I
`
`second thoughts
`on the
`direct-conversion receiver
`""'lig<lol io """'~ -
`""'..,..."""'
`-
`the)ocaiooc:iletot fvfol - ~-in tho_....
`- ""''-olwhic:hiodl•ii•edby-dr·
`.... L2.C2. Fotlll>~lhevfoio-IOIN
`._ . . . . . . .
`. . . .. fO<CWlhevfo
`-~--Hz•lhellidlallhe ..... _
`_,.IO._.
`TheDC-ionot_for __ ol
`•mpli1udl--lotod ligNila. In - - · •m
`be hMrd by IUring - - -
`_ . , -
`lho
`.,.,,.,_ bul ... ii ~ ..... t .,, .-g.n1 solullan. ~
`confmbe-. RTTYmeybe-. - (cid:173)
`ilil-118d• fok. not•..m.
`The mix• lo ' -by •naf-. which •II 1'MI
`..... dvl1y. Fll1e< bllndwidlh .. ~be - · lot lob
`and CW. An •I omp11f.., -
`lhe signal 10 IN
`.,. _
`_
`,ot _phonM ot opeok ... :
`In 1'MI pn1 • .....,.,. ...,. judged meintv on -
`.... c1ivlry Ind Hnlitivity. Nowadays. receiver
`_ . , , in 1'MI .,._...., of "'""'9 lillnoll ii • mejor
`conoldonlllon. II ii in*-ling 10 ....._. tho DC
`-ond tho-""' whh-nl to phooo11••
`lholconbecolod"-signok:"
`
`Stage-by-stage account
`of direct-conversion
`receiver- design -
`a real Dutcfi treat
`for receiver buffs
`
`I
`
`At-.~- 1ho direct._..,.., IDCI
`the claWc ..,_,,.._oc1yne_
`-
`IOb e - - - ·" - -
`radio
`·
`~- doobl ... moinr. ..... ,,lheDC
`""' - " " ' ii lld<s .,, i-f
`-
`·
`-
`~ toconc1-.....bu11t1as-.. ......
`""11111....u;r ~r·---
`....... ,,, • .,. •• _ , , , . __ lloflllo
`direct•con' ·,afon receiver •• well as some
`ielinen!on11 'hot con be . -. SlrOng ...., -
`Poinll of do deoigol •re dlocu.-. Also 11<-.llld
`... IOllll ldf ' -
`..,. u .. ahould - wifh 10 build
`a DC -
`lo IUil VOU< own needs.
`deelgn prtnclple
`
`in Ilg. 1.
`The DC ..,.._ -
`clooig<1 ls -
`•""'
`The..,... -
`""'.,,...,,,. ....... ""' -
`P'l'tlh :doo1 by L1, C1. An tf ~ CCXlld be in--
`- . bulk io not~in-.-. lnlllo
`
`_,._ - o c -
`- .. -- -
`-
`.......... _,. .... -- - -
`...... _ ...
`... -
`- ...
`-
`...
`-
`ol>out lhlo -1 lo10<I. Wholhe< ot not tho DC
`
`,, ...... .....,..
`
`~---
`..... ~ ..............
`
`Becauae the firtt rwo unwanted signals· don•t exist
`with tho DC ....,_, P-lion con - ly be
`- t oimplet than with tho SUperhol (mote
`
`By Dick RollerN, PAISE, v.d. Marcltstraat 5,
`leldeldorp, The Netherlands
`
`- - - -._con-be•mone<
`of opinion. As 1o -
`k,_,. lho -
`in ,,..
`quency between signal ind Image frequency
`amounts to rwice the ._, in 1M Cate of tie tupefhet.
`One . - c:onsidlr IN DC ......,., .. a oupo<hot
`
`whh an ~I o1 _,, heni. In lhot ..... the,_
`rec:eived just as •1 as a llgnal !00 Hz ~ me
`
`differenot between wanted eiignll and imege tr.
`quoncv would also be zero. And thot ii lndood so: •
`signal 500 Hz abo.-e me tuned frequency uy, Is
`
`cin:uit ll, Cl"' llg. 1. ln Eur- ono findtmanfol
`mese suong ,..1.,,. -
`4 •"'a MHz. Du1 o1
`curiosity I once tuned the Input circuit of my rec6ter- ,
`to one of these 1mOons ,.., 4 MHz in the evening.
`The c:Wcuil uooo a powdofed-i•on '°'old lot tho coil.
`Fe>< an antonno I \Md on.-ha~ of my 40-mom long
`invened Ye"9 tranamining 1n1enna 1nd 1n c>pen·line
`feede• 20 ,..,, 16ml long. Tho f - wa1 coupled IO
`the tuned circuit by running the wke once thfough
`the hole in the toroid. At tlmee.1 \1Vm connected to
`tuned'-· Bui ifll- of tOlkiog •bout im-
`the -
`of the oltlOfwioo unlooded circuit •oad 0.7
`-·one could also 110• lhot tho DC ....W. ~voltl
`- -u - •nd - l ldob lnd . I
`•loo how put cur finger on one llf llMI
`Arrd -
`-
`-PoinlS al tho diiecl-- princ:iplo. it. of
`course posdJlil' w ifQJtl>Ot•• ,,.._.to-•·
`....:. "V
`ll)halinol. but 10 my "*"' thio en1inllv
`one -
`spoils tho moin •ttrocbon - oirrc>liolY - of tho DC
`
`--B<eok11wough of 11rOnO ..m 1utiotlo io rypGI lot
`
`tho DC - · Wo wil lhotoughly i<r-li;o• l'1is
`
`n. ......._ ......................... .,,.. _ ....
`
`........... control. . . ............. M....... WA iR
`dll9 C.. end .......... ii MC Mown lft h
`.ct.natlc of
`
`.... .......... ~ ............. .., . ......_.,.,.,.
