Functional Description (continued)
Bit Control
Word
Start Bit 1
Register 0
Select
3 MODE
4 RNGA
5 RNGB
6 RNGC
7 RNGD
8 SIA
9 SIB
10 SIC
11 SID
12 ACT
Table 1
DACA
Write
1
1
0
0
D7
D6
D5
D4
D3
D2
D1
D0
DACB
Write
1
1
DACC
Write
1
1
DACD
Write
1
1
0
1
1
1
0
1
D7
D7 D7
D6
D6 D6
D5
D5 D5
D4
D4 D4
D3
D3 D3
D2
D2 D2
D1
D1 D1
D0
D0 D0
WM5621L
Bit Power-up state
Function
MODE
0
Control serial interface
RNG A
1
DACA range select
(0 = x1, 1 = x2)
RNG B
1
DACB range select
(0 = x1, 1 = x2)
RNG C
1
DACC range select
(0 = x1, 1 = x2)
RNG D
1
DACD range select
(0 = x1, 1 = x2)
SIA
0
DACA shutdown inhibit
SIB
0
DACB shutdown inhibit
SIC
0
DACC shutdown inhibit
SID
0
DACD shutdown inhibit
ACT
0
Software shutdown control
Table 2
SIA
0
0
0
0
1
1
1
1
Table 3
ACT
0
0
1
1
0
0
1
1
HWACT
0
1
0
1
0
1
0
1
DAC status
shutdown
shutdown
shutdown
active
active
active
active
active
Control Register
The control register contains 10 active bits. The MODE bit
controls the operation of the serial interface as described
above. The function of the control register bits, and their
state on power-up, are shown in table 2.
The shutdown state of each DAC is controlled through the
shutdown inhibit bit for that channel (SIx), the ACT bit of the
control register, and the HWACT pin. Table 3 shows the
logical action of these three controlling bits for DAC A. It is
possible, for example, to have any combination of DACs
switched from shutdown to active by the HWACT pin, while
the remaining DACs are held always active (achieve this by
setting ACT=1, SIx=0 for the switching DACs, and SIx=1 for
the always active DACs).
Linearity, offset, and gain error using
single end supplies
When an amplifier is operated from a single supply, the
voltage offset can still be either positive or negative. With
a positive offset, the output voltage changes on the first
code change. With a negative offset the output voltage
may not change with the first code depending on the
magnitude of the offset voltage.
The output amplifier, with a negative voltage offset,
attempts to drive the output to a negative voltage.
However, because the most negative supply rail is GND,
the output cannot drive to a negative voltage.
So when the output offset voltage is negative, the output
voltage remains at ZERO volts until the input code value
produces a sufficient output voltage to overcome the
inherent negative offset voltage, resulting in the transfer
function shown in Figure 5.
This negative offset error, not the linearity error, produces
this breakpoint. The transfer function would have followed
the dotted line if the output buffer could drive to a negative
voltage.
Wolfson Microelectronics
11