文档库 最新最全的文档下载
当前位置:文档库 › 1999. The American Astronomical Society. All rights reserved. Printed in U.S.A. THE FORNAX

1999. The American Astronomical Society. All rights reserved. Printed in U.S.A. THE FORNAX

1999. The American Astronomical Society. All rights reserved. Printed in U.S.A. THE FORNAX
1999. The American Astronomical Society. All rights reserved. Printed in U.S.A. THE FORNAX

The Astrophysical Journal,511:L97–L100,1999February1

?1999.The American Astronomical Society.All rights reserved.Printed in U.S.A.

THE FORNAX SPECTROSCOPIC SURVEY:THE NUMBER OF UNRESOLVED COMPACT GALAXIES Michael J.Drinkwater,1,2Steven Phillipps,3Michael D.Gregg,2,4Quentin A.Parker,5Rodney M.Smith,6

Jonathan I.Davies,6J.Bryn Jones,3and Elaine M.Sadler7

Received1998August21;accepted1998November24;published1998December10

ABSTRACT

We describe a sample of13bright(),compact galaxies at low redshift()behind

18.5!B!20.10.05!z!0.21

J

the Fornax Cluster.These galaxies are unresolved on UK Schmidt sky survey plates,and so they would be missing from most galaxy catalogs compiled from this material.The objects were found during initial observations of The Fornax Spectroscopic Survey.This project is using the Two-degree Field spectrograph on the Anglo-Australian Telescope to obtain spectra for a complete sample of all14,000objects,stellar and nonstellar,with ,in a12deg2area centered on the Fornax Cluster of galaxies.The surface density of compact

16.5!B!19.7

J

galaxies with magnitudes is deg?2,representing of all local()galaxies

16.5!B!19.77?3 2.8%?1.6%z!0.2

J

to this limit.There are deg?2with.They are luminous(?,for

12?316.5!B!20.221.5!M!?18.0

J B

km s?1Mpc?1),and most have strong emission lines(H a equivalent widths of40–200A?)and small H?50

sizes typical of luminous H ii galaxies and compact narrow emission line galaxies.Four out of13have red colors and early-type spectra,and so they are unlikely to have been detected in any previous surveys.

Subject headings:galaxies:compact—galaxies:general—galaxies:starburst

1.INTRODUCTION

Galaxy detection in many optical surveys,especially those

based on photographic data,suffers from strong selection ef-

fects as a function of surface brightness.The dif?culty of de-

tecting low surface brightness galaxies is well accepted(Impey,

Bothun,&Malin1988;Ferguson&McGaugh1995),but at

the other extreme,it has been argued that there is no strong

selection against high surface brightness galaxies(Allen&Shu

1979;van der Kruit1987).Most galaxy surveys that are

based on photographic material()have as-

B!21

J sumed—implicitly—that very few,if any,galaxies are unre-

solved(see,e.g.,Maddox et al.1990b).Morton,Krug,&Trit-

ton(1985)attempted to check this,taking spectra of all606

stellar objects brighter than in an area of0.31deg2,

B?20

but found no galaxies.Colless et al.(1991)found seven gal-

axies among a sample of117faint compact objects,but these

were so faint()that the image classi?cations were B?22.5

J

not conclusive.

Many unresolved galaxies have been found in QSO surveys

(Downes&Margon1981;Koo&Kron1988;Boyle,Jones,

&Shanks1991).More recently,the Edinburgh-Cape blue ob-

ject survey(Stobie et al.1997)and the Anglo-Australian Ob-

servatory,Two-degree Field QSO redshift survey(Boyle et al.

1999)have produced further examples.Many compact galaxies

have also been found among H ii galaxies in objective prism

surveys:some50%of these have starlike morphology(Melnick

1987).The compact narrow emission line galaxies(CNELGs)

1School of Physics,University of New South Wales,Sydney,NSW2052, Australia.

2Visiting Astronomer,Cerro Tololo Inter-American Observatory,National Optical Astronomy Observatories.

3Department of Physics,University of Bristol,Royal Fort,Tyndall Avenue, Bristol BS81TL,England,UK.

4University of California at Davis and Lawrence Livermore National Lab-oratory,Institute of Geophysics and Planetary Physics,P.O.Box808,L-413, Livermore,CA94550.

5Anglo-Australian Observatory,P.O.Box296(167Vimiera Road),Epping, NSW2121,Australia.

6Department of Physics and Astronomy,University of Wales,Cardiff,P.O. Box913,Cardiff CF23YB,Wales,UK.

7School of Physics,A28,University of Sydney,NSW2006,Australia.

found in the Koo&Kron(1988)survey have been studied in detail(Koo et al.1994,1995;Guzma′n et al.1996,1998):35 have been found in an area of1.2deg2to a magnitude limit of.These are very blue,with luminosities,scale B?22.5

J

sizes,and emission-line spectra typical of nearby luminous

H ii galaxies(cf.Salzer,MacAlpine,&Boroson1989;Terlevich

et al.1991;Gallego et al.1997).Similar galaxies have been found at higher()redshifts(Phillips et al.1997),

0.4!z!1

and their distribution may even extend to(Lowenthal et

z≈3

al.1997).

In this Letter,we describe a new sample of bright (),compact galaxies that are unresolved on the pho-B?20.1

J

tographic sky survey plates commonly used to create galaxy catalogs.Unlike previous work,this sample is from a complete spectroscopic survey of all objects in an area of sky,and so we can estimate the fraction of all galaxies that are compact.

A population of compact galaxies that is missing in normal

galaxy surveys(see,e.g.,Colless1998)would be important for several reasons(Schade&Ferguson1994).

2.THE FORNAX SPECTROSCOPIC SURVEY

The Fornax Spectroscopic Survey(see Drinkwater et al.1998 for details)is designed to provide a census of galaxies in the local universe that is free of morphological selection criteria.

We are using the Two-degree Field(2dF)spectrograph on the Anglo-Australian Telescope to obtain spectra for all14,000 objects,stellar and nonstellar,in four2dF?elds(12.5deg2) centered on the Fornax Cluster,with magnitude limits of

(and somewhat deeper for unresolved images).