`COlll'9irl9 IM..,...........,, ne,.. ot ..._ .. .,.. rlilh•
`......
`matter also. There are m°''· but aeea ..,iou1, snags:
`hum on peaking the input tuned cWc11h Ind micr~
`phonics of 1hio ci<cult. Thole will also bo dNft with.
`mixer
`We have alreedy moll- • · m blNkllvough.
`_.,_music lignoll lllo1.,.11"'-idln1 of llfo
`Thie Ill*"'"*"'" rnanllam • - by • multitude of
`tuning. Tho signoll c:omo fnlm u1romoty 1ttong
`insuffic:ionlly ......... llld by tuned
`-
`... -
`
`•
`The most Important factor that determines
`'whethef or not e·m breakthrough occurs is the type
`_ of mixer used (Of product detector. if vou prefer}. In.:
`
`meny desigos fc>< DC ,..,1v9,., • 40673 mQSfol ot
`similar device ii uted 11 'a mixer, probably bec8use
`an fet hes con1ld1r1bl1 freedom from cross·
`modulation •nd intermodulOtion -10 ill quodnllic
`cha<Kterislic. Thia beneflll bot~ aupothel ond DC
`,...;,,e,. H_,, one impofllnt - · • of9>
`OYefk>oked:. quodqlic -.CIO<illic -
`p<Oduces
`detection of ampll1ud1~moiula1ed signets.
`Malhematical bMtmonlof • ......., __
`invariebly ,..,.. on .,,. _,,,,,_ of • ,...,,,.,_
`-
`-
`• - - •ho-lie. Thot.
`mix• with a aecond-powet ctwecl•i!ldc: in a DC
`........, -
`-
`•m ,_ ....,, nol8d by mony
`..-.10 their dilmr/.
`The ume problem oc,cura In a superhet of course,
`but lhO iesuhing •ucllo algnoll In the mix0< output
`can't pa0s tho i·f empllllO< output •nd lhu1 won't be
`noticed. But in tho DC '"'*-· no l·f ampiifiet is
`prnent, '° every 1f component out of the mbcer,
`wholhe< frotn lhO wanted signal °' 1n u.....,,...i
`
`broadcasting station. <NChH .... OUlpUI of tho
`
`--II is~ to lmprow IN rejection of a-m ligNila
`
`by using two .,.,,.."'" in • - - ...... con(cid:173)
`figuntion. Ari • .....- con be found in -
`1.
`'
`
`•• .
`
`TCL EXHIBIT 1017
`Page 1 of 6
`
`

`
`; ~~
`~
`fle,2-~-~--t >p<ii\ci(cid:173)
`pJe. Tu di :i.;.,it p•ut ol IWJs eiicult ls w<Jh
`u..t- 1 : 1"e cieuil should be desluneci for a
`~ •on tho Primorv"llKI ~side
`..ct""'" a <C: l1e< 111>00 "'° inpUi windiiiQ,F"' Cw
`• -
`circlJ'""" be.-, ... Yl1211L ""· but •
`to fond'"'•·
`Iii my"""*"' ""'only r:oaect SQfufon is _, ......
`. . . -.............. moy be -
`mixorlha1-rt0idol*>cf on11~~·
`sliclot-od>gl .... il<fw!ir) bl.Ilona ... ~""''
`--elc-.-1"evfofrequen:y - nolhe<
`_.ts.1~rypt1~1<N.Adi< le-·
`... ....,. bftlught into _ . , , . - - -
`-
`by the vfo ligr1ll can be uted. Dufing nnslc:.r! from
`""".,.,. to•01her. a-m dMectlorl can still occur
`""""°'• """°"*'' to mate !he dura"°" IA ,_
`on !he c:b..C part of 1"* cfiod;t;~-.;.k. It is
`.,......,... .. short .. pooli>le iiv dffling Ille diode
`- wi$. lltong OlcillotO( ~- - · llie
`an-. olgnal is aloo - to n the diode, ~ lhis
`Agoin. ltis .._....,, - " " 'signa) that·(cid:173)
`sil)nal 11\ovld not inflUence "'9 s.-.1chlng
`.,..,_istic. as this wOuld aloo "°""' t6ooble.
`- - · to llie .,,...,,,. slgnel. Ev9n f -
`-nono- taken. """"'a.m de~ wit ocwr
`-
`llling. ti!>gle diode.
`•
`The '"'*"'IS 10 ""' two di<><ies in ' bilanced(cid:173)
`-· s.-,...fi;,bto .;i..,..;,. c8n ""!°""";,, ....
`5w;iure. ~- one - • in common: a
`.., !he balance. My uperience has
`po110n1iome
`been Iha! !he
`• of this pot ~ b<itb-oo the
`·~· Thio le -d. itoc! in,. .Ji'.f,~~ re·
`, ._ • used end !he osciliator·sig
`ampli.
`one m ""' tbe pot on !he ,,;;;;}"'~.
`-
`An¢ie< ~ >ck of .,,..., Single-blfancect l"i•en
`iscorilidetabk !~"-·
`~
`Another P<I. ibilitv is to use a singlf.. Of ouble(cid:173)
`~llced: """81 using b;poiar -
`.... I s., 8Q my
`ex.,.,;inenu "fd"i direct ~ u$lng. a 1eney
`SUl40 rneg~·clrcuil d<J<liile.balanced - · The
`balan<:e in such an IC mixer is inhenin!tt ve.rv good.
