Circuits and Systems, 2011, 2, 74-84
doi:10.4236/cs.2011.22012 Published Online April 2011 (http://www.SciRP.org/journal/cs)
Copyright © 2011 SciRes. CS
A 12-Bit 1-Gsample/s Nyquist Current-Steering DAC in
0.35 µm CMOS for Wireless Transmitter
Peiman Aliparast1,2, Hossein B. Bahar2, Ziaadin D. Koozehkanani2, Jafar Sobhi2, Gader Karimian2
1Young Research Club, Islamic AZAD University of Sofian, Sofian, Iran
2Faculty of Electrical and Computer Engineering, University of Tabriz, Tabriz, Iran
E-mail: p-aliparast@tabrizu.ac.ir
Received December 26, 2010; revised January 25, 2011; accepted March 4, 2011
Abstract
The present work deals with 12-bit Nyquist current-steering CMOS digital-to-analog converter (DAC) which
is an essential part in baseband section of wireless transmitter circuits. Using oversampling ratio (OSR) for
the proposed DAC leads to avoid use of an active analog reconstruction filter. The optimum segmentation
(75%) has been used to get the best DNL and reduce glitch energy. This segmentation ratio guarantees the
monotonicity. Higher performance is achieved using a new 3-D thermometer decoding method which re-
duces the area, power consumption and the number of control signals of the digital section. Using two digital
channels in parallel, helps reach 1-GSample/s frequency. Simulation results show that the spurious-
free-dynamic-range (SFDR) in Nyquist rate is better than 64 dB for sampling frequency up to 1-GSample/s.
The analog voltage supply is 3.3 V while the digital part of the chip operates with only 2.4 V. Total power
consumption in Nyquist rate measurement is 144.9 mW. The chip has been processed in a standard 0.35 µm
CMOS technology. Active area of chip is 1.37 mm2.
Keywords: Wireless Transmitter, 3-D Thermometer Decoding, Current Steering DAC, WLAN, Integrated
Circuits, CMOS
1. Introduction
The rapid improvement in the field of wireless commu-
nications and the image signal processing area requires
the designers to put an increasing amount of design effort
in the integration of digital and analog systems on a chip
(SoC). High performance DACs find applications in the
area of wireless transceivers such as Wireless Local Area
Networks (WLAN) and Wireless Metropolitan Area
Networks (WMAN), image signal processor such as
High Definition Television (HDTV), digital signal syn-
thesizers, and etc. CMOS current mode DACs are the
natural candidate for such applications Because of their
high speed, low power, and cost effectiveness [1]. No-
wadays the WLAN products are increasing in the market.
The WLAN infrastructure such as access points con-
nected to the internet exists now everywhere in homes,
offices, and public spaces such as WLAN hotspots. New
services or applications are being created by connecting
various kinds of WLAN products with the WLAN infra-
structure. Figure 1 shows the typical structure of a direct
conversion (zero-IF) transmission chain for wireless ap-
plications.
Two DACs are needed to convert the I and Q digital
modulated signals coming from the digital signal proces-
sor (DSP) into analog waveforms, which are smoothed by
the following low-pass reconstruction filters. These base-
band signals are then shifted to radio frequency (RF) by
two quadrature mixers, and summed up to obtain the final
waveform to be transmitted at the antenna, after the am-
plification prov ided by the power amplifier (PA) [2]. The
baseband sections of such telecom standard transmitters
typically consist of cascading of a digital-to-analog con-
Figure 1. General block diagram of direct conversion for
wireless transmitter chain .
I
Q
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
75
verter (DAC), receiving the digital signal processor (DSP)
bit-stream, and an analog reconstruction filter, which has