16.5!B!19.7

J

Our targets are drawn from a UK Schmidt sky survey plate

B

J

that is centered on the Fornax Cluster(Phillipps et al.1987) and digitized by the Automated Plate Measuring(APM)Fa-cility(Irwin,Maddox,&McMahon1994).Although we ob-serve objects of all morphological types,we used the automated APM classi?cations of the objects as“stellar”(probably stars) or“resolved”(probably galaxies)in order to optimize our pho-tographic photometry.The magnitudes of the resolved objects were measured by?tting exponential intensity pro?les to the run of area against the isophotal threshold in the APM data (Davies et al.1988;Davies1990).The stellar magnitudes

B

J

L97

L98UNRESOLVED COMPACT GALAXIES Vol.

511

Fig.1.—Histograms showing the completeness of our observations as functions of magnitude and color for stellar and resolved objects.The colors were taken from the APM catalog using magnitudes derived from stellar pro?le ?tting,and so they are only indicative for the resolved objects.In each case,the upper histogram is the total number of objects,and the lower histogram gives the number observed and identi?ed.The triangles indicate the locations of the new compact galaxies.

TABLE 1

Properties of the Compact Galaxies

R.A.(J2000)Decl.(J2000)z V (mag)B ?V (mag)V ?I (mag)B J (mag)a B ?V 0(mag)M B a (mag)[O iii ]/H b [N ii ]/H a W H a (A ?)33445.47?353818.00.045318.73?0.080.52?0.100.48?0.1419.10.43?18.0 4.00.123933453.03?360303.50.213019.20?0.08 1.00?0.160.83?0.1519.90.57?21.1 2.50.214033533.06?350112.80.159318.12?0.050.55?0.060.90?0.0718.50.55?21.5 1.80.2513333856.50?354500.30.115719.42?0.090.22?0.120.71?0.1619.60.22?19.8 4.60.1218933918.37?353240.70.183818.40?0.040.93?0.10 1.27?0.0519.10.55?21.5)0.42633951.33?354752.80.155319.09?0.060.86?0.14 1.23?0.0819.70.55?20.4)))34006.66?360427.10.115618.79?0.050.74?0.090.80?0.0919.30.51?20.6)0.395934025.54?345833.50.103919.11?0.070.47?0.120.90?0.1519.40.47?19.6))834057.25?351034.10.161619.08?0.06 1.30?0.20 1.00?0.1020.00.84?20.4)))34132.89?352008.40.077819.73?0.180.50?0.200.65?0.2820.10.34?18.3 3.290.134434156.94?354401.00.116319.71?0.120.61?0.210.82?0.2020.10.39?19.2 4.010.113134159.59?350901.20.139118.84?0.060.80?0.11 1.05?0.0919.40.53?21.0 1.720.3413634238.68

?35

5621.9

0.1070

19.42

?0.09

0.58

?0.14

0.92

?0.12

19.8

0.37

?19.3

3.10

0.21

138

Note.—Units of right ascension are hours,minutes,and seconds,and units of declination are degrees,arcminutes,and arcseconds.a

K -corrected (Coleman,Wu,&Weedman 1980)using km s ?1Mpc ?1and .

H ?50q ?0.100were taken from the APM catalog data (Irwin et al.1994),which uses internal self-calibration to ?t stellar pro?les,cor-recting for the nonlinear response of the photographic emulsion.Here we present preliminary results from the ?rst ?eld cen-tered at a ?03h 38m 29s ,d ??35?27?01?(J2000)observed in semesters 1996B and 1997B.We observed and identi?ed 992(77%)of the resolved objects to a limit of ,675(38%)B !19.7J of stellar objects to the same limit and a total of 1112(43%)of the stellar objects to the deeper limit of .Figure 1B !20.2J shows the completeness of our observations as a function of magnitude and color.Our main result is that 13of the “stellar”objects have recession velocities of 14,000–60,000km s ?1(see Table 1).These galaxies are well beyond the Fornax Cluster

(km s ?1),and most (nine)have strong emission-line v ?1500spectra.

3.PROPERTIES OF THE NEW GALAXIES

In Figure 2,we compare the distribution of the new compact galaxies with previously detected CNELGs (Koo et al.1994,1995)in magnitude-redshift space.There is considerable over-lap in absolute magnitude,but,as expected from a larger area survey with a brighter magnitude limit,our galaxies occupy a region in this diagram at lower redshift and brighter apparent magnitude.

The compact nature of the new galaxies prohibits a detailed

No.2,1999DRINKWATER ET AL.

L99

Fig.2.—Distribution of absolute and apparent magnitudes of the new com-pact galaxies (triangles )compared with the Koo et al.(1994,1995)CNELGs (crosses )as a function of

redshift.

Fig.3.—Emission-line diagnostic diagram of [O iii ]/H b vs.[N ii ]/H a .The new compact galaxies (triangles )are compared with CNELGs (crosses )and a range of local galaxies from the UCM survey (Gallego et al.1997).

analysis of their scale sizes and central surface brightness using our imaging data.The galaxy images are unresolved on pho-tographic sky survey plates (1?.5seeing FWHM).We therefore estimate conservative upper limits (not correcting for photo-graphic saturation,which occurs at about 21B mag arcsec ?2)to their scale lengths to be ?1?(assuming an image FWHM of 1?.5).This upper limit has been con?rmed by a CCD image of one of the galaxies taken with the Cerro Tololo Inter-Amer-ican Observatory 81.5m Telescope,which was only marginally resolved in 1?.2seeing.At the range of distances indicated,this corresponds to physical scale sizes of 1–4kpc,9which are somewhat smaller than local spiral galaxies (de Jong 1996)and at least as small as CNELGs and luminous H ii galaxies (Phil-lips et al.1997).Despite the small scales,these galaxies are not dwarfs in terms of their luminosities,which are within a factor 10or so of ;indeed,some of them exceed .This L L ??is true for the following reason:their high surface brightnesses;the scale size limits of 1?imply central surface brightnesses 19–21B mag arcsec ?2,as bright as the CNELGs and the lu-minous H ii galaxies (Phillips et al.1997).

We obtained Cousins BVI CCD images of our survey region using the CTIO Curtis Schmidt Telescope.The low spatial resolution (3?FWHM)leaves the compact galaxies unresolved,but the data allow us to calculate the photometry and aperture (8?radius)colors.The K -corrected colors (Table 1)of B ?V the emission-line compact galaxies place them among the CNELG and H ii galaxies,and they are consistent with the relatively high,recent star formation rates (Larson &Tinsley 1978).