`N - , I four)d su~ of a·m break(cid:173)
`nlJlse Ihm> ·-·,_,ling in poor sensi!Mty 1~
`th"'''llh )lisoppciinlil)g. Aleo the IC produ~ more
`
`My exPerimenu led 10 !he COIJCIUsio<o ~ 11 the
`only eiitirely satisfactO<Y mixer for the DC rO..'liver is
`e double-b611nced mixer using four diOdes. 1 tried
`germanium, silicon. and ho1-c-arrier {Sc lOtlky
`barrierJ diodes .•od homemade inpur and QutPut
`ttaOlforrners, bb1h oo ferrite toroid cores and pow.-
`
`cli<ad Iron i:" with two holes. All gave good
`46 a .:o...i 1m
`
`. 11 the mi)u1t1or a 10.dB tlgna~
`· -· The rf
`ratio wn of the Cfdet of 3.1
`plus ncise-
`1.8-21 MHz 1>0nc11:
`~
`micfovohs for
`Suppr~ a.m br"8klhrough wos -•ad
`
`- .
`
`.
`
`perfect and the vto hod nQ . - harmonics •• _ .
`lion could~ l!'>lv on odd hann9nic:s lthe tljrd
`tile - J. In prKtice 1he -
`hlm1onic
`shows ""'"" sensitMly on the second hermonic u
`,
`- · although leSI than on the thlfd. Elq)llriei.ce
`- thet a.;ngle-tuned ciR:uitwith good I~
`at the input is 101ficient to au - an bl.It the
`strOngelt unWari1ad lignols on hlm1onic freq.-.
`cies. The one or rwo signals the! """""' are MSily •
`rec:ognimd u they peak at a -.nt .mng of Cl.
`As""°""" in fill. 3 1 used two - · i<l!Nt tuned
`cln:ui11: one fOI' 15 through 40' meten and the other
`for 1llO al)d 80 mo-. Each circuit.is tuned by one(cid:173)
`half of a oplh·""tor aopeCitof of the type used In
`broadcast reeeivers. The advantage is th.at no
`within m. critii:al clrculls:
`switching ii -
`only input and a.tpUt lir\ko..,. switchod bya good·
`queliiy tqggle .Mtch. H - . a single-wrioble
`capecitor could be used that is switched be-n
`tw0 ot more~-
`No coif.windlnt details a111 giwn for the lnpur cir·
`cuill n ti- coils depend on form factor arid type
`of core. A toroid core of ferrite or poWdered ifOn is
`preferabla. L 1 Ind L3 can. 11 a stan, be med• equal;
`the same appiel to L4 aod LS. If the link coils a111
`sme•. m. circuh will haW tittle daiTif>lng and selac·
`tivity will be good but slgnal lou wil be large. With
`~ny "'':"" fOlodle links, littlO signal will be lost but
`
`by irjectlng a J0.-1 -
`· <400-IU olgnal
`into tf1e ............ noting the_., ot - to f
`amplitude required 10 CIU,M.-• 10 -dB e+n/n ralio
`II !he OO!PUL (The Ill> Rtorwas;.. opotltion for thio
`delul)Od outside the -
`leStl. The slgna4 -
`
`out. but ,,.,., mix1>r1 olJvioully _,, for from perfecl.
`It is for m. -v same ,_ , thet bll8ncecl diode
`mixers are uhd in radar- rec8iver1 without rf
`om!ilification. So it's belier 10 buy o good doubie(cid:173)
`~ milrer right -v unlell vou know the
`seem of rneltlng • good one yourself .
`Input dn:uit;
`
`PMlbend IO that, wilhou1 modulation, no OutPul
`could be .-red. An..._ of about 50 micro(cid:173)
`. volts' of the 30--percent modulated sjgn•I was
`. -. In P'l>Cticaf ..ms this,,_,, Iha~ using""'
`- -•ntenno wiih.., a~na tu- coupled
`to !he DC......_, no •m -
`wn•noted on
`bind al any lime.
`-
`The voltage mentioned IS tho ernl of a signal gen.
`.,.tor wi1h 00 Ohn-.6 OUllNI ~- Noat that
`often the . - - input $lgnllf is mentioned as pro(cid:173)
`ducing a certoin 1 + nfn ratio aod, tor this lidtage,
`half the genera~ emf is teken. This is corTOCt only
`when the -llor is -m e - to the device
`under test .. and in ~ cases- this cannot be rel~
`upon. So it's bone< to state !he generator emf aod
`accepr lhie fact rhat the sensitivity ftgUres look
`poorer.
`A great surprise came when 1 replaced the home-'
`b<ew mi•ers ,.;111 an A._ M011J8 dou61e·balanced
`""""' (11\e most "'°"PensMI on the Dutch marker).•
`This mQ<er is speciffed !or the range !HOO MHz.
`However~ I wanted~ recei\ter to Operate on 1.8 and
`3.5 MHz as - · Any doubts prow<! to be Un·
`founded M suppression of a-m b"°"through wu
`about the same as my own mixers. But sensitivity
`impr-" to an average of 0.82 mlcrOVOlt for !he
`'
`bands 1.8 throog!> 21 M Hz!
`This reveab anolhef very good reason for using a
`well-balanced mixer: i1 appears that with my home(cid:173)
`made mixers, receiver sensitivity was limited by noise
`1ha1 ampirude inodularea the oscilator ~I. In a
`per1..,1 double-balonced mixer this noise is balanced
`• An4 protlfibitV the tncsr enilv ........ 10 Ameril;:.an """*'" .:
`well~ The MOU:B is pricOC$ •t t7.CO (plus PQSWolJ Ind is~
`1,~ me "'*""~. A.rwic Elec.1ronQ,'i39 Green Suwt
`Wa1thlM .. ~huMtts02154,
`.
`
`TCL EXHIBIT 1017
`Page 2 of 6
`
`

`
`- ,....,,.
`
`79-7 t
`
`...a.L~~dlilvfo,
`C2
`C1
`[1
`tloFI
`C.,Hl
`fpF)
`
`lle4145-
`
`s
`""""
`...
`... ~ '"'
`.,._,
`•w.J ~ ,._ ...
`..
`T-3 ·- •
`"
`.. t.4q.w• "
`"'
`, ....
