MASSACHUSETTS INSTITUTE OF TECHNOLOGY
Depart,rnent of Electrical Engineering arlrl Cornpllt,er Science
Spring Terrr~2007
Maglev Design Prot)lrrn
Tlle ~ I I I I I O S Pnf t,l~islt~l)is tn l111ilc1and improve a srrlall rrlagrlet,iclevit,ation (rnaglev) system. Yo11 will t)e
~:
(m11i~:hyou can keep after checkoff). .ift,er l~~lilcling
t,lle l~asicsystern,
given a. kit t,o l)uil(i t l ~ el ~ a s i systerrl
you will evaluate the rlyr~arr~ics
of t,lle syst,ern and design a cornpensator for it. You are e ~ ~ c o l ~ r a gtoe (also
l
improve otller aspects of t,lle syst,ern, sll<:has t l ~ esensor, the elr<:trornagnet,t,lle elect,ronics, arlrl t,lle levit,aterl
ollject. This lat) is an o ~ ~ e r ~ - e n ~design
l e r l project,. so t)e creat,ive!
Basic System Construction
The s t , n ~ < : t ~ feat,ures
~ral
of the t~asicmaglev systerrl are she\%-n in Figure 1. Tlle schernati<:of t l ~ edrive
in Figure 2. Tlle kit yo11 will receive i11c111desall of the part,^ s110rnr1 i11 t,lle s<:herr~ati(:
elect,ror~icsis sllo~vr~
(and list,ecl in Figure 3). t111t does not inchlcle t,lle moorlen starlrl or t l ~ eol)je<:tfor levitatior~(you mill l ~ a v e
t,o prnvide your own). For illis lal). yo^^ will neecl t,o t)~lilrlthe l ~ a s i syst,ern
~:
as sllorvr~in the figures.
This kit is t)ase(i on a lorr-<:(>strlesign ~levelope~l
t)y Guy \1arxler1 [l,21. In t l ~ el~asicsyst,ern, t,lle position
of the levitate(i ol)je<:tis sensecl t)y t l ~ eSS495 Hall-effect sensor. The o ~ ~ t voltage
p ~ ~ tof t l ~ eserlsor rlrives t l ~ e
input of the \IIC502 f t ~ n - r n a ~ ~ a g e r nchip.
e ~ ~ t ,The fan-rnanagernerIt chip pro(ill(:es a p~~lse-rvi(ith
rno(illlater1
(PtVhl) drive sigrlal t,o the L\ID18201 H-l~riclgechip. This PtV\I sigrlal arlj~lstst,lle average cllrrerlt in the
solenoi(i. \%-hichcor~trolsthe rnagneti<:fielrl.
Please use care in t,lle <:onstr~~ct,ion
of yollr syst,ern. Rear1 t l ~ erlat,a slleets for each IC in the scllerr~at,icto
ft~rniliarizeyo~lrself\%-it,h
t l ~ t)eht~vior
e
arlrl operation of t,lle <:ir<:~~it.
tVhile the EECS Department and Nat,ional
Sernicon(il~<:tor
ht~ve11ee11gerlerous in r n a k i ~ ~these
g
t,o you, Ire carlrlot afforrl t,o rlor~atelarge
kits t~vailal~le
W ~ I Oreceives a kit to complete t l ~ e
lal]. tVhile me mill
m~rnt)ersof part,s t,o you. 11-e will expect every st~lrler~t
11e tolerar~tof reasona1)le t111t destr~~ct,ive
errnrs, ~ v ehave only a very srrlall s1111111yof spares. D~lilrlcarefillly!
Testing
For t l ~ el~asicsyst,ern t,o work. t,lle polarities of the solenoi(i. t,lle sensor, arlrl t,lle rrlagrlet rrlust all 1)e <:orrect,.
t t h n <:orrect,.you s11o11l(irlot,ice that the solenoi(i repels t,lle rrlagr~et\%-henthe r n a g ~ ~ is
e t too <:lose. You
may ht~veto experiment with t l ~ epolarit,y of the solerloirl rvinrling, t,lle placement of the sensor, arlrl t l ~ e
orier~t,atior~
of t l ~ ern.agnet,.