The nine emission-line compact galaxies all have strong nar-row H a lines:none are resolved at our resolution of 9A

?or 450km s ?1

.The H a rest equivalent widths listed in Table 1nearly all exceed the mean value for our overall background

sample of emission-line objects of EW(H a )?37A

?,which is typical of local spirals (Kennicutt 1992).The values for the

compact galaxies,~30–190A

?,are more like those seen in low-redshift H ii galaxies (Gallego et al.1997).Given the range

8

CTIO is operated by the Association of Universities for Research in As-tronomy,Inc.,under a cooperative agreement with the National Science Foun-dation as part of the National Optical Astronomy Observatories.9

We adopt km s ?1Mpc ?1and .

H ?50q ?0.100of overall sizes of these objects,it is interesting to consider what Cowie et al.(1996)call the stellar mass doubling time,i.e.,the time it would take for the current star formation rate (SFR)to double the existing underlying stellar mass.This can be derived directly from the equivalent widths:Cowie et al.

note that EW(H a )?60A

?separates galaxies undergoing rapid star formation,with mass doubling times less than 1010yr,from those with moderate SFRs that can be maintained for a Hubble time.At the highest SFR,Cowie et al.?nd that a mass doubling time of yr corresponds empirically to their galaxies

92#10with EW(H a )?125A

?.Our fastest star formers,like J0338?3545,should have mass doubling times of this order.In Figure 3,we show the [O iii ]/H b versus [N ii ]/H a emis-sion-line ratio diagram for our new galaxies compared with the Gallego et al.(1997)emission-line sample and the Koo et https://www.wendangku.net/doc/ea13087071.html,ELGs.This shows that new compact galaxies are actively star-forming as they closely follow the general H ii region relationship.They display generally high excitation as mea-sured by [O iii ]/H b ,putting them in the H ii galaxies with hot spots (H ii H)class more than the starburst nucleus class as de?ned by Gallego et al.The new compact galaxies have very similar properties to the CNELGs.We do not draw any con-clusion from the lack of low-excitation objects:this may be a selection effect,since we are only considering the unresolved galaxies in our survey in this sample.Similar conclusions can be drawn from a plot of the excitation against absolute magnitude.

We also found four compact galaxies that did not have strong emission lines and therefore are not shown in the excitation diagrams.These are unlikely to have been detected in previous work on compact galaxies because of their weak line emission and generally redder colors,which would exclude them from most QSO surveys.One of them,J0339?3547,has a posts-tarburst spectrum with strong Balmer absorption lines.We have used the two ratios of absorption feature strengths Ca ii H ?H e /Ca ii K and H d /Fe i l 4045to estimate the age of the galaxy since the end of the starburst (Leonardi &Rose 1996).For J0339?3547,these ratios are 0.89and 0.69,respectively.For the Leonardi &Rose model of a starburst lasting 0.3Gyr,this is indicative of a very strong starburst about 1Gyr after the end of the burst.This galaxy may represent an intermediate

L100UNRESOLVED COMPACT GALAXIES Vol.511

stage between the CNELG types and the dwarf spheroidal rem-

nants proposed by Koo et al.(1995).By comparison,

J0340?3510has ratios of1.12and0.97,and a composite

spectrum of60normal,early-type galaxies from the survey

has values of1.10and0.95,both consistent with the Leonardi

&Rose values for an old population.

4.NUMBERS OF COMPACT GALAXIES

To estimate the true numbers of compact galaxies from our

sample,we must?rst make a completeness correction.The

color distributions in Figure1show that the compact galaxies

are all bluer than,so the best correction can be

B?R?1.6

J F

taken from the fraction of blue()stellar objects

B?R!1.6

J F

observed;to,we observed31%of the blue stellar B?19.7

J

objects,so the corrected number of compact galaxies to this

limit is,which is equivalent to a surface den-7/0.31?23?9

sity of deg?2.The surface density to(com-7?3B?20.2

J

pleteness of35%)is deg?2.

12?3

We can use our observations of?eld galaxies to estimate the

fraction of normal galaxies represented by the compact gal-

axies.At redshifts,the H a line is shifted out of our

z10.2

2dF spectra,and our galaxy sample is less complete,so we

de?ne a“local”comparison?eld sample to be all galaxies in

the?eld beyond the Fornax Cluster but at redshifts.We

z!0.2

successfully observed992resolved objects to,of

B?19.7

J

which675were“local”background galaxies:this is the min-

imum number of local?eld galaxies.There are1296?

resolved objects still to observe,so the maximum 992?304

number of local?eld galaxies is.The number

675?304?979

of local()background galaxies to our limit is

z!0.2827?

.Therefore,the expected23compact galaxies among the 152

stellar objects constitute of the local galaxy pop-

2.8%?1.6%

ulation.These would be missed by any surveys of objects

classi?ed by the APM as“galaxies”from UK Schmidt pho-

tographic data with.These selection criteria

16.5!B!19.7

J

are typical of previous surveys(Maddox,Efstathiou,&Suth-

erland1990a;Colless1998).

This conclusion is for a magnitude,not volume,limited sam-ple,but in fact the compact galaxies do occupy a similar volume of space to the general run of galaxies to this magnitude limit. For instance,the2dF galaxy redshift survey,limited at a very similar to ours,has a mean redshift of33,000km s(Colless

?1

B

J

1998),which is close to the mean of the compact galaxies.The galaxy catalog used for that survey has a mean surface density of222deg?2at(M.Colless1998,private commu-

B!19.7

J

nication),for which the compact galaxies would represent an additional.

3.2%?1.2%

Only about half of the local galaxy sample exhibits signif-icant emission-line features,so the new compact galaxies con-stitute a larger fraction of emission-line galaxies(3%–5%). They contribute an even larger fraction of strong H a emitters with A?,and so they may make a small but meas-EW(H a)140

urable contribution to the local star formation rate.