`..
`.....
`•
`.....
`•
`sefeclMiy Wll be poor, A. rew ·11iels WW qu;.;Jdy
`btl!lg • 900d <:Omll<oci>ise betwee1>' -
`conflicuog
`~ I'
`variable-frf(luenc:Y oscll)ator
`OscillatO< Jtoquoncy determines the tUl'ing .. _
`· so freq!Jeflcy •lll>if.l'f is ptfhaot the
`of the -
`most ~ Olcilator reqU}ren\81\t: Tw.t oscillator
`llignal~ haW .,,.,.,_, noise modulation. As
`·~~ot;.:!:..i~."""t .. tnY~::.
`et P<Oducei-lets - - a b;poiar
`pelienee,
`is of no CO<lSidofation in a
`1p111 -
`iranslst0<.
`job oC Ille f~ amplifier. whicll is
`vto; !his is
`eove<IMI in• w s>an of the itrtidtl =
`Fot a m •"'tiband, dlfect-c~ IOCeiver the
`oscillator ' gnaJ should b_e 1Jenera(•d directly.
`without 1>111 ,;,.;ng m Jrequeney multl~tlon. This
`_r~uces th:\ risl. of spuri®s sig!~.ali end noise
`"10(fuf0tlon. It is,.of rouse, n01 a ~ Ulslt to limit
`li6q(1611t\i 4ift 6" CM lli(llWlt-fh!Qu~ bands. bul
`
`with an 8'ectroty1ic. capacitor in parallel 10 $Uppnt18
`noise from tl1'J ze-.
`OUtput is uiken fttlrn 11>e drain through a small
`resistor. This method is due to DJIBZ and feanues
`minimum effect on oocillalion frequenqo. 8 -
`c. current pulses flow
`the let _ , ... in -
`1hfOU9h R2. and lhe reM11!ing ~ is far from
`...,.,.._, We can i"""°"" Ille waveohope - agalr)
`aca>rding to DJlBZ - by including C6. It isjjivon
`such a value lhet iu "'""'"""" e ioogt>ly l!<IUOl IO the
`value of R2 in ohms for the tolghe$I frequenc:Y of the
`band conc:emed. With the component ........ ;,.
`fig. I 01Cillat0< output vol- is about 0.5 V rms.
`Rl is requi,..i qnJy ;,. case ot vhf ponisltlcs. which
`~nK .. 1 lhemselvel in-~ freq.-y ju"""' when
`roteting Cl. Rl .should no1 be any higher than
`.......,IV fO< good Sllpprtlllion of parasitics. In my
`
`s!Ock. Ailhough no admirOf ot printed circuitS for
`home Conllruction. I put Ille five vfos on a-piece of
`epoxy board to sava room. The P""'°9'*Jlhs show
`packaging deui;ts. A good--niotiori driw on C1 is
`~· The one I used came from one qt.the
`popular 111 - · In E..,_l war surplus SCR-193'
`tuning uniu. It ""8 a 1:50 worm drive. Originally the
`worm was driven by a thumbwheel protruding_
`thioug~ the front panel. a - this method of
`tuning is very tiring when done regularly, I moved the
`unit thiough 90 degreM. The shaft Of the "l"'"' was
`••tended so ~ can P<Otrude tlwough the .0. wal ol
`lf1e "1o box and me front panel ot the recoivef when!
`II carriOo a big knd> with a cnlflk. The e""'"5ion shalt
`was cemented to the worm shaft wilh epoxy.
`The--motion drive flu one end of the capatjlor
`shaft. The olhe< end of the sheft pro1Ndes through.
`the end wall ot the shield bOx. A drum•type fraquen(cid:173)
`CV dial can be fined to this Shaft and with a separate
`frequency calibfBtion for 88Ch of tliift bandl. The dial
`can be read -through a window in ttle receiver front
`panel.
`buffer amplifier
`
`'"'
`
`..... . ... lpFl ........ (olvMI
`...
`.....
`...
`""
`""
`... ""
`'"
`°'
`,.,
`,,.
`..,
`•
`... ""
`...
`"" ..,
`• . ..
`""
`•
`,., "'
`...
`'"'
`., .. ,.,
`,,.
`.,.
`"° . ,,.
`"°
`
`C2
`
`l"'
`100
`
`f
`
`was mod°ified to a ~ pjaltl in the ro1or. !The 40- ->·
`meter band in the Nether1ancfs cooiprises onty
`'.I000-7100kHr. the flO.me ... band ~1800 klU).
`Tobie 1Cisu1he va!Ues ot Ll. capacit0<0 Cl.c&.
`Bild Rl, R2. lMO<e about C6 laler), ~~you halle .
`for C1-C5 and L 1 avallable
`•deify the .. .,,. -
`!her•'.• a 909d charic:e that one 0< more ot you;
`oscihatOl"S won't oscillate. In this case your feta
`pr-lyhavelets~ ttianl!le0<>nl
`happened to IJave., lt'sthe<efore better not to f (cid:173)
`JTJ'( design blindly but 10 use Ille porta you halle on
`hand or can obtain, then taik>r the c~uit to vour own
`requirements. A few guideines: In principle C4 and
`C5 should be as large .. possible and C3 •small n ,
`~. This-Ml!p"r'bvide-coupfingbe- fet .
`and tu~ cifcuit anc:t wil minimize the influefq of
`the let on Ille oscilation frequency. If you go too far
`In this direction, -bade. will become too small and
`oscillation will l10P, for a suirt. C4. C5 can be made
`equal.
`The indui:i8nce of L 1 is determ~ bY the lOweSt·
`frequency in Ille band desired. LI should re$pf1818 on
`that lrequeocy with the Jotal capaciuince in po111llel
`(Cl set to maxi.-..m). Variation Of Cl wllh respect to
`the total fixed paralel capacilanc::e detemiines me
`bandspread. This means f"oddlino with the capoc~on;
`and coil until tt.6 prcper frequency range is coveted
`while the oscillator stil oocillates.