to suppress the DAC spectral images. Digital interpola-
tions filter to be situated between the DSP (which typi-
cally operates at Nyquist frequency) and the DAC, to
enhance the data-rate to the desired value. The design of
such a baseband section of wideband wireless communi-
cation systems has to optimize the trade-off between two
possible approaches: A low DAC conversion frequency,
implies a low power in terpolation filter, with demand to a
high-order, power-hungry analog reconstruction filter,
and a high DAC conversion frequency, implies a digital
filter with a high interpolation factor, that relaxes the re-
quired performance of the analog smoothing filter. This
trade-off is presently optimized with a DAC data-rate
about 8-10 times the signal bandwidth and a 4-6th order
analog reconstruction filter. For instance, in the case of
the WLAN IEEE 802.11a standard (whose signal band-
width is equal to 10 MHz), the DAC data-rate is around
100 MHz as illustrated in Figure 2 [2-4].
Due to the upcoming higher data rate standards (IEEE
802.16 and 802.11n, for instance), future implementa-
tions will involve with several critical issues on this ba-
seband section architecture. As the new standards will
present a larger signal bandwidth (25 MHz for the up-
coming IEEE 802.16, for instance [5]), the use of tradi-
tional transmission (TX) baseband architectures will re-
sult in a more and more critical design of the analog fil-
ters, since their cut-off frequency has to be increased
(with an increasing sensitivity to the lower CMOS gain
and to the non-dominant poles) [6]. Figure 3 shows this
work which exploits the DAC oversampling ratio (OSR)
to avoid the use of an active analog reconstruction filter
[2]. As a matter of fact, the DAC conversion frequency is
increased up to 1 GHz.
2. High Speed Conventional
Current-Steering DACs
2.1. Binary Weighted Architecture VS. Unary
Decoded Architecture
Current-steering DACs are based on an array of matched
current sources which are unity decoded or binary
weighted [7]. As shown in Figure 4, the reference source
is simply replicated in each branch of the DAC, and each
branch current is switched on or off based on the input
code. For the binary version, the reference current is
multiplied by a power of two, creating larger currents to
represent higher-magnitude digital signals. In the unit-
element version, each current branch produces an equal
amount of current, and thus 2N current source elements
are needed. The performance of the DAC is specified
through static parameters: Integral Non-Linearity (INL),
Differential Non-Linearity (DNL) and parametric yield;
and dynamic parameters: glitch energy, settling time and
SFDR [8]. Static performance is mainly dominated by
systematic and random errors. Systematic errors caused
by process, temperature and electrical slow variation
gradients are almost cancelled by proper layout tech-
niques [9]. Random errors are determined solely by
mismatch due to fast variation gradients.
Advantages and disadvantages of these structures are
Figure 2. Traditional baseband analog section for wireless
transmitters.
Figure 3. This work which exploits impact of the DAC conversion frequency on the filter implementation.
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
76
Figure 4. (a) Unit-element current-steering DAC; (b) Bi-
nary current-steering DAC.
summarized below:
Thermometer:
Advantages
Low glitch energy
Monotonicity
Small DNL errors
Disadvantages
Digital decoding with more area and power
consumption
Increased number of control signals
Binary:
Advantages
Low digital power consumption
Small number of control signals
Disadvantages
Monotonicity not guaranteed
Larger DNL errors
Large glitch energy