The new compact galaxies have very similar absolute mag-nitudes,sizes,and(in most cases)emission-line properties to the Koo et al.(1994,1995)CNELGs.The distributions shown in Figure2suggest that they are a continuation of the CNELGs to lower redshifts and brighter apparent magnitudes.A better way to compare these populations is by the volume density. Koo et al.(1994)derive a CNELG density of Mpc?3

?5

7.5#10 compared with the value of Mpc?3for local H ii H

?5

17#10

?dwarf H ii H galaxies(Salzer et al.1989).For our sample of

compact galaxies,using the1/V

max

method,we obtain a similar value of Mpc?3.These results are consistent

?5

(13?4)#10

with the respective galaxy populations being related,but we prefer not to draw any conclusions until we can analyze the compact galaxies in the context of our complete sample.

We thank our referee for detailed suggestions that greatly improved the presentation of this work.We also thank Lewis Jones for helpful discussions and Jesus Gallego for providing data for Figure3.S.P.acknowledges the support of the Royal Society via a University Research Fellowship.J.B.J.is sup-ported by the UK PPARC.Part of this work was done at the Institute of Geophysics and Planetary Physics,under the aus-pices of the US Department of Energy by Lawrence Livermore National Laboratory under contract W-7405-Eng-48.

REFERENCES

Allen,R.J.,&Shu,F.H.1979,ApJ,227,67

Boyle,B.J.,Jones,L.R.,&Shanks,T.1991,MNRAS,251,482

Boyle,B.J.,Smith,R.J.,Shanks,T.,Croom,S.M.,Miller,L.,&Read,M. 1999,in IAU Symp.183,Cosmological Parameters and the Evolution of the Universe,ed.K.Sato(Boston:Kluwer),in press

Coleman,G.D.,Wu,C.-C.,&Weedman,D.W.1980,ApJS,43,393 Colless,M.1998,Philos.Trans.R.Soc.London A,in press

Colless,M.,Ellis,R.S.,Taylor,K.,&Shaw,G.1991,MNRAS,253,686 Cowie,L.L.,Songaila,A.,Hu,E.M.,&Cohen,J.G.1996,AJ,112,839 Davies,J.I.1990,MNRAS,245,350

Davies,J.I.,Phillipps,S.,Cawson,M.G.M.,Disney,M.J.,&Kibblewhite, E.J.1988,MNRAS,232,239

de Jong,R.S.1996,A&A,313,45

Downes,R.A.,&Margon,B.1981,AJ,86,19

Drinkwater,M.J.,Phillipps,S.,Davies,J.I.,Gregg,M.D.,Jones,J.B.,Parker, Q.A.,Sadler,E.M.,&Smith,R.M.1998,in preparation

Ferguson,H.C.,&McGaugh,S.S.1995,AJ,440,470

Gallego,J.,Zamorano,J.,Rego,M.,&Vitores,A.G.1997,ApJ,475,502 Guzma′n,R.,Jangren,A.,Koo,D.C.,Bershady,M.A.,&Simard,L.1998, ApJ,495,L13

Guzma′n,R.,Koo,D.C.,Faber,S.M.,Illingworth,G.D.,Takamiya,M.,Kron, R.G.,&Bershady,M.A.1996,ApJ,460,L9

Impey,C.,Bothun,G.,&Malin,D.1988,ApJ,330,634

Irwin,M.,Maddox,S.,&McMahon,R.1994,Spectrum,2,14 Kennicutt,R.C.1992,ApJS,79,255Koo,D.C.,Bershady,M.A.,Wirth,G.D.,Stanford,S.A.,&Majewski,S. R.1994,ApJ,427,L9

Koo,D.C.,Guzma′n,R.,Faber,S.M.,Illingworth,G.D.,Bershady,M.A., Kron,R.G.,&Takamiya,M.1995,ApJ,440,L49

Koo,D.C.,&Kron,R.G.1988,ApJ,325,92

Larson,R.B.,&Tinsley,B.M.1978,ApJ,219,46

Leonardi,A.J.,&Rose,J.A.1996,AJ,111,182

Lowenthal,J.D.,et al.1997,ApJ,489,543

Maddox,S.J.,Efstathiou,G.,&Sutherland,W.J.1990a,MNRAS,246,433 Maddox,S.J.,Sutherland,W.J.,Efstathiou,G.,&Loveday,J.1990b,MNRAS, 243,692

Melnick,J.1987,in Starbursts and Galaxy Evolution,ed.X.T.Trinh,T. Montmerle,&J.T.V.Tran(Gif-sur-Yvette:Editions Frontie`res),215 Morton,D.C.,Krug,P.A.,&Tritton,K.P.1985,MNRAS,212,325 Phillipps,S.,Disney,M.J.,Kibblewhite,E.J.,&Cawson,M.G.M.1987, MNRAS,229,505

Phillips,A.C.,Guzma′n,R.,Gallego,J.,Koo,D.C.,Lowenthal,J.D.,Vogt, N.P.,Faber,S.M.,&Illingworth,G.D.1997,ApJ,489,543

Salzer,J.J.,MacAlpine,G.M.,&Boroson,T.A.1989,ApJS,70,479 Schade,D.,&Ferguson,H.C.1994,MNRAS,267,889