`It ,seems Chat many •mateurs sliy away from this
`8pptoach. But in practice It's not as bad as it sounds
`so "?"II .. you're -
`aware of what you're doing.
`Typical ot this is the fact the! the values of L 1 in
`table 1 - • no1 calculated in the design SUIQ8, nor
`were !ll8'/ -'*'· I computed them afterward for
`this artit;le from the frequency ranges covered and
`the capacitor vaJoesJ: so these values are oi very
`limi1ed accuracv.
`Supply voltage •hoold be .. 1ow .. .,Qs.a,1e con(cid:173)
`sistent with reliable oscillalor starting. I (ound 5 vQlts
`10 be a gOOd compromise. Curreot increases quickly
`with voltage; and as dissipation in the let inc._
`with·voltage and current, freQuency stabi1ity suffers.
`An improvement suggested by PAtTW and
`PA9HWE. which I haven•t tried, is to use e smaff
`preset l>Ot between Ille cold end of thli rf chOl<o in
`the source lead and ground (fig. 61. Jhis seems 10
`give smooth osc:illation control. It is absolutety
`nec,.,..ry to sUlbilize the 8"ppl\I voluige tiy a zoner.
`
`TCL EXHIBIT 1017
`Page 3 of 6
`
`

`
`• -
`
`-
`
`- ~-tit-!. -
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`-
`
`,
`
`-
`
`I
`
`r----,.----
`
`' I ... ,
`
`, .. ,r.
`
`... ~ ,, ... ,.. rl•
`
`,.,.,_.., ...
`
`L
`
`- - -------,,, -.J
`...
`..
`
`......
`
`"'"'~..,,-:.....~~~~~
`
`_...,,,I lo-· !hat LO -
`
`'
`""I '"' lhe ...... lion ii ~ i~ many
`-
`. , . -. _.for DC ,~,.,.-mixer
`be sufficiondl'
`S110flll; IO t owr0nwe S lta hlrr •ul th3n inouffi(cid:173)
`cient dri...f
`fpr the M0108, th manufac-turer
`· tPOCifios -
`<N<o<1- 11
`LO drMI ol 7
`dBm; Niii, 5 -1111. lnputimp danceollheLO
`• Before we cen d9t ide on me s-tep(cid:173)
`•
`I, we -
`10 know~,. opliinum load
`03. The 1c vol- 11 03'• eollecror is
`se1 by lhe tooc~lotor ourpu1 volloge end me buffet
`omplif.,.\IWloge lmPl~icotion. The~wltage
`81 Ille°'""""' IUmO OUllD be 1bou)l2.6V,
`.
`ii deli- ,,;y.. ,;.;.... when
`MPlmlf '1 -
`lhe «: ""' !'II" 03 ii mo•lmum. Tht direcl CU'""'1 is
`10 mA, Tllli ""'current•• 'the ooler or Qn therefore
`be obclut f mA 01 mollimum Clhe ~ ok ,,_ii"*'
`10 mAI. The oplimum io.I ~ ce ii - (cid:173)
`-
`7 mA 81 2.CIV. which -
`out 10 be 371
`
`50 D no•- 1177
`
`ohmo. So lhe •lllPdown wlioge ratio of T1 lhould be
`J'l717!111• 2. n . -
`of lhe cons1ruClion of n ,
`lhe rodo can only be• whole number so- c:lloooe 3
`aslhenu,.t.
`Conltrucdon of Tl lo . - , in fig. 5. T1 is en
`autotransformer with • trifilar winding:. The ccwe is
`powdered lton with two holes, as osed in Eure>pe for
`batun tr1ntformera In the input circuit of TV
`receivera. No doubt a tultable toroidal core of
`powdered Iron"' ltrrito would do ju51 as well. The
`
`l)umber of rurno 11 g- bv t~ requinlmon1 of
`
`sufficltn1 lnduc10nce 11 the lowott frequencv to be
`uaed. S -1 orbllrorily I deCided that lhe in·
`duc1""' ruct1nce of the winding 1>e-. connec·
`tiona 1 ond 2 ohould be four times lhe lood im·
`pedonce: 50 ohmo ot 1.8 MHz. A .. mple winding on
`lhe ccn In ~ with t known ~ ond
`COllj)ltd to • grid dippe. -
`that fiw tuma
`-.Id be NQUlrod -
`comtction 1 ond 2. To
`
`obGiin a n•afcwrrm f'ltio of 3 1 IO .. of 15 turrw wil
`be-.ofwhich10--Wl•ltCliol•2
`ond 3. 1 . - Iii<~_,,.._. of 32 AWG
`
`•t ~to b e -. ThnMI
`
`I0.2mml -
`
`pieces of ....... - · wllich • - c (cid:173)
`.,111' with I hend drill, Ind lhe " rapt" IO fo<med lo
`put five times lhrough one hole of the core and bid<
`1"'ough lhe o"'°'. Connections 019 then mode n In·
`clicatod In fig, 15. Make lhe oonnectlona H _, 10 the
`00<0 • ~ Tne compltllld tronaf0<nier is -•v
`cemented to the buffor·omglifltr shield bo•.
`Alfv smol bolt con be UMd for lhe llbflor omplifior.
`10 lilt oteilotor i-.;ng, ..
`The -
`is -
`-.inonool1he~A-leod­
`thmugh carries lhe aignol If"" the -
`Into the
`buffo<.-plifier baa. At ..; ..... f)ltcOUtlon ogoinlt
`I put• l*'ilion- lhe..., _of
`-
`the ompliliw. n -
`fn fig. I . H - . a the
`io only_.. IM-. thlo -~io
`...,._ a1oplilicolioo ......,... 02 -
`ond 03 cof(cid:173)
`-
`'*'-"'"*
`
`buffo<....,iifiot from. --·-aupply.
