Rainbow Electronics MAX11008 - Manual (Page 37)

V GATE_ = (2 x V REFDAC x CODE)/4096
= [2 x V REFDAC x (V SET _
+ LUTTEMP{Temp} +
LUTAPC{APC})]/4096
where:
V
GATE_
= actual gate voltage.
V SET_ = factory-set DAC code at TCAL.
LUTTEMP{Temp} = interpolated lookup value in the
TEMP table for the sampled temperature.
LUTAPC{APC} = interpolated lookup value in the APC
table for the APC parameter.
TCAL = temperature at which LUTTEMP{TCAL} returns
0; i.e., the calibration temperature.
V SET_ is a 12-bit unsigned DAC code (0 to 4095).
LUTTEMP{Temp} and LUTAPC{APC} are the result of
lookup operations and are 16-bit signed numbers in
DAC CODE units. The MAX11008 calculates the sum of
(V
SET_ + LUTTEMP{Temp} + LUTAPC{APC}) with 16-bit
signed arithmetic and limits that result to the 12-bit res-
olution of the DAC (0 to 4095) to arrive at the final out-
put DAC CODE.
The LUT values for Temp (LUTTEMP{Temp}) and APC
(LUTAPC{APC}) are the result of lookup table opera-
tions (LUT operations). The values are directly stored in
the LUT sections of the EEPROM. They are 16-bit
signed (two’s complement) quantities, but to prevent
mathematical overflow, their magnitude should be limit-
ed to 12-bit quantities (-4096 to +4095, which is the full
range of the DAC ignoring the sign).
When averaging is disabled, V GATE_ operations pro-
ceed as follows:
1) A new ADC sample is measured and compared to
the last sample used for a V GATE_ calculation.
2) The absolute difference between the two ADC mea-
surements is compared to the hysteresis setting. If
the difference is equal to or greater than the hystere-
sis setting, the new sample is used to recalculate
V
GATE_ . If the hysteresis setting is not exceeded, the
following steps are bypassed and V
GATE_
is not
recalculated.
3) The ADC sample is converted to a pointer for the
LUT. The mechanism for this is explained in the fol-
lowing section, but the process turns the 12-bit ADC
sample into an n-bit pointer.
4) The lookup operation is performed, and if required,
an interpolation between two table values is calcu-
lated. The result from the lookup table is stored as
either LUTTEMP{Temp} or LUTAPC{APC}.
5) The V GATE_ equation is now calculated and
depending on the status of the LDAC_ bit, output to
the appropriate DAC. The actual value of the DAC
output depends on the values within the LUT. It is
possible that the new value for V
GATE_ is the same
as the last value for V
GATE_
, even though the hys-
teresis in step 2 was exceeded.
If averaging is enabled for either the temperature or
APC parameter, the V
GATE_ calculation process is the
same. The difference is that the value for the ADC sam-
ple (step 1 and step 3) is replaced by an ADC average.
The MAX11008 measures 16 samples to acquire an ini-
tial average. When averaging is enabled, the first 15
samples do not trigger a new average, and a V
GATE_
calculation is not triggered. After the average is
acquired, each new ADC sample produces a new
rolling average. The rolling average is calculated with
the following equations.
In acquire mode:
Average is only valid after 16 samples.
In tracking mode:
Average = 15/16 Average + 1/16 Sample
= 15/16 Average + 1/16 (Average + Difference)
where:
Difference = Sample - Average
= Average + 1/16 Difference
= Average + 1/16 (Limited Difference)
The limited difference between the sample and the
average is a maximum value that is set by the T_LIMIT
and A_LIMIT bits, which are used to reject spurious
noise. Difference limiting may be set from 1 LSB to 64
LSBs, or may be disabled altogether.
By setting the A_AVGCTL and T_AVGCTL bits, the
average tracking formula can be altered to add 1/4 of
the difference on each calculation, rather than 1/16.
This reduces the filter’s time constant and allows the
average to track faster moving signals, and is most
suited to the APC channel. The A_AVGCTL and
T_AVGCTL bits do not alter the formula for acquiring
the initial average.
If the APC[11:0] value is used instead of an ADC sam-
ple for the APC sample, all averaging and hysteresis
functions are bypassed. The serial interface directly
controls the APC[11:0] value and triggers a V
GATE_
cal-
culation each time it is written.
Average Sample
j
= ∑
=
/16
0
15
MAX11008
Dual RF LDMOS Bias Controller with
Nonvolatile Memory
______________________________________________________________________________________ 37