Vous êtes sur la page 1sur 21

GRRStudy Users Manual

Revision 1 July 27, 2009


Hemant Ashok Deshmukh (hemant.deshmukh@ti.com) e!as "nstruments

Table of Contents "ntroduction...............................................................................................................# So$t%are Architecture................................................................................................# em&erature Data......................................................................................................' (a&a)ility Data...........................................................................................................* Re&eata)ility Data......................................................................................................+ est ,a)el -ormat....................................................................................................... So$t%are Setu&........................................................................................................../0 So$t%are Run 1hase "............................................................................................// User "n&uts Sym)olsAnd,imits.........................................................................../' User "n&uts DS1lots............................................................................................../* Data De&endency...................................................................................................../2 So$t%are Run 1hase "".........................................................................................../+ So$t-unctions User"n&uts...................................................................................../. So$t-unctions Guard)ands.................................................................................../3 So$t-unctions Data1lots......................................................................................./3 1lot4ditor ool.........................................................................................................#0 Re5ision History.......................................................................................................#/ Re$erences.................................................................................................................#/

"ntroduction
6GRRStudy7 is a Microso$t 4!cel )ased utility $or 8au8e re&eata)ility and re&roduci)ility (GRR) analysis. he tool ser5es multi&le $unctions9 "t can )e used to calculate 8uard)ands $or the de5ice test limits "t can )e used to 8enerate the &lots re:uired $or &roduct characteri;ation re5ie% (1(R) "t can )e used to 8enerate datasheet &lots his manual %ill assist the user in understandin8 the ste&s and &rocedures need to )e $ollo%ed %hile usin8 this utility.

So$t%are Architecture
-i8ure / illustrates the architecture o$ 6GRRStudy7 so$t%are. he so$t%are &rocesses tem&erature data< ca&a)ility data (also called as re&roduci)ility data)< re&eata)ility data and user in&uts. A$ter &rocessin8 the in&uts< the so$t%are 8enerates Guard)ands< 1(R 1lots and Datasheet 1lots.

em&erature Data (a&a)ility Data Re&eata)ility Data So$t%are

Guard)ands 1(R 1lots Datasheet 1lots

User "n&uts
-i8ure / So$t%are Architecture

em&erature Data
em&erature data is collected to see ho% de5ice &er$orms %ith tem&erature 5ariation. he &rocedure to collect tem&erature data is as $ollo%s9 Take N number of devices. Test t ese devices at a tem!erature T1 de"ree #elsius on a "iven tester. $t t e end, t e results for all t e devices s ould be stored in a file. Re!eat t e above % ste!s to collect t e data at ot er tem!eratures &T2, T% ' etc.( for t e same N devices and on t e same tester. T us, t e results are available in multi!le files, one corres!ondin" to eac tem!erature. he $ormat o$ the $iles %ill usually )e 6.std$7 $ile $ormat. =ut< the 6GRRStudy7 so$t%are re:uires that the in&ut $iles should )e in 6.!ls7 $ile $ormat. here are $e% tools a5aila)le< like 6 a)%in7< %hich may )e used to con5ert 6.std$7 $iles to 6.!ls7 $iles. 45en a$ter con5ersion to 6.!ls7 $ormat the data in the $ile may need some &ost>&rocessin8 to )e com&ati)le %ith 6GRRStudy7 so$t%ares data $ormat. he 6GRRStudy7 so$t%are su&&orts a ma!imum o$ ? tem&eratures< namely /< #< ' and ?. he data corres&ondin8 to these tem&eratures should )e made a5aila)le to the so$t%are in /.!ls< #.!ls< '.!ls and ?.!ls res&ecti5ely. -i8ure # illustrates the $ormat o$ a /.!ls $ile. he same $ormat is a&&lica)le to other tem&erature data $iles.