`
`y.ltlo,..... .... -10-the
`
`It is \IOfl' impofant lhllt the buffet • - opero•
`within the linotr pan o1 llWlt tronlfor chetKteriotic.
`As soon n • tronUtor bottomt out the -tion
`olferild is gone. Tnia can be eotlv chtd<od: pulling•
`le>od on !he amplifier in fig, I should have linlo or no
`effect on lrequencv. If frtquency shift occurs ~
`shillding and decoupHng 110 olright, then ovord""9
`of the buffer me,v bt tlit cou11. Oeeraeling R2
`lowers the Input~ 10 the buffer. I( this'--'"
`10 be the rOmoclv then C8 .. , to be conecuid •welt,
`n expla1nod eal1iof.
`audio filters
`1 began mv .. perirnfnrswith DC~ usng on
`SUMO IC .,._ modt by P*My, wliich l1n 350
`ohmsOUqM~· fhe miur-f-.clb'(
`a filler for ub will> a 2100-Hl cutoff ' - · The
`a S<>COled C..... or elliptic Cunclion dtsign.
`filter -
`which 011'111 the ..,,.., DOUi>le uansition .,._,
`passband and sropbend ""'1h • glwn numbe< of cois
`and capacitors~ Ill d1Ndv1n11ge is lhet all
`componenlS have odd values:. so coitl IT'fUt1 be tab4
`made b'( poralleting pot CO<OI 1nd copecitors of
`11aodlrd ·valu.es. The fi~er did an 1.111:ceUent job.
`
`•
`
`-"'·
`I ,.pla_cod the SUMOwlllt I double· bolanood mllcor
`using OA 1540 gormenlum dladff, wtilch had a
`..-.reel output lmpodenco of obout 125 ohms.
`Since tlie filter ... d to be ,-.ivnet1 for lhe diffo111<1t
`the -Jinown 118-mH
`lmpodonce, I dtcldod to -
`toroids for coe with 110.p.rd·velue copecitors. This
`would mike dupticolion"llv othett Oltio<. Tho -
`-
`occoidlntltt dtligned to lhe rulta of - . i
`~-thnl<y.
`
`TI-. fi10n do not -
`uansition
`lhe -
`- ~ lftd IUlpbtnd o f f -b ' ( -
`-
`. IO-c__.;on _ _.,. b'(-·
`Ing the cutoff'- to - . t 2llOO Hz. which e
`oulficient fOf ub -lion. AfW aomt - . the fil.
`... ol fig • • -ooc1. Tho 22·mH coils - - .._
`b'( piecing lhe ..., windings
`from 118-mH -
`
`...... ,.._ ... -
`
`in perotlel.' It II lmporUint !hot the coil .,. c.,,.. .
`nected.., thot lhe two windings don't -
`-
`pe-nd onenuotlon ii of lhe
`other. In lhe -
`of 0.5 dB; cutoff occurs - r 1900 Hz. H you
`-
`wilh to chtdt""' dtlign . . . fil10t -
`-
`be
`found in ..,.,._ 4. It lo built from ' - cc>nlWll-k
`hadsectionl.
`•
`For the CW filw I concl ICltd 1ho1 the ~
`not only is it diffi(cid:173)
`should not b e - - · -
`cult to get the aigr10i within lhe _....,.i ond .._ it
`but the ...... ii otwevt of the -
`pitdl, wliiclt
`-
`- - !iring during P'oloneod lillllning. This !Kt
`Q f fig. 7. which contlots of foll' 3-
`led ... "'" -
`....,_. _.. haN - - Tho .,-nd C- 3 dBi
`is be-. obout 580 ond 900 Hr. In muticol - ·
`this '!""'"" lhllt lhe lignel con be tunetl tlifough
`about 1 fifth without no-ble varitlion in ampli(cid:173)
`tude. Attenuation inert• ... faster. (>n the high(cid:173)
`frequencv lido of lhe peaobtnd. The CW fll•r lo elso
`usefuf as 1n outboard unit between any teceivef and
`h6adphones in cau the receiver or transceiver has
`insufficient i~f selectivity tor CW use. Some resist·
`
`·s...-0 ...... 1...Hc:hOil'OYldlS~~~
`~"'Gin b9 obUINd by ~ ~ fn:lm . . . ciqu(cid:173)
`
`.......... ~
`
`........... 1177 a s1
`
`TCL EXHIBIT 1017
`Page 4 of 6
`
`

`
`larger than - pr--
`
`· As '"''*' be(cid:173)
`
`c-- spread in the chilractarillb of lets. It
`
`..-,.s. Optimum sourc:t1 -n c a for the al fn.
`out .,....,...., ~ on 4lvU IVP9 and collecto<
`cunvit and can vary between a few thousand and
`-•-oflhouMnd"'--
`From 1:heM die. • 11.1itlble transfonntf can be
`°""· or couid \'OU buy
`speofiod Bu• -.Id \'OU -
`"mmawhl'97 Perhlpa. 8u1 don't run out 10 get one
`·Nol odr ohould the
`-
`-
`-
`ii -
`•f ..,......., he\19 • ceRllf'I resistance •t its input;
`at me ...,,. ,.,..., the af Rw tn use should have the
`PfOPIJf """'9"11Jb0n rws11tano1: It its output. The load
`on the-wlbe the~tancel!lf the al a111>
`
`- · -<Kmed by the •nsf- - so tl1is ....
`
`cpremen1 allo flll• the trt ndonner ratio. tt wouk:t
`be rnoet unlikely that the nnsformer nnio so found
`would coincide wid\ rhl one for optimum noise
`metching, so-,_,, a p<Oblem.