Figure 5 also summarizes aforementioned points
graphically.
2.2. Segmented DAC Structure
Usually, to leverage the clear advantages of the ther-
mometer-coded architecture and to obtain a small area
simultaneously, a compromise is found by using seg-
mentation [10]. The DAC is divided into two sub-DACs,
one for the MSBs and one for the LSBs. Thermometer
coding is used in the MSB where the accuracy is needed
mostly. Because of the reduced number of bits in this
section, the size is considerably smaller than a true ther-
mometer coded design. The LSB section can either be
done using the binary-weighted or the thermome-
ter-coded approach. We will refer to a fully bi-
nary-weighted design as 0% segmented, whereas a fully
thermometer-coded design is referred to as 100% seg-
mented. The design of current-steering DAC starts with
an architectural selection to find the optimum segmenta-
tion ratio (m over n) that minimizes the overall digital
and analog area [10-12]. The INL is independent of the
segmentation ratio and depends only on the mis-
match if the output impedance is made large en ough [7].
Figure 5. Binary weighted versus Unary-decoded.
The DNL speciation depends on the segmentation ratio
but it is always satisfied provided that the INL is below
0.5 LSB for reasonable segmentation ratios. The glitch
energy is determined by the number of binary bits b, be-
ing the optimum architecture in th is sense a totally unary
DAC. However, this is unfeasible in practice due to the
large area and delay that the thermometer decoder would
exhibit. The minimization of the glitch energy is then
done in circuit level design and layout of the switch and
latch array and current source cell [13].The optimum
segmentation is workout 75% in [10,12] so we have used
this segmentation to achieve the best performance in
high-speed design. Thus we consider 9-bit as thermome-
ter-coded and 3-bit as binary-weighted. Figure 6 shows a
typical block diagram of an n-bit segmented cur-
rent-steering DAC which uses the advantages of both
architectures. Input word is segmented between b less
significant bits that switch a binary weighted array and
m= n – b most significant bits that control switching of a
unary current source array. The m input bits are ther-
mometer decoded to switch individually each of the un-
ary sources [14-16]. A dummy decoder is placed in the
binary weighted input path to equalize the delay. A latch
is placed just before the switch transistors of each current
source to minimize any timing error [10].
3. New Thermometer Decoding Arch it ec t ur e
Figure 7 shows a block diagram of a conventional row
and column decoded 12-b it current-steer ing DAC. In this
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
77
Figure 6. A typical segmented current-steering DAC archi-
tecture.
Figure 7. Block diagram of a conventional row and column
decoded 12-bit current-steering DAC.
block diagram, the lower significant bits are applied to a
dummy decoder [17]. This decoder creates a delay pro-
portional to the Binary-to-Thermometer decoder and
causes the signal to arrive at the switches synchro nously.
The f ive L SB b its are colu mn de coded an d the f our MSB
bits are row decoded. Column decoder is a 5-input
31-output Binary-to-Thermometer Decoder and row de-
coder is a 4-input 15-output Binary-to-Thermometer
Decoder. Outputs of the decoders control 511 current
cells in the main matrix. But if we think about Bi-
nary-to-Thermometer Decoder structure we understand
that β-bit increase of the input of the decoder cause the
area, complexity, number of control signal and power
consumption of the decoder increase with 2β. In fact
power and area are doubled with only one bit increase in
the input of the decoder and we can write:

4to15BTD 23to7BTDP P (1)

4to15 BTD23to7BTDA A (2)
Thus:

5to 31 BTD43to7 BTDP P (3)
 
5to31BTD43to7 BTDA A  (4)
where BTD is Binary-to-Thermometer Decoder, P is
the power consumption of the decoder and
A
is active
area that the decoder uses. Now consider Figure 8 that
shows a 3D decoding architecture. In this block diagram
three BTD have been used. Three bits for height, three
bits for row and three bits for column and every cell is
selected with 3 parameters (R, C and H). In fact we have
only used three (3to7 BTD) instead of two (5to31 BTD)
and (4to15 BTD) thus power consumption and area of
the circuit have been improved two times because:
 
4to15 BTD23to7 BTDP P
+
 
5to31BTD 43to7BTDP P
 
4to15BTD5to31 BTD63to7BTDP P P
(5)
And for area we have:
 
4to15BTD 23to7BTDA A
+
 
5to31BTD43to7BTDA A 

4to15BTD5to31 BTD63to7BTDA A A
(6)
In this structure 3 LSB bits are column decoded, 3
middle bits are row decoded and 3 MSB bits are height
decoded. On the other hand, we have only used 21 con-
trol signals instead of 46 control signals thus the number
of control signals has been decreased by 55 percent
hence we can achieve the best speed and performance.
4. The Current Cell, Latch and Driver
Static and dynamic performance of current-steering
Figure 8. Block diagram of a novel method row and column
and height decoded 12-bit 3-D DAC.
I
out+
I
out
I
out+
I
out
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
78
DACs is mostly determined by the accuracy of the cur-
rent sources, finite output impedance, and switching time.
Figure 9 shows a current source transistor MCS, an addi-
tional cascode transistor MCAS that increases the output
impedance and two complementary switch transistors
MSW. This figure shows cascode current source and
switch structure for 1LSB while for realizing unary cur-
rent source cell (8LSB) we used same structure with 8
parallel transistors. In proposed 12-bit DAC three bits are
binary weighted so it uses the current source of Figure 9
and remaining 9 bits are thermometer decoded and need
unary current sources. Since two D/A converters pro-
cessed in the same technology do not necessarily have
the same specifications due to technological variations,
therefore it is of the utmost importance to know th e rela-
tionship that exists between the specifications of the cir-
cuit and the matching properties of used technology. For
a current-steering D/A converter, the INL is mainly de-
termined by the matching behavior of the current sources.
A parameter that is well suited for expressing this tech-
nology versus DAC specification relation is the INL
yield [16]. This INL yield is defined as the ratio of the
number of D/A converters with an INL smaller than 1
LSB to the total number of tested D/A converters. As
defined by Pelgrom, mismatch “is the process that causes
time- independent random variations in physical quanti-
ties of identically designed devices” [18]. Pelgrom’s pa-
per has become the de facto standard for analysis of tran-
sistor matching, and thus his formula for the standard
deviation of saturation current for two identically sized
devices has been used for the design. This formula is:
 


222
222
4T
GS T
IV
IVV


(7)
where
Figure 9. Current source cell topology.

2
222
VT
TVT
A
VSD
WL
 (8)
and
2
222
2
ASD
WL

 (9)
Most of these variables are process-dependent con-
stants. Using these results, an equation for the minimum
size device that still provides a reasonable current stan-
dard deviation can be determined [13]:

2
22
22
4
1
2
VT
cs minGS T
A
IA
WL
IVV





(10)
where Aβ, AVT , VGS and VT are process parameters, while
I is the current generated by a given source and σI is the
relative standard deviation of one current source. The
same aspect ratio can be obtained for different areas
W×L, except for the MCS transistor, because the usual
INL-mismatch specification eliminates one degree of
freedom. The relative standard deviation of a unit current
source σI/I has to be small enough to fulfil the INL< 0.5
LSB specification given a parametric yield [17]:
2
21
05 2
2
upper bound
N
INL
I
yield
Iinv _ normal..



(11)
where inv_normal is the inverse cumulative normal dis-
tribution. The MCS transistor size is found by:
22
2
22 4
4VT
nox
AA
I
WVV
I
CI









(12)
2222
24
4
nox
VT
C
LA.VA
I
I. I




(13)
where µnCOX is the MOS transistor gain factor and ΔV =
(VGS VT ). Applying Equations (12) and (13) we arrived
in Wmin = 4 µm and Lmin = 5 µm for the current source.
But in design of cascode current sources, to achieve high
speed, we need to choose the size of cascode transistor as
small as possible. With different size for WCAS and WCS,
we have to use contact in node Y (Figure 9) which in-
creases parasitic capacitance and decreases the speed . So
in a trade off, we decided to decrease WCS as small as
possible and use the same size with WCAS. In other words
we have chosen WCAS = WCS = 2 µm, and avoid using
contact in node Y. To compensate for reduction in WCS
in Equation (14), we increase the values of LCS (LCS =
10 µm) and ΔV while we keep LCAS at its minimum size
0.35 µm. Thus we do not use these Equations (12) and
(13) and use only mismatch Equation (10) to reach a
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
79
minimum sizing of current cell. With this method, the
speed of switch is high also INL < 0.5 LSB is satisfied.
The small-signal output impedance for the current source
topology of Figure 9 is given by:
outSWCAS dsSWdsCAS dsCS
Rgm .gm.r.r.r (14)
The optimum MSW and MCAS gate bias voltages con-
cerning the output impedance are found by differentiat-
ing Rout with respect to VgSW and VgCAS. For the SW and
CAS gate bias voltages that maximize output impedance
are found as:

122
3
gCASTo minCSCASSW
VVVVV V

  (15)

12
3
gSWTo minCSCASSW
VVVVV V
 
  (16)
Figure 10 shows the biasing scheme for the cascoded
current sources. The PMOS sections of the biasing cir-
cuits are labeled as Global biasing while the NMOS
sections are labeled as Local biasing. In the actual im-
plementation, the global biasing is realized using a
common-centroid layout to reduce effects of gradients.
The local biasing is separated into four quadrants. There
is no direct connection between any two quadrants. This
will improve both DNL as well as INL performance [10].
A driver circuit with a reduced swing placed between the
latch and the switch reduces the clock feed-through to
the output node as well [19,20]. Figure 11(a) shows a
current source, switch, latch and driver cell. A new
swing-reduced-device (SRD) circuit is designed (shown
in Figure 11(b) ). The latch circu it co mplemen tary outpu t
levels and non-symmetrical cross point are designed to
minimize glitches [13]. The waveforms of the different
nodes are shown in Figure 11(c) without SRD circuit
and Figure 11(d) with SRD circuit. Signals with sym-
metrical crossing point are fed from the left and SRD
makes a non-symmetrical crossing point which reduces
the spike at node VX considerably. In SRD circuit, MSRD1
is always on and when MSRD2 is off, VgSW approaches
2.4 V (power supply value of digital part). When MSRD2
is on with proper sizing of MSRD2, VgSW can be set to de-
sired value because VgSW in this case will be equal to VSG
of MSRD2 transistors. In this circuit for complete switch-
ing of MSW transistors we need 350 mV differential volt-
age, so VSG of MSRD2 is set to 2.05 V. On the other hand
for non-symmetric crossing it’s enough to choose bigger
size for MSRD1 than MSRD2. Size of MSRD1 and MSRD2 has
been given in Table 1, also SRD output wave forms and
its effect in reducing spike in node VX is shown in Figure
11(d). The capacitive coupling to the analog output is
minimized by limiting the amplitude of the control sig-
nals just high enough to switch the tail current com-
pletely to the desired output branch of the differential
Figure 10. Biasing scheme for current sources.
pair. In addition the switch transistors are kept relatively
small in order to avoid large parasitic capacitances.
Table 1. Current source and SRD transistors dimensions
and currents.
Transistor Size ID
MCS W = 2 µm, L = 10 µm 5 µA
MCAS W = 2 µm, L = 0.35 µm 5 µA
MSW W = 0.5 µm, L = 0.35 µm -
MSRD1 W = 1.5 µm, L = 2 µm -
MSRD2 W = 1 µm, L = 2 µm -
5. Layout and a Few Techniques to Achieve
High Speed
Clock distribution for 1 GHz is very difficult and getting
data in this speed is very hard thus we have used 2 chan-
nels for digital section. Every channel works at 500MHz
and then results of two channels are combined at the in-
put of the switch to get 1 GHz. Figure 12 shows the
structure used for digital section of the DAC. Channel 1
samples input data with clock and channel 2 samples
input data with clock-not. A buffer just before switch
combines the output of two digital channels. It sends the
output of digital channel 1 with clock and the output of
digital channel 2 with clock-no t to the input of switch. In
fact in one period of clock we take 2 samples of the input
code and at the output it seems that the circuit works at
1 GHz. On the other hand, we use master-slave operation
in all digital circuits and use pipelining scheme, so in