-i8ure # -ormat o$ /.!ls -ile

Rules 8o5ernin8 the $ormat o$ tem&erature data $iles are stated )elo%. 45ery user o$ this so$t%are should a)ide to these rules. -e% rules stated here are de$ined %ith 6 /.!ls7 as an e!am&le. hese rules are also a&&lica)le to other tem&erature data $iles. /. he name o$ the $ile must )e an element o$ the set @ /.!ls< #.!ls< '.!ls< ?.!lsA #. Data in /.!ls must corres&ond to the data taken at tem&erature / '. he %ork)ook /.!ls must ha5e a %orksheet 6Sheet/7 and the data must )e stored in the %orksheet 6Sheet/7 ?. "$ there are any other %orksheets in the %ork)ook< those %ill )e i8nored *. he data must start at the cell %hose coordinates are (Ro% /< (olumn /) 2. he data must )e in the $orm o$ a t%o>dimensional array %ith the ro%s corres&ondin8 to test la)els (test la)els %ill )e discussed later in this manual) and the columns corres&ondin8 to de5ice num)ers +. -irst column must contain the test la)els startin8 $rom ro% # on%ards .. -irst ro% must contain de5ice num)ers startin8 $rom column # on%ards 3. he de5ice num)ers must start $rom / and must increment only )y / as %e &roceed to ne!t hi8her column /0. he num)er o$ de5ices in /.!ls and all other tem&erature data $iles must )e the same //. De5ice num)er / in $ile /.!ls must corres&ond to de5ice num)er / in all other tem&erature data $iles and so also the other de5ice num)ers /#. ,et ! B num)er o$ de5ices and y B num)er o$ test la)els< then there must )e a total o$ C(! D /) E (y D /)F num)er o$ cells in the t%o dimensional array. 4ach o$ these cells it may )e de5ice num)er or test la)el or data &oint must )e $illed. he data should )e conti8uous and none o$ these cells should )e le$t )lank. /'. here should )e no other data any%here in the %orksheet 6Sheet/7 o$ %ork)ook /.!ls e!ce&t the one mentioned in rule /# a)o5e.

(a&a)ility Data
(a&a)ility data is collected to see %hether the de5ice &er$ormance is re&roduci)le i$ tested )y di$$erent testers. he &rocedure to collect ca&a)ility data is as $ollo%s9 Take N number of devices. Test t ese devices on a tester #1 at a "iven tem!erature. &T e device !arameters are usually tested only at t e test conditions for ) ic "uardbands need to be calculated.( $t t e end, t e results for all t e devices s ould be stored in a file. Re!eat t e above % ste!s to collect t e data on ot er testers &#2, #% ' etc.( for t e same N devices and at t e same tem!erature. T us, t e results are available in multi!le files, one corres!ondin" to eac tester. 1lease note that the de5ices and the num)er o$ de5ices used $or ca&a)ility study may )e totally di$$erent $rom the de5ices and the num)er o$ de5ices used $or collectin8 tem&erature data. he 6GRRStudy7 so$t%are su&&orts a ma!imum o$ 2 test setu&s (testers)< namely (/< (#< ('< (?< (* and (2. he data corres&ondin8 to these test setu&s should )e made a5aila)le to the so$t%are in $iles (/.!ls< (#.!ls< ('.!ls< (?.!ls< (*.!ls and (2.!ls res&ecti5ely. he minimum re:uirement is ca&a)ility data collected on at least t%o testers i.e. (/.!ls and (#.!ls. hou8h other $iles are o&tional< it is )etter to collect data on as many test setu&s as &ossi)le to 8et relia)le results. (ommon &ractice is to collect data on $our test setu&s. he $ormat o$ (/.!ls is same as /.!ls (see $i8ure #). he same $ormat is a&&lica)le to other ca&a)ility data $iles. A similar set o$ rules 8o5ernin8 the $ormat o$ ca&a)ility data $iles are stated )elo%. -e% rules stated here are de$ined %ith 6(/.!ls7 as an e!am&le. hese rules are also a&&lica)le to other ca&a)ility data $iles. /. #. '. ?. *. 2. he name o$ the $ile must )e an element o$ the set @(/.!ls< (#<!ls< ('.!ls< (?.!ls< (*.!ls< (2.!lsA Data in (/.!ls must corres&ond to the data collected on tester (/ he %ork)ook (/.!ls must ha5e a %orksheet 6Sheet/7 and the data must )e stored in the %orksheet 6Sheet/7 "$ there are any other %orksheets in the %ork)ook< those %ill )e i8nored. he data must start at the cell %hose coordinates are (Ro% /< (olumn /). he data must )e in the $orm o$ a t%o>dimensional array %ith the ro%s corres&ondin8 to test la)els (test la)els %ill )e discussed later in this manual) and the columns corres&ondin8 to de5ice num)ers

+. -irst column must contain the test la)els startin8 $rom ro% # on%ards .. -irst ro% must contain de5ice num)ers startin8 $rom column # on%ards 3. he de5ice num)ers must start $rom / and must increment only )y / as %e &roceed to ne!t hi8her column /0. he num)er o$ de5ices in (/.!ls and all other ca&a)ility data $iles must )e the same //. De5ice num)er / in $ile (/.!ls must corres&ond to de5ice num)er / in all other ca&a)ility data $iles and so also the other de5ice num)ers /#. ,et ! B num)er o$ de5ices and y B num)er o$ test la)els< then there must )e a total o$ C(! D /) E (y D /)F num)er o$ cells in the t%o dimensional array. 4ach o$ these cells it may )e de5ice num)er or test la)el or data &oint must )e $illed. he data should )e conti8uous and none o$ these cells should )e le$t )lank. /'. here should )e no other data any%here in the %orksheet 6Sheet/7 o$ %ork)ook (/.!ls e!ce&t the one mentioned in rule /# a)o5e.