Stobie,R.,et al.1997,MNRAS,287,848

Terlevich,R.,Melnick,J.,Masegosa,J.,Moles,M.,&Copetti,M.V.F.1991, A&AS,91,285

van der Kruit,P.C.1987,A&A,173,59

嵌入式系统的低功耗设计

第27卷第6期增刊 2006年6月 仪 器 仪 表 学 报 Chinese Journal of Scientific Instrument Vol.27No.6 J une.2006  嵌入式系统的低功耗设计 3 杨天池 金 梁 王天鹏 (解放军信息工程大学 郑州 450002) 摘 要 嵌入式系统的电源管理是系统设计中关键部分,合理的电源管理方案可以减少系统的功耗并提高整体性能。本文提出了一种层次化的电源管理结构,分别为硬件层、驱动层、操作系统层、电源管理层和应用层。本文同时引入了动态的电源管理方法来解决电源功耗的动态管理问题。通过在实际的系统中的测试表明,该电源管理机制的有效性。关键词 嵌入式系统 低功耗设计 动态电源管理 PXA255 Low pow er design in embedded system Yang Tianchi Jin Liang Wang Tianpeng (Universit y of I nf ormation Engineering ,Zhengz hou 450002,China ) Abstract Proper power management mechanism is important when designing embedded system.It is helpful to reduce power consumption and improve performance.This low power model adopt s five 2layer architecture ,which are hardware platform ,driver layer ,operating system ,power manage mechanism and application program.Dynamic power management (DPM )technology is also introduced to solve the problem of power consumption.The experiment on embedded system demonstrates t hat this power management mechanism is feasible.K ey w ords embedded system low power design dynamic power management PXA255  3基金项目:河南人才创新基金(0421000100) 1 引 言 随着嵌入式系统的发展以及应用面的不断扩展,功耗控制是系统设计中必不可少的组成部分。如何最大限度的降低系统功耗、减少不必要的能源损失、延长电池使用时间已经成为嵌入式系统特别是便携式系统设计中研究的热点问题。系统的低功耗设计,并非是某一方面、某一角度的解决方案,而应当从系统级的设计考虑功耗的节省,是一个硬件设计与软件控制相互结合的协调过程。 2 低功耗电路模型 低功耗设计对于无线设备、PDA 等便携式设备的实际应用具有重要的意义。低功耗元件的发展和系统设计的进步使得通用计算技术可以用到表、无线电话、 PDA 和桌面计算机中。在这些系统中的电源管理技 术传统上集中在休眠模式和设备能源管理这2个方面上[1]。但是,这样的电源管理缺乏直观性和灵活性,而且功耗的降低,并非单独软件、硬件单方面可以解决的[2],因此设计并建立如图1所示的系统低功耗设计模型。整个模型由硬件平台,驱动层,操作系统层,电源管理机制层和应用程序五个部分组成。 2.1 硬件平台 几乎所有系统功耗都集中于硬件平台,因此降低硬件平台的功耗是实现低功耗的基本所在。公式(1)为系统功耗的表达式: P ∞CV 2 f (1) 式中:C 是负载电容,V 是器件电压,f 是工作频率[3]。系统功耗同负载电容、器件电压平方以及工作频率成正比。因此,硬件平台设计多选用低电压,电压、频率可调器件,以及采用SOC 设计来进一步降低功耗[4,5]。另外,模式可控器件在空闲状态消耗的能量为运行状

ARM低功耗设计_全面OK

嵌入式系统中的低功耗设计 2008-12-31 18:19:55 作者:电子之都来源:电子之都浏览次数:59 网友评论 0 条 经过近几年的快速发展,嵌入式系统(Embedded system)已经成为电子信息产业中最具增长力的一个分支。随着手机、PDA、GPS、机顶盒等新兴产品的大量应用,嵌入式系统的市场正在以每年30%的速度递增(IDC预测),嵌入式系统的设计也成为软硬件工程师越来越关心的话题。 在嵌入式系统的设计中,低功耗设计(Low-Power Design)是许多设计人员必须面对的问题,其原因在于嵌入式系统被广泛应用于便携式和移动性较强的产品中去,而这些产品不是一直都有充足的电源供应,往往是靠电池来供电,所以设计人员从每一个细节来考虑降低功率消耗,从而尽可能地延长电池使用时间。事实上,从全局来考虑低功耗设计已经成为了一个越来越迫切的问题。 那么,我们应该从哪些方面来考虑低功耗设计呢?笔者认为应从以下几方面综合考虑: 1.处理器的选择 2.接口驱动电路设计 3.动态电源管理 4.电源供给电路的选择 下面我们分别进行讨论: 一、处理器的选择 我们对一个嵌入式系统的选型往往是从其CPU和操作系统(OS)开始的,一旦这两者选定,整个大的系统框架便选定了。我们在选择一个CPU的时候,一般更注意其性能的优劣(比如时钟频率等)及所提供的接口和功能的多少,往往忽视其功耗特性。但是因为CPU 是嵌入式系统功率消耗的主要来源---对于手持设备来讲,它几乎占据了除显示屏以外的整

个系统功耗的一半以上(视系统具体情况而定),所以选择合适的CPU对于最后的系统功耗大小有举足轻重的影响。 一般的情况下,我们是在CPU的性能(Performance)和功耗(Power Consumption)方面进行比较和选择。通常可以采用每执行1M次指令所消耗的能量来进行衡量,即Watt/M IPS。但是,这仅仅是一个参考指标,实际上各个CPU的体系结构相差很大,衡量性能的方式也不尽相同,所以,我们还应该进一步分析一些细节。 我们把CPU的功率消耗分为两大部分:内核消耗功率PCORE和外部接口控制器消耗功率PI/O,总的功率等于两者之和,即P=PCORE+PI/O。对于PCORE,关键在于其供电电压和时钟频率的高低;对于PI/O来讲,除了留意各个专门I/O控制器的功耗外,还必须关注地址和数据总线宽度。下面对两者分别进行讨论: 1、CPU供电电压和时钟频率 我们知道,在数字集成电路设计中,CMOS电路的静态功耗很低,与其动态功耗相比基本可以忽略不计,故暂不考虑。其动态功耗计算公式为: Pd=CTV2f 式中,Pd---CMOS芯片的动态功耗 CT----CMOS芯片的负载电容 V----CMOS芯片的工作电压 f-----CMOS芯片的工作频率 由上式可知,CMOS电路中的功率消耗是与电路的开关频率呈线性关系,与供电电压呈二次平方关系。对于一颗CPU来讲,Vcore电压越高,时钟频率越快,则功率消耗越大。所以,在能够满足功能正常的前提下,尽可能选择低电压工作的CPU能够在总体功耗方面得到