overall the digital circuit only senses one gate-delay. For
example the structure of one of 3-input 7-output Bi-
nary-to-Thermometer Decoder has been shown in Figure
13. Layout of all digital section has been done manually
to guarantee the best speed, low power and minimum
area. Figure 14 shows the complete layout of the DAC,
latches and switches which are grouped in a separated
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
80
(a) (b)
(c)
(d)
Figure 11. Non-symmetrical crossing point reduces current source drain spike and clock feed-through scheme, (a) current
source, switch, latch and driver cell, (b) SRD circuit, (c) drain spike and driver voltages without SRD circuit, and (d) drain
spike and driver voltages with SRD circuit.
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
81
Figure 12. Using two 500 MHz digital channels to achieve
1 Gsample/s.
Figure 13. Gate level structure of 3-input 7-output Binary-
to-Thermometer Decoder.
array placed between the decoders and the current source
arrays to isolate these noisy digital circuits fro m the sen-
sitive analog circuits that generate the current. A guard
ring has been used to separate analog section from digital
section. Layout of the decoder circuit has been drawn
manually and pipelining used to reach the maximum
speed and improvement of the parasitic capacitance and
sizing of transistors has been done with simulation. For
reduced systematic errors each unary current source is
divided into 16 sub-current sources and Q2 Random
Walk distribution scheme is applied [21].
6. Simulation Results
Simulations have been performed on a differential 50-
load. The internal node interconnection capacitance has
been estimated to be 400 fF, and the output capacitance
1pF. The analog vo ltage supply is 3.3 V while the digital
part of the chip operates at only 2.4 V. Total power con-
sumption in the Nyquist rate measurement is 144.9 mW.
SFDR is better than 64 dB in Nyquist rate. Figure 15
shows differential output spectrum where DAC worked
with 1 GSample/s speed and input code near to Nyquist
rate (495 MHz) with 1 mV (rms) noise voltage on analog
Figure 14. Layout of designed 12-bit DAC.
power supply. Also Figures 16 and 17 show differential
outputs spectrum for 1 GSample/s speed with input sig-
nals in 100 MHz and 25 MHz respectively. Measured
SFDR for both of them was better than 70 dB. Figure 18
shows the measured SFDR versus various input fre-
quency for the proposed DAC at a 1 GHz sampling fre-
quency. In Figure 19, a dual-tone SFDR measurement is
shown. Two sinusoidal signals around 15 MHz with
5-MHz spacing have been applied to the D/A converter
at an update rate of 1 GSample/s. The SFDR equals 71 dB.
In order to make simulation of glitch energy transitio n of
input digital codes from 01111111 1111 has been made to
100000000000, such that the glitch energy has been ob-
tained to be 2 .3 pV.s. Figures 20 and 21 show DNL and
INL characteristics of designed DAC for in creasing
Figure 15. Sinewave spectrum for Fs = 1 GSample/s, Fsig =
495 MHz.
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
82
Figure 16. Sinewave spectrum for Fs = 1 GSample/s, Fsig =
100 MHz.
Figure 17. Sinewave spectrum for Fs = 1 GSample/s, Fsig =
25 MHz.
input code from 0 to 4096. The INL and DNL obtained
from post layout four corners Monte-Carlo simulations
considering process mismatch parameters are better
than 0.74 LSB and 0.49 LSB, respectively. Table 2