Re&eata)ility Data
Re&eata)ility data is collected to see %hether the de5ice &er$ormance can )e re&eatedly achie5ed on a 8i5en tester. he &rocedure to collect re&eata)ility data is as $ollo%s9 Take a sin"le device. Test t e device re!eatedly on a "iven tester at a "iven tem!erature. &T e device !arameters are usually tested only at t e test conditions for ) ic "uardbands need to be calculated.( $t t e end, t e results for all t e iterations )ill be stored in a file. he 6GRRStudy7 so$t%are su&&orts re&eata)ility data $or only one de5ice. he data collected iterati5ely on this sin8le de5ice should )e made a5aila)le to the so$t%are in R/.!ls. he $ormat o$ R/.!ls looks similar to that o$ /.!ls sho%n in -i8ure #. he di$$erence here is only that the num)ers in $irst ro% o$ each column (e!ce&t $irst column) re&resent iteration num)er instead o$ de5ice num)er. he set o$ rules a&&lica)le to R/.!ls are mentioned as $ollo%s9 /. #. he name o$ the $ile must )e R/.!ls he %ork)ook R/.!ls must ha5e a %orksheet 6Sheet/7 and the data must )e stored in the %orksheet 6Sheet/7. '. "$ there are any other %orksheets in the %ork)ook< those %ill )e i8nored. ?. he data must start at the cell %hose coordinates are (Ro% /< (olumn /). *. he data must )e in the $orm o$ a t%o>dimensional array %ith the ro%s corres&ondin8 to test la)els (test la)els %ill )e discussed later in this manual) and the columns corres&ondin8 to de5ice num)ers. 2. -irst column must contain the test la)els startin8 $rom ro% # on%ards. +. -irst ro% must contain iteration num)ers startin8 $rom column # on%ards .. he iteration num)ers must start $rom / and must increment only )y / as %e &roceed to ne!t hi8her column 3. ,et ! B num)er o$ de5ices and y B num)er o$ test la)els< then there must )e a total o$ C(! D /) E (y D /)F num)er o$ cells in the t%o dimensional array. 4ach o$ these cells it may )e de5ice num)er or test la)el or data &oint must )e $illed. he data should )e conti8uous and none o$ these cells should )e le$t )lank. /0. here should )e no other data any%here in the %orksheet 6Sheet/7 o$ %ork)ook R/.!ls e!ce&t the one mentioned in rule 3 a)o5e.