基于MSP430的极低功耗系统设计

基于MSP430的极低功耗系统设计 摘要:MSP430是TI公司出品的一款强大的16位单片机,其显著特点是具有极低的功耗。本文对构造以MSP430为基础极低功耗系统作为有益的探讨,对于设计各种便携式设备都具有较高的参考价值。 对于一个数字系统而言,其功耗大致满足以下公式:P=CV2f,其中C为系统的负载电容,V为电源电压,f为系统工作频率。由此可见,功耗与电源电压的平方成正比,因此电源电压对系统的功耗影响最大,其次是工作频率,再就是负载电容。负载电容对设计人员而言,一般是不可控的,因此设计一个低功耗系统,应该考虑到不影响系统性能前提下,尽可能地降低电源的电压和使用低频率的时钟。下面对TI公司新出MSP430来具体探讨这个问题。 MSP430具有工业级16位RISC,其I/O和CPU可以运行在不的时钟下。CPU功耗可以通过开关状态寄存器的控制位来控制:正常运行时电流160μA,备用时为0.1μA,功耗低,为设计低功耗系统提供了有利的条件。 图1是我们设计的以MSP430为CPU的“精密温度测试仪”(下面简称测试仪)。该产品使用电池供电,体积小巧,携带方便。 在使用时应该尽可能地选择最低的电源电压。对于MSP430而言,可用的最低电压是很低的,最低可达1.8V。我们使用TI公司推荐使用的3V。通常的电源只提供5V电压,因此,需要将5V电压由一个3V的稳压管降压后给CPU供电,也可以直接锂电池供电。3V不是标准的TTL电平,因此,在使用时需要用接口电路使CPU的非TTL标准电平能与TTL标准电平的器件连接。这些接口电路应该也是低功耗的,否则会造成一方面使用低电压降低了功耗,另一个方面使用额外的接口电路又增加了系统的功耗。或者直接使用支持3V电压的外围芯片。图1 (2)时钟频率 从低功耗的角度看,需要较低的频率,但是在实时应用中为了快速响应外部事件

单片机MSP430的极低功耗系统设计

单片机MSP430的极低功耗系统设计

超低功耗系统设计

超低功耗系统设计 学院: 学号: 姓名:

基于MSP430单片机的开关稳压电源设计 MSP430系列单片机是美国TI公司生产的新一代16位单片机,是一种超低功耗的混合信号处理器(MixedSignal Processor),它具有低电压、超低功耗、强大的处理能力、系统工作稳定、丰富的片内外设、方便开发等优点,具有很高的性价比,在工程控制等领域有着极其广泛的应用范围。开关Boost稳压电源利用开关器件控制、无源磁性元件及电容元件的能量存储特性,从输入电压源获取分离的能量,暂时把能量以磁场的形式存储在电感器中,或以电场的形式存储在电容器中,然后将能量转换到负载。对DC—DC主回路采用Boost升压斩波电路。 2 系统结构和总设计方案 本开关稳压电源是以MSP430F449为主控制器件,它是TI公司生产的16位超低功耗特性的功能强大的单片机,其低功耗的优点有利于系统效率高的要求,且其ADCl2是高精度的12位A/D转换模块,有高速、通用的特点。这里使用MSP430完成电压反馈的PI调节;PWM波产生,基准电压设定;电压电流显示;过电流保护等。 系统框图如图1所示。 3 硬件电路设计 3.1 DC/DC转换电路设计 系统主硬件电路由电源部分、整流滤波电路、DC/DC转换电路、驱动电

路、MSP430单片机等部分组成。交流输入电压经整流滤波电路后经过DC/DC变换器,采用Boost升压斩波电路DC/DC变换,如图2所示: 根据升压斩波电路的工作原理一个周期内电感L积蓄的能量与释放的能量相等,即: 式(1)中I1为输出电流,电感储能的大小通过的电流与电感值有关。在实际电路中电感的参数则与选取开关频率与输入/输出电压要求,根据实际电路的要求选用合适的电感值,且要注意其内阻不应过大,以免其损耗过大减小效率采样电路。对于电容的计算,在指定纹波电压限制下,它的大小的选取主要依据式(2): 式(2)中:C为电容的值;D1为占空比;TS为MOSFET的开关周期;I0为负载电流;V’为输出电压纹波。 3.2 采样电路 采样电路为电压采集与电流采集电路,采样电路如图3所示。其中P6.O,P6.1为MSP430芯片的采样通道,P6.O为电压采集,P6.1为电流采集。 电压采集因为采样信号要输入单片机MSP430内部,其内部采样基准电压选为2.5 V,因此要将输入的采样电压限制在2.5 V之下,考虑安全裕量则将输入电压限制在2 V以下,当输入电压为36 V时,采样电压为:12/ (12+200)×36=2.04 V,符合要求。 电流采集采用康铜丝进行采集。首先考虑效率问题,康铜丝不能选择过大,同时MSP430基准电压为2.5 V,且所需康铜丝需自制。考虑以上方面在康铜丝阻值选取上约为O.1Ω。 3.3 PWM驱动电路的设计 电力MOSFET驱动功率小,采用三极管驱动即可满足要求,驱动电路如图

DSP电源系统的低功耗设计

DSP电源系统的低功耗设计 自从美国TI公司推出通用可编程DSP芯片以来,DSP技术得到了突飞猛进的发展。DSP电源设计是DSP应用系统设计的一个重要组成部分,低功耗是DSP电源系统设计的发展方向。由于DSP一般在系统中要承担大量的实时数据计算,在CPU内部,频繁的部件转换会使系统功耗大大增加,降低DSP内部CPU供电的核电压是降低系统功耗的有效方法,因此TI公司的DSP大多采用低电压供电方式。 从一定程度上说,选择什么样的DSP就决定系统处于什么样的功耗层次。在实际应用中,电源系统直接决定了DSP能否在高性能低功耗的情况下工作,因此,一个稳定而可靠的电源系统是至关重要的。 TI公司最新推出的TPS6229X系列开关电源芯片有两种工作模式:PWM 模式和节能模式。在额定负载电流下,芯片处于PWM模式,高效稳定的为DSP 供电,当负载电流降低时,芯片自动转入节能模式,以减小系统功耗,适宜于DSP系统的低功耗设计,本文主要介绍了该芯片的特点,并给出了基于此芯片的DSP电源电路。 l DSP电源特点 1.1电源要求 TI公司的DSP需要给CPU、FLASH、ADC及I/O等提供双电源供电,分别为1.8V或2.5V核电源和3.3V的I/O电源,每种电源又分为数字电源和模拟电源,即数字1.8V(2.5V)、模拟1.8V(2.5V),数字3.3V,模拟3.3V。