.in arrlrrlet,er car1 assist in verifying correct operat,ion. The i n p ~ currer~t
~t
sho111(it)e large (1111 t o 0.5 A) if
t,lle rrlagr~etis too far frorr~or t,oo <:loseto t,lle solenoirl. At t,lle operating ~ ~ o i nt,lle
t , <:~lrrent
<:onsllrr~ption
for
t,l~isk it is aro~mcl100 rr1.i.
Tlle correct l~ehavioris also very ser~sitivet,o t l ~ eweight of the levitat,e~lot)ject. The actual lifting
<:apal~ility
of t,lle l~asicsysterrl is very srrlt~ll(at)out half-a-pendl \%-it,h
t,lle r n a g ~ ~ e tUe
) . pre11axr1to experiment.
Compensation
E v t ~ l ~ ~t al t~eerlyr~arr~ics
of t l ~ esystern and design a <:ornpensat,orfor it. To evaluate t,lle t,rar~sier~t
l~ehaviorof
t,n provi(ie ~LII i r q ) ~ sigrlal.
~t
as shn\%-r~
i11 Figure 4.
t,lle irnprove(i system. yo^^ will neecl to modify the (:ir(:~~it
kleas~lrethe st,ep response of t,lle systern to srrlall i n p ~ st,eps.
~t
Due tn the PtVhl drive t,n the ele<:tr~rr~agnet,
serlsnr 111t~yirlchlde a lot nf ripple. f i l l
t,he O I I ~ ~ nf
I I t,lle
~
may wish to filter out illis ripple wit11 a lorn-pass filt,er l~eforet l ~ eos<:illoscope. Note illat this filt,er cloes riot
go insicle t l ~ efee(it)a<:kloop (for o l ) v i o ~ reasons).
~s
1)ut goes l~et\%-een
t,lle serlsor o ~ ~ t and
p ~ the
~ t sc:ope.
Cite as: James Roberge, course materials for 6.302 Feedback Systems, Spring 2007.
MIT OpenCourseWare (http://ocw.mit.edu/), Massachusetts Institute of Technology. Downloaded on [DD Month YYYY].
analog control
and power
electronics
permanent magnet
levitated object /
Figure 1: An assrrnt)le(i rrlagr~rticlevitat,ion systrrn
"SW
"TF
OND
C , =470
p elecholykc
C2 = 1 pF ceramic
C3 = 0.1 pF ceramic
C4 = 0.01 pF ceramic
-
Mount C , and C, close to U3.
"S
Mount C, and C, close to UZ.
OND
OUT
"EEDBACX P A m
"4 ss495.4
Figure 2: Schernati<:of rrlagrlet,ic levit,ation systrrn
2
Cite as: James Roberge, course materials for 6.302 Feedback Systems, Spring 2007.
MIT OpenCourseWare (http://ocw.mit.edu/), Massachusetts Institute of Technology. Downloaded on [DD Month YYYY].
U1
U2
U3
U1
C1
C2
C3
C1
Lh1780S
voltage reg~llator
MIC502
fan-rnanagerner~tIC
LhID18201 rrlotor H-11riclge IC
SS49S.4
Hall-effect serlsor
170 lrF
ele<:trolyticcapacitor
1 ILF
ceramic: capacitor
0.1 lrF
ceramic: capacitor
ceramic: capacitor
0.0111F
~~rervo~lnrl
solenoi(i
soft-steel carriage l~olt,
neo(iyrni~~rn
r r ~ a g ~ ~(2)
ets
l~eatsir~k
for LMD18201
srrlall circuit 11oarcl
Figure 3: List of kit <:ontents
voltage
Compensator
Maglev
Circuit
Electromagnet
t
Sensor
,
position
Figure 1: Block (liagrarrl of rno(iifie(1rrlagrletic levitat,iorl systern
Design a sirnple rlorr~ir~a~~t-~)ole.
lag, or lead corrlpensat,nr t,o stal~ilizet,lle syst,errl and irrlprove the t,rarlsierlt,
response. Design for a 11eak overslloot of less tllarl 20% in the st,ep response.
