaded. Also, ESS dacs have scaling feature so that if using dacs in some configuration where gain matching is needed (say, perhaps, dual mono) then the DAC chip with the higher output can be attenuated with an internal register to match the lower output DAC chip. IIUC that's part if not all of what the internal ADC is for, measuring the actual output level. Therefore, maybe as another feature of operating the DAC chip in stereo mode, there is a more sophisticated level scaling system? Just a wild guess, nothing more.
Thanks @Markw4 for your input! The gain matching is supposed to be done via the master trim, ESS claiming that there will only be channel output variations between different chip dyes.
I was thinking of a balanced pair as being 2 pairs, but I guess you are right. IIRC the ES9038pro has 4 of those pairs, giving it the higher current output compared to the ES9028.
Maybe the registers somehow control this? That would also explain why they are not mentioned in the latter datasheet.
I was thinking of a balanced pair as being 2 pairs, but I guess you are right. IIRC the ES9038pro has 4 of those pairs, giving it the higher current output compared to the ES9028.
Maybe the registers somehow control this? That would also explain why they are not mentioned in the latter datasheet.
For clarity sake in case anyone is reading along, ES9038PRO has 8-channels, with each channel having two dacs, one for the inverting output, an the other dac for the non-inverting output. Thus there are a total of 16-dacs inside the chip.