相对与模拟电源和数字电源,也要求有模拟地和数字地。数字电源与模拟电源单独供电,数字地与模拟地分开,单点连接。 DSP大多采用数字电源供电,可以通过数字电源来获得模拟电源,主要有两种方式:(1)数字电源与模拟电源、数字地与模拟地之间加电感或铁氧体磁珠构成无源滤波网络。铁氧体磁珠在低频时阻抗很低,在高频时很高,可以抑制高频干扰,从而消除数字电路的噪声。(2)采用多路稳压器。方法(1)结构简单,能满足一般的应用要求,方法(2)有更好的去耦效果,但电路复杂成本高。 1.2 供电次序 TI公司DSP采用双电源供电,因此,需要考虑上电、掉电顺序。大部分DSP 芯片要求内核电压先上电,I/O电压后上电。因为如果只有CPU内核获得供电,周边I/O没有供电,对芯片不会产生损害,只是没有输入输出能力而已;如果周边I/O获得供电而CPU内核没有加电,那么DSP缓冲驱动部分的三极管处于未知状态下工作,这是很危险的。但是也有要求I/O电压先上电,内核电压后上电,如TMS320F2812。 在设计不同DSP芯片的电源系统时,要根据其不同的电源特点,否则可能造成整个电源系统的损坏。 2 TPS62290芯片介绍 2.1 芯片特点 TPS62290是TI公司最新推出的高效率同步降压DC/DC转换器,应用于手机、掌上电脑、便携式媒体播放器以及低功耗DSP电源设计中,其主要有以下特点: 输出电流高达1000mA

单片机低功耗设计

单片机系统的低功耗设计策略 摘要:嵌入式系统的低功耗设计需要全面分析各方面因素,统筹规划。在设计之初,各个因素往往是相互制约、相互影响的,一个降低系统功耗的措施有时会带来其他方面的“负效应”。因此,降低系统整体功耗,需要仔细分析和计算。本文从硬件和应用软件设计两个方面,阐述一个以单片机为核心的嵌入式系统低功耗设计时所需考虑的一些问题。 关键词:低功耗设计硬件设计应用软件设计低功耗模式 在嵌入式应用中,系统的功耗越来越受到人们的重视,这一点对于需要电池供电的便携式系统尤其明显。降低系统功耗,延长电池的寿命,就是降低系统的运行成本。对于以单片机为核心的嵌入式应用,系统功耗的最小化需要从软、硬件设计两方面入手。 随着越来越多的嵌入式应用使用了实时操作系统,如何在操作系统层面上降低系统功耗也成为一个值得关注的问题。限于篇幅,本文仅从硬件设计和应用软件设计两个方面讨论。 1 硬件设计 选用具有低功耗特性的单片机可以大大降低系统功耗。可以从供电电压、单片机内部结构设计、系统时钟设计和低功耗模式等几方面考察一款单片机的低功耗特性。 1.1 选用尽量简单的CPU内核 在选择CPU内核时切忌一味追求性能。8位机够用,就没有必要选用16位机,选择的原则应该是“够用就好”。现在单片机的运行速度越来越快,但性能的提升往往带来功耗的增加。一个复杂的CPU集成度高、功能强,但片内晶体管多,总漏电流大,即使进入STOP 状态,漏电流也变得不可忽视;而简单的CPU内核不仅功耗低,成本也低。 1.2 选择低电压供电的系统 降低单片机的供电电压可以有效地降低其功耗。当前,单片机从与TTL兼容的5 V供电降低到3.3 V、3 V、2 V乃至1.8 V供电。供电电压降下来,要归功于半导体工艺的发展。从原来的3 μm工艺到现在的0.25、0.18、0.13 μm工艺,CMOS电路的门限电平阈值不断降低。低电压供电可以大大降低系统的工作电流,但是由于晶体管的尺寸不断减小,管子的漏电流有增大的趋势,这也是对降低功耗不利的一个方面。 目前,单片机系统的电源电压仍以5 V为主,而过去5年中,3 V供电的单片机系统数量增加了1倍,2 V供电的系统也在不断增加。再过五年,低电压供电的单片机数量可能会超过5 V电压供电的单片机。如此看来,供电电压降低将是未来单片机发展的一个重要趋势。 1.3 选择带有低功耗模式的系统 低功耗模式指的是系统的等待和停止模式。处于这类模式下的单片机功耗将大大小于运行模式下的功耗。过去传统的单片机,在运行模式下有wait和stop两条指令,可以使单片机进入等待或停止状态,以达到省电的目的。 等待模式下,CPU停止工作,但系统时钟并不停止,单片机的外围I/O模块也不停止工作;系统功耗一般降低有限,相当于工作模式的50%~70%。 停止模式下,系统时钟也将停止,由外部事件中断重新启动时钟系统时钟,进而唤醒CPU继续工作,CPU消耗电流可降到μA级。在停止模式下,CPU本身实际上已经不消耗什么电流,要想进一步减小系统功耗,就要尽量将单片机的各个I/O模块关掉。随着I/O模块的逐个关闭,系统的功耗越来越小,进入停止模式的深度也越来越深。进入深度停止模式无异于关机,这时的单片机耗电可以小于20 nA。其中特别要提示的是,片内RAM停止供电后,RAM中存储的数据会丢失,也就是说,唤醒CPU后要重新对系统作初始化。因此在让系统进入深度停止状态前,要将重要系统参数保存在非易失性存储器中,如EEPROM中。深度停止模式关掉了所有的I/O,可能的唤醒方式也很有限,一般只能是复位或IRQ中断等。 保留的I/O模块越多,系统允许的唤醒中断源也就越多。单片机的功耗将根据保留唤醒方式的不同,降至1μA至几十μA之间。例如,用户可以保留外部键盘中断,保留异步串行

SoC系统的低功耗设计

SoC系统的低功耗设计 摘要:功耗问题正日益变成VLSI系统实现的一个限制因素。对便携式应用来说,其主要原因在于电池寿命,对固定应用则在于最高工作温度。由于电子系统设计的复杂度在日益提高,导致系统的功耗得到其主要功耗成分。其次,以该主要功耗成分数学表达式为依据,突出实现SoC低功耗设计的各种级别层次的不同方法。 关键词:VLSI SoC CMOS集成电路低功耗设计 引言 从20世纪80年代初到90年代初的10年里,微电子领域的很多研究工作都集中到了数 字系统速度的提高上,现如今的技术拥有的计算能力能够使强大的个人工作站、复杂实时语音和图像识别的多媒体计算机的实现 成为可能。高速的计算能力对于百姓大众来说是触指可及的,不像早些年代那样只为少数人服务。另外,用户希望在任何地方都能