est ,a)el -ormat


A test la)el in a test &ro8ram is used to associate a 8i5en data &oint %ith its &arameter. 4!tra in$ormation in the test la)els can also )e used to descri)e the test conditions at %hich the data %as collected. he rules that 8o5ern the test la)el $ormat $or 6GRRStudy7 so$t%are are s&eci$ied as $ollo%s. 45ery user o$ this so$t%are should )e a%are o$ this la)elin8 $ormat )e$ore %ritin8 the test &ro8ram and )e$ore actually startin8 data collection. /. 4ach test la)el can )e decoded )y a 8eneric 6-ormat Strin87 9 -"4,D/G-"4,D#G...G-"4,DHG1ARAM4 4R a. 4!am&le9 "$ AIDDG=IDDGIR4-G1ARAM4 4R is the $ormat strin8 and *.0G'.'G?.032GSHR is a test la)el< then the test la)el can )e decoded as 6SHR s&eci$ied at AIDD B *.0< =IDD B '.' and IR4- B ?.0327 ). Hote that it is still not kno%n that %hat AIDD is or %hat SHR is or %hat the unit o$ SHR is. All these thin8s %ill )e s&eci$ied )y user in&uts %hich %ill )e discussed later in this manual. c. Here AIDD is an e!am&le o$ 6-ield Hame7 %hile *.0 is an e!am&le o$ 6-ield Ialue7. #. he last $ield in the $ormat strin8 must )e 1ARAM4 4R. ("$ there is only one $ield in the $ormat strin8 it must )e 1ARAM4 4R.) '. -"4,D/ J -"4,D# J K J -"4,DH J 1ARAM4 4R i.e. all the $ield names in the $ormat strin8 must )e distinct. ?. An underscore si8n (6G7) must )e used to se&arate the adLacent $ields in the $ormat strin8 (and test la)els). *. here must )e only H underscores in a $ormat strin8 (and test la)els) containin8 (H D /) $ields. 2. Underscore si8n (6G7) should not )e a &art o$ $ield name or $ield 5alue. +. Mrder o$ the $ields must al%ays )e &reser5ed %hile %ritin8 the test la)els. a. 4!am&le9 ,et AIDDG=IDDGIR4-G1ARAM4 4R is the intended $ormat strin8. Su&&ose SHR needs to )e measured at AIDD B *.0< =IDD B '.' and IR4- B ?.032 then test la)el must )e %ritten as *.0G'.'G?.032GSHR. Any other in5alid com)inations like ?.032G*.0G'.'GSHR or '.'G?.032G*.0GSHR %ill lead to %ron8 results or so$t%are mal$unction. .. he num)er o$ $ields in the $ormat strin8 should )e e:ual to the ma!imum num)er o$ distinct 5aria)les (e!am&le9 AIDD< =IDD< etc.) across %hich the de5ice &arameters %ill )e s%e&t &lus one (the 1ARAM4 4R $ield). a. 4!am&le9 Su&&ose SHR needs to )e measured at AIDD B @?.*< *.0< *.*A and IR4- B @?.0< *.0AN then the $ormat strin8 should )e AIDDGIR4-G1ARAM4 4R. Also< su&&ose ",eaka8e $or di8ital in&ut &ins needs to )e tested at =IDD B @'.'< *.0AN then the $ormat strin8 $or ",eaka8e test could sim&ly )e =IDDG1ARAM4 4R.

Since no% there are total ' 5aria)les across %hich the de5ice &arameters %ill )e s%e&t< the $ormat strin8 should )e %ritten as =IDDGAIDDGIR4-G1ARAM4 4R (any other com)ination o$ $irst three $ields is also acce&ta)le) ). "n the e!am&le a)o5e %e see that $or ",eaka8e measurement AIDD and IR4- $ields are irrele5ant )ut still $or the sake o$ decodin8 the test la)els %e %ill ha5e to &ut some 6dummy7 5alues $or AIDD and IR4- $ields. -or e!am&le %e can &ut AIDD B 0.0 and IR4- B 0.0 in =IDDGAIDDGIR4-G1ARAM4 4R $ormat strin8 and %rite the test la)els as '.'G0.0G0.0G",eaka8e and *.0G0.0G0.0G",eaka8e $or the t%o test conditions o$ =IDD B @'.'< *.0A. he dummy 5alues are irrele5ant and can )e any 5alid 5alue $or that $ield. 3. Oey%ords should not )e a &art o$ $ormat strin8. P 4M14RA UR4< PD4I"(4< P 4S S4 U1< P" 4RA "MH< PIA,U4 and PDA A are key%ords and should not )e used as $ieldnames. /0. P1ARAM4 4R is also a key%ord and can a&&ear only as the last $ield in the $ormat strin8. //. Units should not )e a &art o$ the test la)el. a. 4!am&le9 -or AIDDG=IDDG1ARAM4 4R< *.0IG'.'IGSHR is not a 5alid test la)el< instead %rite *.0G'.'GSHR. ( he Units %ill )e taken care o$ later %hen user in&uts are &ro5ided.)

So$t%are Setu&
Mnce all the in&ut data is collected in res&ecti5e $iles in the &ro&er data $ormats discussed a)o5e< user can &roceed %ith the so$t%are setu&. M&eratin8 System Re:uirements9 Microso$t Qindo%s R1 1ro$essional (Iersion #00# or )etter) So$t%are Re:uirements9 Microso$t 4!cel (Iersion #00# or )etter) he ste&s to setu& 6GRRStudy7 so$t%are are as $ollo%s9 /. (reate a ne% $older. ( he name o$ the $older and the &ath to its location is not im&ortant. his $older %ill )e re$erred to as 6current %orkin8 directory7 throu8hout this manual.) #. his $older should include $ollo%in8 $iles9 a. he main so$t%are $ile9 GRRStudy.!ls ). he em&erature Data $iles9 @ /.!ls< #.!ls< '.!ls< ?.!lsA c. he (a&a)ility Data $iles9 @(/.!ls< (#.!ls< ('.!ls< (?.!ls< (*.!ls< (2.!lsA d. he Re&eata)ility Data $ile9 R/.!ls hats allS he so$t%are is ready to run no%S 1lease note that< out o$ all the data $iles listed a)o5e< $e% $iles may )e o&tional< de&endin8 on %hat out&ut results are intended to )e 8enerated $rom the so$t%are. 1lease re$er to data de&endency ta)le (discussed later in this manual). he 6GRRStudy7 so$t%are runs in t%o &hases 1hase " and 1hase "".

