revisiting the power-efficiency trade-off on a dc voltage...
TRANSCRIPT
Revisiting the Power-Efficiency Trade-Offon a DC Voltage Source
Arturo Fajardo Jaimes 1,2, Fernando Rangel de Sousa 1
1- Radiofrequency Laboratory| Department of Electrical and Electronics Engineering | UFSC | Florianpolis, Brazil
2- Department of Electronics Engineering | Pontifical Xavierian University (PUJ) | Bogota, Colombia
Table of Content RFLaboratory
IntroductionPower-efficiency trade-offThis work
Efficiency in a voltage source
Power-efficiency trade-offPETO for lossless voltage source (Rss = 0,Rsp =∞PETO for voltage source w/o loss of self consumption(Rss 6= 0,Rsp =∞)PETO for general voltage source (Rss 6= 0,Rsp 6=∞)
Experimental verification
Conclusions
References
2UNIVERSIDADE FEDERALDE SANTA CATARINA
Power-efficiency trade-off I RFLaboratory
Energy conversion processes are characterizedby a trade-off between output power and effi-ciency.
Earliest work about electric-mechanic conversion (1978)Edison and his chief assistants re-searched the design techniques for DCelectric generators, and determinedthat a generator with smaller internalresistance than its load is more effi-cient than a generator with internal re-sistance equal to its load [1].
3UNIVERSIDADE FEDERALDE SANTA CATARINA
Power-efficiency trade-off II RFLaboratory
Nanoscale energy conversion Molecular motors (2016)At the nanoscale new possibilities and phe-nomena arise. For example, chemical energycan be converted directly into useful work(molecular motors). This diagram shows thecalculated output power versus efficiency ofa molecular motor (dashed) near equilibriumand (solid) far from equilibrium for (blue) per-fect coupling and (red) weak loss processes[2].
4UNIVERSIDADE FEDERALDE SANTA CATARINA
Power-efficiency trade-off III RFLaboratory
PETO and Portable Electronic devices
• Most of the electronic devices must beportable and must be connected to theInternet [3]
• Typically, the device portability is associ-ated to the energy technologies that hasbeen adopted [4].
• The designer of PEDs must face a trade-off between functionality (i.e. dissipatedpower) and portability (i.e. battery run-ning time, weight, size) [5].
5UNIVERSIDADE FEDERALDE SANTA CATARINA
About this work RFLaboratory
The problemThe designer of PEDs need to take into account the PETO in orderto achieve the required specifications and, as a result, an optimumoperating point (i.e. optimum load) could be found and then could beused as a constraint in the electronic design of the system.
The Proposed SolutionIn this work an analytic expression is developed for quantifying thePETO on a DC Voltage Source with and without limited energy.
ηsc =1− pL
ξ
1 + ξk·pL
6UNIVERSIDADE FEDERALDE SANTA CATARINA
Efficiency in a voltagesource I
RFLaboratory
Modeling the DC sourceA common technique for modeling a real voltage source is theuse of the Norton or Thevenin equivalent circuit as its model.
VthRth
RL
RthRL
Vth/Rth
ModelsRS
RL
Real
DC
Voltage
Source
Rth<<RL
Rth>>RL
Vth
RL
RL
Vth/Rth
7UNIVERSIDADE FEDERALDE SANTA CATARINA
Efficiency in a voltagesource II
RFLaboratory
ObservationThe equivalent circuit predicts only the current and voltage per-formance of the source, from the load point of view.
Efficiency values predicted by both models
Without knowledge about thephysical laws that govern the in-ternal source behaviors, we can-not predict accurately its effi-ciency.
8UNIVERSIDADE FEDERALDE SANTA CATARINA
PETO for voltage source RFLaboratory
The PETO is discussed for any energy source that can be modeledusing the circuits shown:
With infinite energy
V RssRsp RLvL
iL
vs
is
DC source
With finite energy
V
QnRss
Rsp RLvL
iL
vs
isDC source
9UNIVERSIDADE FEDERALDE SANTA CATARINA
Ideal battery model RFLaboratory
In [6] was proposed a battery model based on the circuit shown:
Symbol
V
Qnsv
is
Circuit model
C=Qn
SoE
-isVz=1
Vf=0
0
0 0
V SoE
SoEsv>ì
= í=î
is
10UNIVERSIDADE FEDERALDE SANTA CATARINA
PETO Rss = 0,Rsp =∞ RFLaboratory
Efficiency and power expressions
For this circuit the efficiency is aload independent variable, there-fore PETO does not occur.