访问到这种计算能力,而不是被一个有线的物理网络所束缚。便携能力对产品的尺寸、重量和功耗加上严格的要求。由于传统的镍铬电池每磅仅能提供的能量,因而功耗就变得尤为重要。电池技术正在改进,每5年最大能将电池的性能提高30%,然而其不可能在短期内显着地解决现在正遇到的功耗问题。 虽然传统可便携数字应用的支柱技术已经 成功地用于低功耗、低性能的产品上,诸如电子手表、袖珍计算器等等,但是有很多低功耗、高性能可便携的应用一直在增长。例如,笔记本计算机就代表了计算机工业里增长最快的部分。它们要求与桌上计算机一样具有同样的计算能力。同样的要求在个人通信领域也正在迅速地发展,如采用了复杂语音编解码算法和无线电调制解调器的带袖 珍通信终端的新一代数字蜂窝网。已提出的未来个人通信服务PCS应用对这些要求尤其明显,通用可便携多媒体服务是要支持完整的数字语音和图像辨别处理的。在这些应用中,不仅语音,而且数据也要能在无线链路

MSP430低功耗设计大作业

MSP430低功耗设计大作业

MSP430低功耗设计 1.MSP430单片机有几种工作模式,在中断子程序中如何设置,可以使系统从LPM4 模式进入活动模式。 答:五种低功耗功耗模式,分别为LPM0~LPM4;cpu的活动状态成为AM. LPM0模式:关闭CPU;LPM1、LPM2模式:通过开启、关闭不同时钟源控制系统功耗;LPM3模式:时钟开启时的最低功耗模式,仅低频时钟处于运行状态。Lpm4模式工作时只保存RAM区数据,CPU只能通过I/O口外部中断唤醒。在主函数中进入休眠模式并打开总中断,然后在中断程序里面执行你想要的操作就可以了。 2看门狗用于看门狗监测的原理是什么? 答:工作原理是在系统运行以后也就启动了看门狗的计数器,看门狗就开始自动计数,如果到了一定的时间还不去清看门狗,那么看门狗计数器就会溢出从而引起看门狗中断,造成系统复位。所以,在使用有看门狗的芯片时要注意清看门狗。在由单片机构成的微型计算机系统中,由于单片机的工作常常会受到来自外界电磁场的干扰,造成程序的跑飞,而陷入死循环,程序的正常运行被打断,由单片机控制的系统无法继续工作,会造成整个系统的陷入停滞状态,发生不可预料的后果,所以出于对单片机运行状态进行实时监测的考虑,便产生了一种专门用于监测单片机程序运行状态的芯片,俗称"看门狗"看门狗电路的应用,使单片机可以在无人状态下实现连续工作,其工作原理是:看门狗芯片和单片机的一个I/O引脚相连,该I/O引脚通过程序控制它定时地往看门狗的这个引脚上送入高电平(或低电平),这一程序语句是分散地放在单片机其他控制语句中间的,一旦单片机由于干扰造成程序跑飞后而陷入某一程序段不进入死循环状态时,写看门狗引脚的程序便不能被执行,这个时候,看门狗电路就会由于得不到单片机送来的信号,便在它和单片机复位引脚相连的引脚上送出一个复位信号,使单片机发生复位,即程序从程序存储器的起始位置开始执行,这样便实现了单片机的自动复位。 3.捕获/比较寄存器CCR0有什么特殊性? 答:用做捕获时:捕获的同时TAR的值会传给CCRx,用来测算周期是很好的方法,一般捕获用法时无须设置参数。 用作比较时:CCR0一般用来设置输出电平的转换时机,就是TAR计数到CCR0时输出电平发生相应的变化(输出方式可以设置);也可以将CCR0设置为最大值,此时CCRx(x:1或2)用来设置输出电平转换的时机,即TAR计数到CCRx 时输出电平转换,计数到CCR0时重新开始计数。 4.定时器A工作于捕获,比较,定时方式时,中断标志在什么情况下置位? 捕捉:看你如何设定的控制寄存器,上一个上升沿捕捉还是一个下降沿捕捉还是多个上升沿捕捉,捕捉到了中断标志置位。 比较:当定时器的值(随系统运行在不断的加1)和比较寄存器的值相同的时候,中断标志置位 定时:定时器值溢出(从全1变成全0)的时候中断标志置位。

系统低功耗设计

系统低功耗设计 论文关键词:集成电路低功耗设计SoC 论文摘要:功耗问题正日益变成VLSI系统实现的一个限制因素。对 便携式应用来说,其主要原因在于电池寿命,对固定应用则在于最高 工作温度。因为电子系统设计的复杂度在日益提升,导致系统的功耗 得到其主要功耗成分。其次,以该主要功耗成分数学表达式为依据, 突出实现SoC低功耗设计的各种级别层次的不同方法。 引言 从20世纪80年代初到90年代初的10年里,微电子领域的很多研究 工作都集中到了数字系统速度的提升上,现如今的技术拥有的计算水 平能够使强大的个人工作站、复杂实时语音和图像识别的多媒体计算 机的实现成为可能。高速的计算水平对于百姓大众来说是触指可及的,不像早些年代那样只为少数人服务。另外,用户希望在任何地方都能 访问到这种计算水平,而不是被一个有线的物理网络所束缚。便携水 平对产品的尺寸、重量和功耗加上严格的要求。因为传统的镍铬电池 每磅仅能提供20W.h的能量,因而功耗就变得尤为重要。电池技术正 在改进,每5年最大能将电池的性能提升30%,不过其不可能在短期内显著地解决现在正遇到的功耗问题。 虽然传统可便携数字应用的支柱技术已经成功地用于低功耗、低性能 的产品上,诸如电子手表、袖珍计算器等等,但是有很多低功耗、高 性能可便携的应用一直在增长。例如,笔记本计算机就代表了计算机 工业里增长最快的部分。它们要求与桌上计算机一样具有同样的计算 水平。同样的要求在个人通信领域也正在迅速地发展,如采用了复杂 语音编解码算法和无线电调制解调器的带袖珍通信终端的新一代数字 蜂窝网。已提出的未来个人通信服务PCS (PersonalCommunicationServices)应用对这些要求尤其明显,通用 可便携多媒体服务是要支持完整的数字语音和图像辨别处理的。在这 些应用中,不但语音,而且数据也要能在无线链路上传输。这就为实

相关文档