So$t%are Run 1hase "


o run this utility kee& the macro security o$ Microso$t 4!cel so$t%are as 6Medium7 C ools T Macro T SecurityF. (lose all runnin8 Microso$t 4!cel a&&lications. M&en the 6GRRStudy.!ls7 $ile. (A$ter o&enin8 this $ile< i$ &rom&ted to ena)le macros or to chan8e macro security le5el< should )e done so.) Mnce the 6GRRStudy.!ls7 $ile is o&ened< a 8ra&hical user inter$ace (GU") namely 6GRRStudy7 %ill &o& u& (see $i8ure ').

-i8ure ' Gra&hical User "nter$ace So< lets 8et $amiliar %ith the 6GRRStudy7 GU" (&lease re$er $i8ure '). /. his check)o! should )e checked only i$ tem&erature data ( /.!ls< #.!ls K etc.) is a5aila)le. (heckin8 this check)o! %ill ena)le #. #. Select the num)er o$ test tem&eratures $or %hich data is a5aila)le. he minimum re:uirement is that data $or at least one tem&erature should )e a5aila)le. Data $or a ma!imum o$ $our tem&eratures can )e &rocessed.

'. Mnce the num)er o$ test tem&eratures is selected< the corres&ondin8 num)er o$ te!t)o!es %ill 8et ena)led. "n te!t)o! /< tem&erature (in de8ree centi8rade) at %hich data in $ile /.!ls %as collected should )e %ritten. Same lo8ic a&&lies $or other te!t)o!es. ?. his check )o! should )e checked only i$ ca&a)ility data ((/.!ls< (#.!ls K etc.) is a5aila)le. (heckin8 this check)o! %ill ena)le *. *. Select the num)er o$ test setu&s used $or ca&a)ility study. he minimum re:uirement is that data collected on at least t%o test setu&s (testers) should )e a5aila)le. Data $or a ma!imum o$ si! test setu&s can )e &rocessed. Usual &ractice is to collect data on $our test setu&s. 2. his check)o! should )e checked only i$ re&eata)ility data is a5aila)le in R/.!ls. +. =y de$ault the num)er o$ de5ices used $or re&eata)ility is set to one. .. -ormat strin8 should )e entered here. his %ill )e used to decode the test la)els in the in&ut data to retrie5e in$ormation. 3. Mnce all the in$ormation discussed a)o5e is $illed>in< click this 6H4R 7 )utton. he so$t%are %ill $irst check $or 5alidity o$ the in&uts &ro5ided to it and then %ill &roceed accordin8ly. Also< the GU" %ill disa&&ear. /0. User may de$er to do any o&eration )y clickin8 this 6(ancel7 )utton. //. "$ the so$t%are $inds any errors a$ter 6H4R 7 )utton is clicked< those %ill )e dis&layed here and &ro&er 8uidance %ill )e 8i5en to correct those. "$ the in&uts &ro5ided )y user and the in&ut data $iles are 5alid< the so$t%are %ill run and com&lete its &hase ". At the end o$ &hase "< user needs to &ro5ide $e% more in&uts to the so$t%are. he in&uts are &ro5ided )y $illin8>in t%o %orksheets< namely< 6Sym)olsAnd,imits7 and 6DS1lots7.

User "n&uts Sym)olsAnd,imits


At the end o$ 1hase "< the so$t%are %ill land on the sheet 6Sym)olsAnd,imits7 (see $i8ure ?).

-i8ure ? Sheet 6Sym)olsAnd,imits7 (=lank) -irst column o$ the sheet 6Sym)olsAnd,imits7 contains test &arameter names deri5ed $rom the test la)els a5aila)le in the tem&erature data. he $irst column also contains the s%ee& 5aria)les across %hich tem&erature data is s%e&t and are deri5ed $rom the $ormat strin8 &ro5ided )y the user. Mther columns ask $or the in$ormation associated %ith these &arameters. User should $ill>in this sheet a&&ro&riately )e$ore heUshe &roceeds $urther. -i8ure * illustrates e!am&le o$ a $illed>in 6Sym)olsAnd,imits7 %orksheet. he $i8ure * also illustrates ho% the in$ormation &ro5ided in a &roduct datasheet (ADS.'/3 in this case) can )e used to $ill>in this %orksheet.

