Hi,
i use DAC9881SBRGET do my test,
Parameters are as follows
VDD：+5V
IOVDD：3.3V
VrefH-F & VrefH-S：4.096V
VrefL-F & VrefL-S：0V(GND)
GAIN：0V
RSTSEL：0V
PDN：0V
LDAC：0V
SDOSEL：3.3V
USB/BTC：3.3V
but now write code between 0h~550h,Vout only keep 8.9mV,No change
if code between 550h~262144,the Vout is normal change
i need Vout at code 0h~550h normal change
Please give me some guidance thank you!!
Circuit is as follows
Hi Hardy,
The near zero-scale and full-scale error is due to the internal amplifier of the DAC9881 device. Near zero-scale and full-scale the internal amplifier does not have enough headroom resulting in non-linear behavior, which prevents full rail-to-rail output swing. All electrical characteristics pertaining to the DAC9881 device are characterized from code 2048 to 260096, this is shown on page 3 of the DS. This is further explained and illustrated in page 26 of TIDU022, figure 29 and 30.
One solution for true full-scale use is to change VrefL to a value higher than 0V, and slightly above ~10mV offset that occurs for codes 0-550.
If you have any other questions please let me know.
Best Regards,
Matt
Hardy,
The internal amplifier obtains it's negative supply from AGND, therefore the output will more than likely saturate at ~10mV for codes 0 to 550.
Regarding your search into another DAC device, do you have any criteria for the device, such as resolution, linearity, offset error, etc?
If you do need 18 bits of resolution and the entire span, one option is to increase VrefL and decrease VrefH. This will ensure that the internal amplifier operates within the supply rails of 5V and GND. You can then add an external difference amplifier to adjust gain and offset so that the end span is 0V to 5V. When you select the external amplifier and passive components, please verify that the error sources are acceptable as they contribute to gain and offset to final output. I've provided an example below with an ideal op-amp and passives:
In the example below, VrefL was increased to 0.1V, which is in the range specified in the datasheet. Also note that for the example below VS2 is a fixed 0.2V source, and VS1 represents the DAC output, with a VrefH value of 2.6V and VrefL value of 0.1V. When choosing an op amp verify that it can reach rail to rail, or increase the headroom to both rails by increasing and decreasing the AVCC and AVSS supplies of the op amp.
If 16 bit devices are acceptable for your application, I would encourage you to look at DAC8830/8831. Additionally, unbuffered DACs, which do not include an internal buffer can also be used. An external amplifier is required for these devices, so just as suggested above, verify that the error sources specified from the amplifier datasheet is acceptable for your end application.
Best Regards,
Matt