PLa =V 2
RL; (1)
ηsa = 1. (2)
(a) Load power and efficiency vs load (b) Efficiency vs power
11UNIVERSIDADE FEDERALDE SANTA CATARINA
PETO Rss 6= 0,Rsp =∞ I RFLaboratory
Efficiency and power expressions
PLb= 4Pavs
RLRss(
1 + RLRss
)2; (3)
ηsb =ηsa
1 + RssRL
; (4)
Pavs =V 2
4 · Rss; (5)
pL =PLb
Pavs
=4 · RL
Rss(1 + RL
Rss
)2. (6)
(a) Load power and efficiency vs load (b) Efficiency vs power
12UNIVERSIDADE FEDERALDE SANTA CATARINA
PETO Rss 6= 0,Rsp =∞ II RFLaboratory
Quantifying the PETO
• The better compromise be-tween efficiency and poweroccurs in the load rangewhere the PETO exists.
• The limits are: 100% of ηand 100% of pL.
• We cannot achieve theselimits simultaneously. (pL =0, η = 1) or (pL = 1, η = 0.5)
• A good compromise of thisPETO is pL = 0.75, η =0.75.
In all the operating points of thisPETO, the relationship betweenthe pL and the η is described by:
ηsb = 1− pLξ
;
where, ξ=2(1 +√
1− pL), and
RL ≥ Rss.
13UNIVERSIDADE FEDERALDE SANTA CATARINA
PETO Rss 6= 0,Rsp 6=∞ I RFLaboratory
Efficiency and power expressions
PLc = PLb= 4Pavs
RLRss(
1 + RLRss
)2; (7)
ηsc =ηsb(
RssRsp
(RLRss + 1
)+ 1
) ; (8)
Pavs =V 2
4 · Rss; (9)
pL =PLb
Pavs
=4 · RL
Rss(1 + RL
Rss
)2. (10)
(a) Load power and efficiency vs load (b) Efficiency vs power
14UNIVERSIDADE FEDERALDE SANTA CATARINA
PETO Rss 6= 0,Rsp 6=∞ II RFLaboratory
Quantifying the PETO
• The PETO range is limited bythe maximum η and minimumpL point and the minimum ηand maximum pL point.
• Defining the source quality fac-tor as k = Rsp
Rss :
maxηsc
=
k√k + 1(
1 + k +√k + 1
) (k +√
k + 1)
max pL = 1
• The extreme Load val-ues are:
RL = RLp = Rss
RL = RLη = Rss√k + 1
• RLp is the load valuethat maximizes pL. RLηis the load value thatmaximizes η
15UNIVERSIDADE FEDERALDE SANTA CATARINA
PETO Rss 6= 0,Rsp 6=∞ III RFLaboratory
Quantifying the PETOWhen RL is in the PETO range (RLp ≤ RL ≤ RLη), the load powercan increase only if the efficiency decrease and vice versa. In all theoperating points of this PETO, the relationship between the pL and theη is described by:
ηsc =1− pL
ξ
1 + ξk ·pL
where, Rss√k + 1 ≥ RL ≥ Rss and ξ=2
(1 +√
1− pL). Having a high
k value means that the power supply is good, i.e. both the self con-sumption and the internal conduction losses are low.
16UNIVERSIDADE FEDERALDE SANTA CATARINA
Experimental verification RFLaboratory
Experimental Setup
RL
Real DC
voltage
source
vL
iL
• The implemented system is a demon-strator of the validity of the analytic func-tion rather than targeting a specific ap-plication.
• The real DC source (k = 90) was emu-lated using a power supply (1 V < V <10 V) and two discrete resistors (Rsp =100 kΩ, Rss = 1.1 kΩ), the electric loadwas an adjustable resistor (100mΩ <RL < 10 MΩ).
• Using this setup, both the resistive loadand the input voltage were swept.
17UNIVERSIDADE FEDERALDE SANTA CATARINA
Experimental andtheoretic results I
RFLaboratory
!"#
!"$
!"%
!"&
!"'
!
!(
!('
!(&
!(%
!($
!(#
!()
!(*
!(+
,-./01
2334546758
.