-i8ure * -illin8 6Sym)olsAnd,imits7 Qorksheet Usin8 1roduct Datasheet 45ery datasheet %ill ha5e datasheet sym)ol $or almost e5ery &arameter mentioned in it. "$ the datasheet sym)ol is not s&eci$ied< $or e!am&le Gain 4rror in this case< &lease &ut an asterisk si8n 6V7 in the corres&ondin8 cell (see $i8ure *). Sym)ol Descri&tion is (MM1U,SMRW $or e5ery &arameter. he asterisk si8n 6V7 also stands $or 6Hot A&&lica)le7. -or e!am&le< the &arameter SHR has only M"H>limit (3#d= in this case) and MAR>limit is not a&&lica)le to this &arameter at all. his is indicated )y &uttin8 asterisk si8n 6V7 in the 6MAR ,imit7 column $or SHR &arameter (see $i8ure *). DM HM lea5e any in&ut cell )lank. Use 6V7 si8n as and %hen necessary. "n the 6@ 4M14RA UR47 column user should s&eci$y only the 6HMM"HA, em&erature7 at %hich 1(R &lots should )e 8enerated. Data in this column is used only $or 8eneratin8 1(R &lots. he $ormat strin8 $or this &articular e!am&le ($i8ure *) is AIDDGIR4-G1ARAM4 4R. he 6@ AIDD7 and 6@ IR4-7 columns a&&ear in the %orksheet only )ecause the &arameters %ere tested )y s%ee&in8 across these 5aria)les and they %ere mentioned in the $ormat strin8. hat means there %ill )e as many columns a$ter 6@ 4M14RA UR47 column as the num)er o$ s%ee& 5aria)les. he 5alues s&eci$ied in the s%ee& 5aria)le columns (6@ AIDD7 and 6@ IR4-7< in this case) are the 5alues at %hich the test &arameters are s&eci$ied in the datasheet. -or e!am&le< in ADS.'/3 datasheet SHR is s&eci$ied at AIDD B

*I and IR4- B *I< so in the ro% corres&ondin8 to &arameter SHR< cell in column 6@ AIDD7 %ill ha5e 6*7 %ritten in it and the cell in column 6@ IR4-7 %ill ha5e 6*7 %ritten in it (see $i8ure *). he test &arameter 8uardands $or the de5ice %ill )e calculated only at the test conditions s&eci$ied in the s%ee& 5aria)le columns (6@ AIDD7 and 6@ IR4-7< in this case) %hich a&&arently means only at the test conditions s&eci$ied in the de5ice datasheet. At this &oint it is also im&ortant to kno% ho% 1(R &lots %ill )e 8enerated. "n $i8ure *< the SHR o$ ADS.'/3 is s&eci$ied 6@ AIDD B *7 and 6@ IR4- B *7. hou8h the tem&erature ran8e $or SHR s&eci$ication is 6>?0 o( to D.* o(7< still a nominal tem&erature e:ual to #* is s&eci$ied in 6@ 4M14RA UR47 column. he num)er o$ 1(R &lots 8enerated &er &arameter %ill )e e:ual to the num)er o$ s%ee& 5aria)les across %hich the &arameter %as s%e&t. So in this case %e %ill ha5e ' 1(R &lots namely 6SHR across 4M14RA UR47< 6SHR across AIDD7 and 6SHR across IR4-7. -or &lot 6SHR across AIDD7 other t%o 5aria)les %ill )e ke&t constant at s&eci$ied 5alues 4M14RA UR4 B #* and IR4- B *. Hote that $or 1(R &lots R>a!is re&resents 1ARAM4 4R and W>a!is re&resents D4I"(4 Hum)er.

User "n&uts DS1lots


-i8ure 2 illustrates an e!am&le o$ )lank and $illed>in 6DS1lots7 %orksheet. he in&uts &ro5ided throu8h this %orksheet are used only $or &lottin8 Datasheet 1lots.