Modifications
Yn-I are encol~rage(it,o improve otller aspects of t,l~esyste111. hlo(iificat,ions to <:onsi(ieri r ~ c l ~ ~serlsor
r l e selectior~
size
desigrl, the ~:l~r%rr%~:teristics
nf t,lle levitat,e~lol~ject,.and t,l~erlesign
a1111 placement, the elect,rnrr~agr~et
of t,lle power elect,ronics.
For exarnple. orle asl~ectof the design \%-hi<:hcar1 11e irr~proveclis the position sensor. Tile Hall-effect
sensor rletects the st,rerlgtll of the rrlagneti<:fielrl pro(il~(:ed11y the rrlagrlet 011 the levit,ated nl~ject,.Ho~vever,
IIPC>LIISPthe Hall-effect serlsnr is (iirect,ly 11elo\%the elect,rornagnet, it,s O I I ~ ~ will
I I ~ rlepencl not only on t,lle tlle
r~lagrlet1)ut r%lsn011 the strerlgt,h nf ~rlag~leti(:
~listance1)etsveen the serlsnr
t,lle per~rla~lerlt
field generat,ecl
1137 tlle solenoi(i.
Tile lr%rgerr~aglevsyst,errl a~~r%lyze(i
i11 lect,llre uses a photo-(iet,e<:tor t,o sense position 13. 41. In this
approach, t,lle steel t~alllevit,ates t)et,wee~~
a. collirnate(i light sollrce and t,lle photo-(iet,e<:torto prorl~~ce
a
as she\%-n in Figure S. This serlsor scllerrle is sirnple and relial~learlrl can prorl~~ce
a <:lose-to-linear
sl~arlo\%-,
estirrlat,iorl of position. Horrever. it req~lirescareful aligr~rr~er~t
and acl(iitiona1 parts, in<:l~l(iing
a. light 111111)or
LED. a photo-(iet,e<:tor(CdS cell), fo<:~lsing
opt,ics, and assorterl rle<:troni<:s.
~
~
Cite as: James Roberge, course materials for 6.302 Feedback Systems, Spring 2007.
MIT OpenCourseWare (http://ocw.mit.edu/), Massachusetts Institute of Technology. Downloaded on [DD Month YYYY].
r\
light bulb
cardboard tube
collimator
strip
photo-detector
Lab Report
tVrit,e a. report that cornplet,ely docllrnents your clesign. 111 ~ ~ a r t i < : ~ l l a r ,
1. Descril~eyour efforts to evaluat,e, charact,erize. arlrl rnodel the 1)asic: syst,errI. Provide l~orleplots arlrl
trarlsfer fim<:tions,if appropriat,e.
2. Docl~rr~er~t
all cllar~gest,llat yo^^ rnacle to tile l ~ a s i ~systern,
:
s11~:has serlsor type anrl lo<:ation, t l ~ e
elr<:trornagnet <:oil rlesign, the t,ype of levit,ated ollject,. the porrer electronics. arlrl (of course) t l ~ e
cornpensator.
3. Provide cornplet,e schernati<:s of any <:ir<:~~itry
t,llat you <:hanger1in or acldecl to the l ~ a s i ~
design
:
arlrl
rlescril~eits filnction.
4. Descril~et,lle t)eh>~vior
of your irnprovecl syst,ern. Provirle step respnnses, 1)ocle plots, and trarlsfer
filnct,ions.
5. I n < : l ~ ~refererlces
de
to arly l~ooks.articles, d,ltasheets, or \%-el]sit,es illat
rlesign.
yo^^
ilsecl in yollr researcl~arlrl
Y(mr report sho~llcll)e <:ornplete. ~letaile~l,
and neat. Usir~gyour report, a reasonat)le tecllr~iciar~
sl~ol~lrl
l ~ ael ~ l eto recreate your design and r e s ~ l l t , ~ .