Table 2. Performance summary.
Technology 0.35 µm (1P4M) TSMC Mixed Mode
CMOS
Resolution 12-bit
Update rate Up to 1 GS/s
Max. output swing 2Vpp diff.
DNL Better than 0.49LSB
INL Better than 0.74LSB
SFDR (495 MHz@1 GS/s) 64 dB
SFDR (100 MHz@1 GS/s) 70 dB
SFDR (25 MHz@1 GS/s) 71 dB
SFDR (1 MHz@1 GS/s) 72 dB
ENOB (25 MHz@1 GS/s) 10.7-bit
ENOB (1 MHz@1 GS/s) 11-bit
Analog Power consumption
(at 1 GS/s) 69.3 mW (21 mA from 3.3 V)
Digital Power consumption
(at 1 GS/s) 75.6 mW (31.5 mA from 2.4 V)
Total Power consumption (at
1 GS/s) 144.9 mW
Analog/Digital voltage
supply 3.3 V/2.4 V
Active area 1 306 µm×1052 µm
Figure 18. SFDR versus input frequency for the proposed
DAC at 1 GHz sampling frequency.
Figure 19. Simulated dual-tone spectrum for Fs = 1 GSample/s,
Fsig = 20 MHz and 10MHz.
Figure 20. DAC INL characteristic.
summarizes some of important performance parameters
of the DAC.
7. Conclusion
In this article a 3.3 V, 12-bit, current-steering, 9 + 3 seg-
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
83
Figure 21. DAC DNL characteristic.
mented architecture digital to analog converter for
base-band of wireless transmitter circuits has been pre-
sented. A new 3-D thermometer decoding scheme has
been used in digital section which reduces the area power
consumption and number of control signals considerably.
Simulations have been performed to analyze and solve
some of important dynamic linearity limitations. Using
two digital channels in parallel, one operating with clock
and the other operating with clock-not for the sampling
rate of 1 GS/s while each channel operates only at 500 MHz.
This clocking strategy makes clock distribution much
easier. Analog switches and SRD circuits have been op-
timized not only to get minimum area and maximum
speed but also to improve dynamic behavior of the DAC.
Segmentation (75%) decreases DNL error and glitch
energy considerably and guarantees needed improvement
of SFDR. Separate power supplies have been used for
digital and analog parts. Digital section operates at lower
supply voltage than an alog p art. This increases speed and
reduces power consumption of the digital part and at the
same time decreases power supply noise and improve the
performance of the analog part. The technology used is a
0.35 µm, single-poly four-metal, 3.3 V, standard TSMC
Mixed Mode CMOS process. The active area of the
DAC, as shown in Figure 14, is 1052 µm × 1306 µm.
8. References
[1] S. M. Ha, T. K. Nam and K. S. Yoon, “An I/Q Channel
12-bit 120 Ms/s CMOS DAC with Three Stage Ther-
mometer Decoders for WLAN,” Proceedings of the IEEE
Asia Pacific Conference on Circuits and Systems, Singa-
pore, 4-7 December 2006, pp. 355-358.
doi:10.1109/APCCAS.2006.342443
[2] N. Ghittori, et al., “1.2-V Low-Power Multi-Mode
Dac+Filter Blocks for Reconfigurable (WLAN/UMTS,
WLAN/Bluetooth) Transmitters,” IEEE Journal of Solid-
State Circuits, Vol. 41, No. 9, 2006, pp. 1970-1982.
doi:10.1109/JSSC.2006.880602
[3] S. Khorram, et al., “A Fully Integrated SOC for 802.11 b
in 0.18 µm CMOS,” IEEE Journal of Solid-State Circuits,
Vol. 40, No. 12, 2005, pp. 2492-2501.
doi:10.1109/JSSC.2005.857419
[4] S. Mehta, et al., “An 802.11 g WLAN SOC,” IEEE
Journal of Solid-State Circuits, Vol. 40, No. 12, 2005, pp.
2483-2491. doi:10.1109/JSSC.2005.857418
[5] C. Eklund, R. Marks, K. Stanwood and S. Wang, “IEEE
Standard 802.16: A Technical Overview of the Wireless
Man Air Interface for Broadband Wireless Ac ce ss,” IEEE