-i8ure 2 =lank and -illed>in 6DS1lots7 Qorksheet he $irst column o$ the 6DS1lots7 %orksheet asks $or $i8ure num)er (as a &ractice< these $i8ure num)ers should )e ke&t as the user %ants them to a&&ear in

the datasheet). "n the ne!t column (61ARAM4 4R7) should contain the &arameter that needs to )e &lot. he &arameter name should corres&ond to the &arameter name %hich a&&ears in the test &ro8rams test la)els and not the one in datasheet. All the other columns stand $or s%ee& 5aria)les. An asterisk si8n 6V7 should )e %ritten in the column o$ s%ee& 5aria)le 65ersus7 %hich the &arameter needs to )e &lot. here must )e one and only one asterisk si8n 6V7 in each ro%. he num)er o$ &lots 8enerated %ill corres&ond to num)er o$ ro%s in the %orksheet. Hone o$ the re:uired cells should )e le$t )lank. -or e!am&le (see $i8ure 2)< 6-i8ure #7 %ill )e a &lot o$ 6Si8nal>to>Hoise Ratio 5ersus -ree>Air em&erature7 at AIDD B *I and IR4- B *I. Hote that the units< sym)ols and sym)ol descri&tions re:uired $or datasheet &lots %ill )e retrie5ed $rom the 6Sym)olsAnd,imits7 %orksheet.

Data De&endency
Since< no% the user is $amiliar %ith in&ut data $iles and user in&uts< its time to kno% the de&endency o$ out&uts on in&uts. he data de&endency ta)le ( a)le /) e!&lains %hat the minimum in&ut re:uirement is $or each out&ut. Mut&ut Guard)and em&erature Data Re:uired /.!ls > (om&ulsory Re:uired /.!ls > (om&ulsory Re:uired /.!ls > (om&ulsory a)le / Data De&endency a)le (a&a)ility Data Re:uired (/.!ls< (#.!ls > (om&ulsory M&tional "n&ut Re&eata)ility Data Re:uired User "n&uts Re:uired

1(R 1lots

R/.!ls Sym)olsAnd,imits > (om&ulsory > (om&ulsory M&tional Re:uired Sym)olsAnd,imits > (om&ulsory Re:uired Sym)olsAnd,imits < DS1lots > (om&ulsory

Datasheet 1lots

M&tional

M&tional

So$t%are Run 1hase ""


A$ter $illin8>in 6Sym)olsAnd,imits7 and 6DS1lots7 %orksheets 8enerated at the end o$ 1hase "< user is no% ready to 8o throu8h 1hase "" o$ this so$t%are. =e$ore startin8 any $urther &rocessin8< user can al%ays ha5e a look at the data 8enerated at the end o$ &hase ". he charts 6 (hart7< 6((hart7 and 6R(hart7 re&resent a 8ra&hical 5ie% o$ tem&erature data< ca&a)ility data and re&eata)ility data res&ecti5ely. All these charts are 1i5ot (harts and 8i5e $le!i)ility to 8o throu8h all the data at once. ,ookin8 at these charts< user may scrutini;e the data 5isually and may discard $e% de5ices (i$ the user thinks those are outliers) or add discarded ones (see $i8ure +).

-i8ure + Scrutini;e Data Usin8 1i5ot (harts Mnce data ins&ection is com&lete< user may &roceed %ith $urther data &rocessin8. 1R4SS 6(trl D m7S hats allS

1ressin8 6(trl D m7 %ill &o&>out 6So$t-unctions7 menu %hich allo%s access to all the $unctions &ro5ided )y the 6GRRStudy7 so$t%are (see $i8ure .).

-i8ure . So$t-unctions Menu Here on%ards< 6So$t-unctions7 %ill )e discussed in a ste& )y ste& manner.

So$t-unctions User"n&uts
4!&ort User in&uts &ro5ided at the end o$ 1hase " throu8h 6Sym)olsAnd,imits7 and 6DS1lots7 %orksheets can )e stored $or $uture use. Qhen 64!&ort7 )utton is &ressed< the data $rom 6Sym)olsAnd,imits7 and 6DS1lots7 %orksheets o$ 6GRRStudy.!ls7 $ile is stored in 6User"n&uts.!ls7 $ile in the current %orkin8 directory. his e!&orted $ile can later )e im&orted to sa5e time $illin8>in 6Sym)olsAnd,imits7 and 6DS1lots7 %hen user %ants to &rocess data collected $or ne!t re5ision o$ de5ices or Lust ne!t characteri;ation data. "m&ort "$ the user has a $ile 6User"n&uts.!ls7 8enerated usin8 6GRRStudy7 so$t%are in some &ast run< then the $ile can )e im&orted to $ill in data in 6Sym)olsAnd,imits7 and 6DS1lots7 %orksheets o$ 6GRRStudy.!ls7 $ile. o do this< user should co&y the old 6User"n&uts.!ls7 $ile in current %orkin8 directory< )e$ore usin8 the 6"m&ort7 $unction. Qhen 6"m&ort7 )utton is &ressed< only the 5alid data $rom 6User"n&uts.!ls7 $ile in the current %orkin8 directory is im&orted into

6Sym)olsAnd,imits7 and 6DS1lots7 %orksheets o$ 6GRRStudy.!ls7 $ile. (han8es in test &ro8ram< datasheet s&eci$ications or s%ee& 5aria)les $rom old re5ision o$ de5ices to current re5ision may render the im&orted data in5alid. Hence< the im&orted data must )e care$ully ins&ected )y the user< )e$ore startin8 any $urther data analysis.