Contest
.it checkoff, you rrlust rlerr~or~strate
yollr improved syst,ern to t,lle TAs. .i mnnl~erof systerrl chara<:trrist,ics
nil1 t)e tested and evaluat,ecl. Prizes will t)e t ~ ~ ~ ~ a rfor
d e cthe
l l ~ e s tdesigns in t,lle follorrir~gcat,egories:
Best Transient Response
20% extra credit,
tVidest Dyr~arr~ic:
Range
20% extra credit,
Best Dist~lrl~ance
Rejectior~
20% extra credit,
Heaviest Ol~jectLifter1
10% extra credit,
Lorrest Power C o n s l ~ r n p t i o ~ ~
10% extra credit,
klost .irtist,ic Syst,ern
Tosci's gift certificate
.ill erltrar~t,srrlust 1)e ready for clleckoff startir~ga t 4prn on hIay 11. La11 report,s rrlust l)e cornplete a t
will t)e disq~lt~lified.
t,l~istirrle. Late e~lt,r>~r~t,s
1. Widest Dynamic Range: Largest ~~erioclic
rrloverrlerlt of t l ~ ot)ject,
e
r n e a s ~ ~ rwit11
e ~ l a. nller,
for a sqllare-\%-aveor sine-mt~vei n p ~ ~ t .
2. Best Disturbance Rejection: Largest ratio l~etsreer~
het~viestol)je<:tlevit,aterl to liglltest
ol)je<:tlevitated. ~lsirlgt l ~ esarrle mnnl~erof rnagnet,s.
3. Heaviest Object Lifted: tVeight of ollject, rnaz;irm~rnof orle rnagnet,.
4. Lowest Power Consumption: \Ieas~ned\%-it,harrlrrlet,er on the single 15-\' s ~ ~ p p l y .
5. Most Artistic System: B e a ~ ~ is
t yin the eye of the l~eholder.
Each st,l~<lent
rnay er~ter111)to two cat,egories. Decisior~sof the j~lrlgesare final, and rrlaji t)e arl~itrary.
Cite as: James Roberge, course materials for 6.302 Feedback Systems, Spring 2007.
MIT OpenCourseWare (http://ocw.mit.edu/), Massachusetts Institute of Technology. Downloaded on [DD Month YYYY].
Acknowledgements
Special t,klar~kst,o Kational Serni<:onrl~~ctor
for a generolls (ionat,ion of LhID18201s
References
[I] Guy \Iars<len, .'Levitation!" Kilts arlrl \'olt,s Magazine, vol. 24, no. 9, 1111. 58-61, Septernt~er2003.
dvailal~leorllirle at http: / / w w w . arttec .net/Press/N&V/Levitation.html
[2] K.A. Lilier~kt~rnp
arlrl K.H. Llln(it~erg,"Lorr-<:ost rrlagr~eti<:
levit,ation project kit,s for teaching fee(it~a<:k
syst,err~design." 01114 Airir'r.fcti~~,
Coiitr.<ilCoi~,ff:i'f:ricf!,
Dost,on, 11.i. Jllrle 20, 2004.
Dyr~,arnfcs-
P m t I. New lilrk: 1Viley. 1968.
[3] H.H. 11-oo~lsonand J.R. hlelcher. Eiectr.omf!cli<i~~ic~~,i
111,. 193-200. [A] J.K. Rol~erge.O~if:r~itroii~ii
Air~,piifir'r.,s:Tlieory
P~~ictrcf!.
New York: n51ey. 197:. p11. 214-217.
Cite as: James Roberge, course materials for 6.302 Feedback Systems, Spring 2007.
MIT OpenCourseWare (http://ocw.mit.edu/), Massachusetts Institute of Technology. Downloaded on [DD Month YYYY].