`It would be nice u we could choose a uansformer
`
`.._P9dd 'II .... ~ be~ 10 -Ith the
`filwto•..,~endt1 clpto••• ..... a
`The"9"' -N•IM-d ....... no : -.
`
`_ .. ____ .......,191*•-
`
`"" 1-di>t.- -
`A ..,..._ al the CW
`_
`..... , ........ - . . - ...... The
`1.,.. .-_ LI, comlMne to """' cools o1 -.c
`
`-21.1. By-"'....-. flt. t -
`
`• bendpMa filter
`•-in-DC-~W10$Tend
`
`NNIRL - · Wi1h - - odr""" hlllf
`- - -
`,......., The ....... "' ... · -
`--the~ il--odr """tlwt
`
`llope ii on
`
`---...,UC-.__ ob -
`
`_...,.,,.,_.,,fie, •. ,......,.,_,.
`ol N ,_ ., fie. t ii lhot die -
`die_,,__.,, tlw~. - · th e
`,,_ ~ the """ -
`onlot -
`coil if the -
`·
`"""'*' i.- become II U In,,_
`In. t hod -
`jcWIOd ...,. coll In thet caM
`Th e - - ........ tion'-...... al .... -
`~140.ohml '"'the Ill> filter •nd 104 ohms
`~~;..~L.t~~=~i::::~:
`
`-
`-
`
`-
`
`... ,. MJoo-ed -~.,cw-.
`
`wtich'-f"iy50ohmloutputim*"'._· Luckily,
`liters baSlllt on lho " ' - pet......., c.sign mtlho<I
`a.ea COfnlllOmiM on ""'ttlwlg onywey. end on PIK•
`lice lho """'"tell on lho lnpst lido -
`nol notice(cid:173)
`l l l ly - 4 rom -
`petfonnonce. Output ..., .. ,,.
`Ing ii~ in"'" ..... - ; " ' ·
`. '
`62 • "'1-
`
`1S17
`
`Sonw- rm .,.,ed why 1 use old , _
`coil-apocilOr liters -
`!hot ...... -
`Wdbout
`eoila are ftllillble. In &he first place rm not sure lhat
`. . . , ,_ equol to my~
`. ._ -
`-
`.-o1 .... _..,..,......,.,__Butl(cid:173)
`-wouldbeso....,..;rm.,,..,ee0<-.(cid:173)
`
`ello •more fundoi1•10ll objection'° the uoe ol ec·
`- - in lliil . . . -................. ThespeetlWn
`o119*- to the -
`by the mixw COIT'C>rises
`•'"'Vo~ - - - easily be -111
`d8 or more. I fell' 1tW range 11 more tfw\ an active
`---·-"'"-~ ....
`-
`.. die ..... "'"'"- CK the..._ ....
`it. One lhould not l"'lfO' that. _...,In
`-
`cutoff-.. some pons al lho cir·
`the CMe al -
`cuit wlJ QIJY mucll higher voltages ,,_ -
`output ten\1lnels -
`-
`.. -
`points
`"""" -
`the ci._. ol - ' I I ii greewt. ~I
`.,,.-the.-LC-.ConstnicWnlhould,,,_
`no~·• end the 88-mH ton>id& ar. iMxpenoiw
`lndplontit\11.
`
`on
`
`1udlo-frequ.cy 1mpllfler
`In I DC,_ the mput lignal is oo1y •-tltd
`
`In the fira1 stages. Amplification oca.111 for the fi.st
`lime altet the al filters, so signal -
`roecl!M a
`minimum 11 the input of the If amplifM1r. U.-1he
`mixer is PoOr end oscillator noise. dominates.. the
`by the ai
`ligna~tc>-r1"'8tt rado is -
`-
`-
`amplifier Input 1"'90. I egeirl gel the imJl'eaion that
`lhil lmportent consideration did not ~ ........,
`in_.., of the DC·receive<
`lhe enontion It -
`designs I've-..
`If a bipolar transistor is uMd in the first stave of
`the If omplifier, ii should be 8 low-noise device; e.g.,
`• type sul.- '"' the input 51- of • tape or cas(cid:173)
`sene - · From the menufac1un1r's dlta - .
`one can rm the collector cunent for oplimum noise
`factor. usualv some tellS of 11tic1aempc. But these
`.._.. . ._ lnOthet feet: optimum noise lactO< ii
`obtained ,..;,,, • specific output -
`al •he sig(cid:173)
`nal '°"""' for '-Inv the lnnSistorl Agreed, the
`curve for noiM fKtor as a function of source resist·
`.._ ..,_. • ralhe< b<oi8d minimum, but if the of
`done, noise -= may be so
`,_ • c..-..:tod directly 10 the output al the al
`,.__ .. ii -
`that lfn ratio ii dograded by_,,I dB.
`-
`Prof-..i designe1s of ._._ al amplifier$
`use input transformers to obtain an optimum noise
`match if the source resist.ance differs widety from the
`oP""-'m. •in the case of• dynamic 11oioop.°1o1K' for
`........... We could do 1he same in our DC niceiwr.
`Source resistance in this case is the SO-ohm output
`· - - of the double-balanced -
`.. ...,,
`""OU9h the al m ... -
`the major part al the filler
`
`'
`wtth an audio _.tor end el voltmeter showed
`
`the " " ' - step up from pnmery 10 -ondlry lo be
`al the°""" of ffty. So ligno~to-noile rano • -
`fifty times comp1red to a straight connec"°" w.thout
`a tninsformet.
`•
`The matter of Rutt Cl!fYTW\8tion ii •rill to be N11lld
`Tu;s is siml>lv eccoumed for by R2 ., ,,..-,_.th the
`~ windng. R2 is transl- to tM inout
`side of the trandonner. _...,the -
`of the
`in my cau. The liters ara _.,.t undlt·
`tranSfonner iano. The . - of 1 rNgOlwn -
`"'"'
`-
`loaded ., their ootputs. but tigl\Ol lrom the Rtarl •
`
`fore. liter ""'tcllirig is not -v ctilical.