So$t-unctions Guard)ands
(hDe5 (alculate Guard)ands Mnce the data in &i5ot charts ( (hart< ((hart< R(hart) has )een scrutini;ed and de5ices ha5e )een chan8ed (discardin8 outliers or addin8 discarded ones etc.) &ress the )utton 6(hDe5 (alculate Guard)ands7 to calculate 8uard)ands. 45en i$ no chan8es in de5ices %ere done< same )utton must )e used to calculate 8uard)ands. A$ter &ressin8 the 6(hDe5 (alculate Guard)ands7 the so$t%are %ill $ind out %orst case tem&erature and test tem&erature on its o%n and %ill calculate the 8uard)ands accordin8ly. he results %ill )e stored in 6Results7 %orksheet o$ the 6GRRStudy.!ls7 $ile. ,ookin8 at the 6Results7 sheet< user may %ant to chan8e the de5ices and calculate 8uard)ands once a8ain. "$ so< in5oke 6So$t-unctions7 menu )y &ressin8 6(trl D m7 and then &ress 6(hDe5 (alculate Guard)ands7 )utton. his &rocess can )e re&eated any num)er o$ times. (h em& (alculate Guard)ands ,ookin8 at the 6Results7 sheet< user may also %ant to chan8e the test tem&erature or %orst tem&erature and calculate 8uard)ands once a8ain. "$ so< in5oke 6So$t-unctions7 menu )y &ressin8 6(trl D m7 and then &ress 6(h em& (alculate Guard)ands7 )utton. his &rocess also can )e re&eated any num)er o$ times.

So$t-unctions Data1lots
1(R 1lots Qhen this 61(R 1lots7 )utton is &ressed< 1(R &lots are 8enerated and stored in 61(R1lots.!ls7 $ile in the current %orkin8 directory. Any e!istin8 61(R1lots.!ls7 $ile in the directory %ill )e o5er%ritten. Datasheet 1lots

Qhen this 6Datasheet 1lots7 )utton is &ressed< datasheet &lots are 8enerated and stored in 6Datasheet1lots.!ls7 $ile in the current %orkin8 directory. Any e!istin8 6Datasheet1lots.!ls7 $ile in the directory %ill )e o5er%ritten.

1lot4ditor ool
1lot4ditor is a tool meant $or t%eakin8 &lots stored in 6Datasheet1lots.!ls7 8enerated usin8 6GRRStudy7 so$t%are. o use this tool 61lot4ditor.!ls7 $ile must )e &resent in the current %orkin8 directory. his $ile is a standalone $ile and is neither 8enerated )y 6GRRStudy7 so$t%are nor is re:uired $or normal o&eration o$ 6GRRStudy7 so$t%are. "t Lust comes as a utility alon8 %ith 6GRRStudy7 so$t%are. o start the utility9 (lose all the currently runnin8 Microso$t 4!cel a&&lications. M&en 61lot4ditor.!ls7 $ile. (4na)le macros i$ &rom&ted to do so.) M&en 6Datasheet1lots.!ls7 $ile. (lick on any &lot in any sheet o$ 6Datasheet1lots.!ls7 %hich needs to )e edited. 1ress 6(trl D m7 to in5oke the 1lot4ditor menu.

-i8ure 3 1lot4ditor Menu -i8ure 3 illustrates the 1lot4ditor Menu. he menu 8i5es access to most commonly needed &arameters o$ the &lot in a sin8le &alette< %hich sa5es a lot o$ time.

he source data $or the &lots stored in each %orksheet o$ 6Datasheet1lots.!ls7 is also stored in the corres&ondin8 %orksheets startin8 at location (Ro% /< (olumn /). he source data can also )e edited.

Re5ision History
Re5ision / Xuly #+< #003

Re$erences
htt&9UU%%%/.msl&.ti.comUh&aGn&dUdocU estG(a&a)ilityGStandardGYGR e$erenceGGuide.doc

Vous aimerez peut-être aussi