96:;.<47=! <..>=+!(#* +
?@6-.<47=! <..>=+!(+!+
96:;.<47=!'<..>=+!(#* +
?@6-.<47=!'<..>=+!(+!+
96:;.<47=!&<..>=+!(#* +
?@6-.<47=!&<..>=+!(+!+
96:;.<47=!%<..>=+!(#* +
?@6-.<47=!%<..>=+!(+!+
96:;.<47=!$<..>=+!(#* +
?@6-.<47=!$<..>=+!(+!+
96:;.<47=!#<..>=+!(#* +
?@6-.<47=!#<..>=+!(+!+
96:;.<47=!)<..>=+!(#* +
?@6-.<47=!)<..>=+!(+!+
96:;.<47=!*<..>=+!(#* +
?@6-.<47=!*<..>=+!(+!+
96:;.<47=!+<..>=+!(#* +
?@6-.<47=!+<..>=+!(+!+
96:;.<47= !<..>=+!(#* +
?@6-.<47= !<..>=+!(+!+
(a) Efficiency vs load power
18UNIVERSIDADE FEDERALDE SANTA CATARINA
Experimental andtheoretic results II
RFLaboratory
!" !# !$ !% &' !&
!&
!"
!(
!#
!)
!$
!*
!%
!+,
,-
.//0102314
5
5
62785903: &955;:+ !$%&+
<=2>5903: &955;:+ !+ +&
62785903: "955;:+ !$%&+
<=2>5903: "955;:+ !+ +&
62785903: (955;:+ !$%&+
<=2>5903: (955;:+ !+ +&
62785903: #955;:+ !$%&+
<=2>5903: #955;:+ !+ +&
62785903: )955;:+ !$%&+
<=2>5903: )955;:+ !+ +&
62785903: $955;:+ !$%&+
<=2>5903: $955;:+ !+ +&
62785903: *955;:+ !$%&+
<=2>5903: *955;:+ !+ +&
62785903: %955;:+ !$%&+
<=2>5903: %955;:+ !+ +&
62785903: +955;:+ !$%&+
<=2>5903: +955;:+ !+ +&
62785903:& 955;:+ !$%&+
<=2>5903:& 955;:+ !+ +&
(b) Efficiency vs normalized load power
19UNIVERSIDADE FEDERALDE SANTA CATARINA
Conclusions RFLaboratory
• In this paper an analytic expression was developed and used forquantifying the power-efficiency trade-off involved in DC voltagesource that supplies energy to an electric load.
• The expression was validated by an experimental demonstrator,with good agreement between the results and the predicted val-ues.
• The analytical work developed would be a useful tool for design-ers and students in order to understand and solve a lot of confu-sion and myths about power supply efficiency and life time of thebattery.
20UNIVERSIDADE FEDERALDE SANTA CATARINA
Conclusions RFLaboratory
• In this paper an analytic expression was developed and used forquantifying the power-efficiency trade-off involved in DC voltagesource that supplies energy to an electric load.
• The expression was validated by an experimental demonstrator,with good agreement between the results and the predicted val-ues.
• The analytical work developed would be a useful tool for design-ers and students in order to understand and solve a lot of confu-sion and myths about power supply efficiency and life time of thebattery.
21UNIVERSIDADE FEDERALDE SANTA CATARINA
Conclusions RFLaboratory
• In this paper an analytic expression was developed and used forquantifying the power-efficiency trade-off involved in DC voltagesource that supplies energy to an electric load.
• The expression was validated by an experimental demonstrator,with good agreement between the results and the predicted val-ues.
• The analytical work developed would be a useful tool for design-ers and students in order to understand and solve a lot of confu-sion and myths about power supply efficiency and life time of thebattery.
22UNIVERSIDADE FEDERALDE SANTA CATARINA
References I RFLaboratory
[1] R. E. Stross, The wizard of menlo park: how thomas alva edison invented the modern world. New York: ThreeRivers Press (CA), 2008.
[2] [Online]. Available: http://www.physics.otago.ac.nz/research/epg/research
[3] Z. Pang, “Technologies and architectures of the internet-of-things (iot) for health and well-being,” Dept. Elect.Eng, Ph.D. thesis, Royal Institute of Technology Univ., Stockholm, Sweden. 2013.
[4] R. W. Erickson and D. Maksimovic, Fundamentals of power electronics. New York: Springer, 2001.
[5] S. D’Ambrosio, S. De Pasquale, G. Iannone, D. Malandrino, A. Negro, G. Patimo, A. Petta, V. Scarano, L. Serra,and R. Spinelli, “Mobile phone batteries draining: Is green web browsing the solution?” in Proc. 2014 GreenComputing Conf., Nov 2014, pp. 1–10.
[6] A. Fajardo and F. Rangel de Sousa, “Ideal energy power source model and its implications on batterymodeling,” presented at the XXII IBERCHIP Workshop, Florianopolis, Brasil. 2016.
23UNIVERSIDADE FEDERALDE SANTA CATARINA
ObrigadoE-mail: [email protected]
Site: http://lrf.ufsc.br/