Communications Magazine, Vol. 40, No. 6, 2002, pp.
98-107. doi:10.1109/MCOM.2002.1007415
[6] N. Ghittori, et al., “An IEEE 802.11 and 802.16 WLAN
Wireless Transmitter Baseband Architecture with a 1.2-V,
600-Ms/s, 2.4-mW DAC,” Analog Integrated Circuits
and Signal Processing, Vol. 59, No. 3, 2009, pp. 231-242.
doi:10.1007/s10470-008-9262-x
[7] B. Razavi, “Principles of Data Conversion Systems,”
Wiley-IEEE Press, New Jersey, 1995.
[8] P. Hendriks, “Specifying Communication DACs,” IEEE
Spectrum, Vol. 34, No. 7, 1997, pp. 58-69.
doi:10.1109/MSPEC.1997.609817
[9] Y. Cong and R. Geiger, “Switching Sequence Optimiza-
tion for Gradient Error Compensation in Thermome-
ter-Decoded DAC Arrays,” IEEE Transaction on Circuits
and Systems-II, Vol. 47, No. 7, 2000, pp. 585-595.
doi:10.1109/82.850417
[10] C. Lin.and K. Bult, “A 10-bit, 500-Ms/s CMOS DAC in
0.6 mm2,” IEEE Journal of Solid-State Circuits, Vol. 33,
No. 12, 1998, pp. 1948-1958. doi:10.1109/4.735535
[11] J. Vandenbussche, et al., “Systematic de Sign of
High-Accuracy Current-Steering D/A Converter Macro
Cells for Integrated VLSI Systems,” IEEE Transaction
on Circuits and Systems II: Analog and Digital Signal
Processing, Vol. 48, No. 3, 2001, pp. 300-309.
doi:10.1109/82.924073
[12] J. Gonzalez and E. Alarcon, “Clock-Jitter Induced Distor-
tion in High-Speed CMOS Switched-Current Segmented
Digital to Analog Converters,” International Symposium
on Circuits and Systems (ISCAS’01), Sydney, 6-9 May
2001, pp. 1512-1515.
[13] J. Bastos, M. Steyaert and W. Sansen, “A High Yield
12-Bit 250-Ms/s CMOS D/A Converter,” IEEE Custom
Integrated Circuits Conference (CICC), San Diego, 5-8
May 1996, pp. 431-434.
[14] L. Sumanen, M. Waltari and K. Halonen, “A 10-Bit
High-Speed Low-Power CMOS D/A Converter in 0.2 mm2,”
IEEE International Conference on Electronics, Circuits
and Systems, Lisboan, 7-10 September 1998, pp. 15-18.
[15] Y. Nakamura, T. Miki, A. Maeda, H. Kondoh and N.
Yazwa, “A 10-b 70-Ms/s CMOS D/A Converter,” IEEE
Journal of Solid-State Circuits, Vol. 26, No. 4, 1991, pp.
637-642. doi:10.1109/4.75066
[16] M. Albiol, J. Gonzalez and E. Alarcon, “Mismatch and
Dynamic Modeling of Current Sources in Current-Steer-
ing CMOS D/A Converters: An Extended Design Proce-
dure,” IEEE Transactions on Circuits and Systems I:
Regular Papers, Vol. 51, No. 1, 2004, pp. 159-169.
doi:10.1109/TCSI.2003.821287
P. ALIPARAST ET AL.
Copyright © 2011 SciRes. CS
84
[17] A. Bosch, M. Borremans, M. Steyaert and W. Sansen, “A
10-Bit 1-Gs/s Nyquist Current Steering CMOS D/A
Converter,” IEEE Journal of Solid-State Circuits, Vol. 36,
No. 3, 2001, pp. 315-324. doi:10.1109/4.910469
[18] M. Pelgrom, A. Duinmaijer and A. Welbers, “Matching
Properties of MOS Transistors,” IEEE Journal of Sol-
id-State Circuits, Vol. 24, No. 5, 1989, pp. 1433-1440.
doi:10.1109/JSSC.1989.572629
[19] H. Kohno, et al., “A 350-Ms/s 3.3V 8-Bit CMOS D/A
Converter Using a Delayed Driving Scheme,” IEEE Cus-
tom Integrated Circuits Conference (CICC), Santa Clara,
1-4 May 1995, pp. 211-214.
[20] L. Luh, J. Choma, J. Draper, “ A High-Speed Fully Dif-
ferential Current Switch,” IEEE Transactions on Circuits
and Systems-II: Analog and Digital Signal Processing,
Vol. 47, No. 4, 2000, pp. 358-363.
doi:10.1109/82.839672
[21] G. Van der Plas, et al., “A 14-Bit Intrinsic Accuracy Q2
Random Walk CMOS DAC,” IEEE Journal of Solid-
State Circuits, Vol. 34, No. 12, 1999, pp. 1708-1718.
doi:10.1109/4.808896