`The -a l the I f emplifier ii simc>le. Thof9 ii
`
`is therefore bener to use a ..,.,..._ teSiator • aource
`resistot R3 first and to e4',tst it for • de vot19gll of 10
`
`'
`n1, I. Tht cw flt1., W • combin11tion ol '°"' halftMCdon&. 1hown to the ,?gin of tM ~Cllf ,,.., N01• how c°'h and c•p.chon
`i,,, th• indlvldU91 MCllOtl• •r• combined In the •ctllMI filter •t thti i.tt o f tN d~;-9 v..-:tc..l llne.
`volts at ttte· drah of 01. The reststor ls then meesured
`and a frxed resistor of narest st1ndlrd value sub·
`~tiruted.
`The major share of the tota1 1mplificatlon is pro·
`vidded by the popular 741 op amp. Resil1<>ra R4 and
`RS are ulacted so that de "*II• a1 the ooq>ul !pin
`61ishall the"4>P1Yvoltage; I.e .. 6volto. 11--ry
`R4 and/or RS can be changed H the vol'ilge II pin 6
`differs to0 mu:h from haH supply vo4-. Volume
`
`ratio for best sin re1io at the at·input siag'!_ and con·
`trol flft8f terminetion separatetv. We can by using at\
`fet. An even f>etter 1/n ra1io than tMth • b~r ttan·
`liltof atso occura. The input res$tance of an fe1 at
`audio frequendes cen be considered infinite for our
`pul"POM. ao 1here'1 no k>eding at the input. Noise of
`an fet ii lower thin in ,even km-noise bipolar tran·
`altitora. NoiM In the let can be lhoughl of as being
`gener•UKI In a noise-voltage source in series with the
`
`Input (91tol. The highof the lignel input vol- is
`
`. - , the-thea/nratiobec<KneS.aaP9naland
`...,.. voll"90 _ . .. in -
`on the gate. So the
`hlol* the step up of the..,....,_ between al fil.
`tar end al emplifier. th e - the s/n ratio-.
`The limiting fector ii the ..-""""' itself. W.th high
`mtioo the capecita1 '°" al the - ry wWling
`limits the high-~ ....,.,._. But this ii of
`more_., 10 the hi-fi -ipnwnt desiilr*. es the
`~t eudio i._..,.,. _.,. in-UICI in ii about
`:zllO()Hz.
`Very suitlb'8 for our purpose are microphone
`
`..-om.. from c:ommlPcetiooa equipnwnt using
`tubes twor-"'rplua tn1ntloinara. obsolete mobile
`. -1. Translonner Tl ir'fl9. 3 came from w..(cid:173)
`Set 1ype 19. • himou• f l1ilh WW II wtaran. The
`pl>o-11 ahows the UM in its shielded box. A "'5t
`
`control R6 ~ the I-ck: this -t •
`
`unusual svsten has - - with op 11'11j>O. Sili·
`con diodes CR t and CR2 wll protect \'Olfr oera in
`cne a stronq oignal a-ra u-tadly; they llmi1
`ootpul vol- "' • maximum of mout 1.2 vol• pell<
`to peak.
`Bec:ause the OU1PU1 resilaance ol an OCI •mp is so
`low end~ - - -with f-. connect·
`ing ""'diodes in peraw with ""'~- -
`not be _., efllctive. Tha1's why they ,_,, -
`in·
`81JO.oflln
`citcurt. -
`corporated in the , _
`......., headph:in. ire used,..;,,, both -
`In pet·
`don't conduct II nonnal lillOning
`alel, the -
`Jewl. II \'OU prefe< J><IP'9'0UP _.nd - · -
`omit CR1 and CR2.
`When stani>g my tallS with direct convenlon I
`used a speak•. Ins- of the 741 op emp I triod 1
`
`novomber 1S17 a 63
`
`TCL EXHIBIT 1017
`Page 5 of 6
`
`

`
`""°'* much -
`
`S - TAAJOO ond ltle Plesol), SUl30 ff OUtO<ll
`IC ......,_, but I like the - " ' frO'TI good l\6lcl
`- M-tn ,....... headohonet
`r•oducilon, Mn vt<Y 19ht
`i.w ...., good -
`""~· ft. ddtN;ult to · (cid:173)
`hum ii -
`hum ...i..cuon "°""""""'· -..-Tl d - - 10
`...... *'l'9Y ~ r.-..o1 ~· line ~onnet'S.
`.... ' . · - CIOcrnl -
`- Fbt .... -
`C2. "'
`---"'8outOU~is ..-'9Cho< ...... Thetre(cid:173)
`___ ., 6_d8 l)('rOC- -350
`
`""'Y pleasant. but o nuty side effect spoiled pet·
`formenc:e: when tu!ling • 11tong carrie< the age volt·
`oge "'°'"'"""' In I oeriel of llll)O, I lried chlnging the
`timo consun11 in 1he loop bu1 nolNng helped. Also
`the Ptessey -'""""" .__.could not lhnk ol a
`rerr>e<ly. It is -
`in which m.y
`tion ..., . . ~led f\,Jt tf• .......... 11#-t -
`perfonn ex-dy
`but in the DC -
`they
`
`11111 the Sl.810-SLil21 c(cid:173)
`
`obviouoly do not fool II home. r-......., I -
`
`age. First. lhe rf amplrf* was ncn needed from a
`sensilMty- o1-. ond secondly my -
`•
`pan of a ORP lelogrophy t~ and it"""' ....
`docn used for listening to ub.
`tunable hum
`Many DC ,_.,.,. hive bum that apoears when
`the input circun i:t tuned to the signal (oscillator)
`freQuencv. I .,,,nk the exl)lanotion iS that socne oteil·
`IMkl tlvough the mil<er