lot has appeared in recent issues on the subjectof the ESR (equivalent series resistance) of anelectrolytic capacitor. The Capacitor Wizard

was reviewed by Martin Pickering in June 1998. It'sdesigned to measure a capacitor's ESR in-circuit while

'

ignoring any components that are connected to it. The.unit described'in this. artible performs the same task, ..and a lot of work has been put into achieving the end 'result. Even if you don't get around to building themeter, this article will give you insight into the designcriteria and the way in which the instrument works. Butbuild it if you can: it's effective, very useful and inex-pensive.

ESR : . , '' In view of Martin's review and also the articles by Ray

Porter on a capacitor's ESR (January and April 1993) Iwon't say a lot about ESR here: it would simply be rep-etition. To put it in a nutshell, a capacitor's measuredESR (in ohms) is an indication of its 'goodness'. TheIower the ohms reading, the better the capacitor. AnESR check can give an early indication ofcapacitor fail-ure, and is far more useful than a capacitance measure-ment. Indeed many faulty electrolytics show OK whenchecked with a conventional capacitance meter.

In recent months I've talked to many people whodon't appreciate the importance of ESR and in whatsense it differs from capacitance. So I feel it worthwhileincluding an extract from a technical bulletin on theCapacitor Wizard written by Doug Jones, the Presidentof Independence Electronics Inc. It sums up the ques-tion of ESR well.

"ESR is the dynamic pure resistance of a capacitor toan AC signal. High ESR can cause time-constant prob-lems, capacitor heating, circuit loading, total failure etc.A switch-mode power supply may not start reliably - orstart at all. Slight hum bars appear in the video of a

The subiecl of qnelecfrolytic copocitor's ESRhos generqted <r lot ofinteresf in recent issues.Alon Witlcox hos tqken it ostoge further in designingo procticol ESR meler. Thisfirsf porf deols wirh theoperotion of fhe circuitryused in fhe meter .

Design.,of '�uh''E$M6tBrVCR ot monitor. A TV display may be pulled in fromthe sideVtop/bottom. Diode and transistor failure canoccur over a period of time.

These and many other problems are often caused bycapacitors , with normal capacit'nce but high ESR,which does not exist as a static quantity and thereforecannot be measured using a conventional capacitancemeter or a:DC ohmmeter. ESR exists onlv when alter-nating current is applied to a cdpacitorbr when a capac-itor's dielebtric charge is changing state. It can be con-sidered as the total in-phase AC resistance ofa capaci-tor, and includes the DC resistance of the leads, the DCresistance ofthe connCction to the dielectric, the capac-itor plate resistance and the in-phase AC resistanc-e ofthe dielectric'material at a panitular frequency (myitalics) and temperature. . ,' , ' r

The component combination that constitutes ESR canbe thought ofds a resistor in series with a capacitor: theresistor does not exist as a physical entity, so a directmeasurement across the,'ESR resistor, is not possible.If, however, a method of correcting for the elfects ofcapacitive reactance is provided, aid considering thatall resistances are in phase, the ESR can be calculatedand measured using the basic electronics formula E = Ix R! This is the basis of the design of the CapacitorWizard."

Design CriterioCapacitor manufacturers quote ESR values measured atl00kHz. So this is the test frequency I chose. Theimpedance of inductors in the microhenries region canbe measured at this frequency, enabljng the c-onditionof video heads to be gauged --as they w6ar and the gapdeteriorates, their inductance falls.

The Wizard has a buzzer that sounds when the ESRis below lCl or so. A capacitor with an ESR of lessthan about lC) is generally considered to be good, so

ri'i

r 1 6March 1999 TELEVISION

this is a very useful feature in situations where youwaht to cheik a number of suspect components,. itmeans that you need refer to the meter ohly whenthere's no beep. I've incorporated this facility, b'ut you'must bear in mind that a lot of the capacitors in whichwe'ire interested have ESR values of less than 0.5C)when good. More on this later.

I'd like to stress this basic point before going any fur-then as with the Capacitor Wizard, the meter describedin this article doein't measure a capacitor's micro-farads. It simply lets you know if the capacitor is orisn't up to the job. After gaining some practical experi-ence with the meter, you will soon get to know"whatreading to expect from a good capacitor - taking intoaccount its capacitance and voltage rating. But in'anycase the reading obtained with a faulty capacitor usual-lf lea1e1 little doubt as to its condition.

The Op-AmpThe circuit'uses the basic opamp as an oscillator,amplifier;' dslsctor, voltage-follower and .'comparator.So'it's appropriate to devote some space to a description'of the op-amp and its associated circuitry. Incidentallythe term'opcrational amplifier' relates to its use in ana-logue computers and appeared in a paper by Ragaziuiniand others in 1947. The first general-purpose op-amp,with.differential inputs and using the familiar triangularsymbol 'for circuit rcprcsentation, was :introduced;in.1952 r 6Model K2-W, by'.George A. ' PhilbrookRe.searches Inc.). It's sobering to think that almost fortyyears ago an early op-amp, the P2, cost $227 - an eighth'of the cost of a VW Beetle at that time: now a superiordevice can tie bought for less than a pound.

The op-amp is a high-gain (x100,000 or so) amplifi-er that usually has two 'inputs, on-e non-invertingQabelled +) and the other inverting (labelled !). Forpractical'purposes.the gain.rcan be considered asinfinitely high, with no cuirent flow at the inputs. Theopamp'isr,designed primarily to' operate stably'withhcarlr- nefative. feedback' In 'fact from 'the' hisfelisal:;poiriio"f:vlcw:de opamp andithe concept of negative'feddb4k(the, invention :of fi,S.'r Blach-'workingrforr'Bell_'[5t*iratories, in 1927) are synon5rmous. Black'wa5''w6ikihg'on telephones,' tis bu;ective biing to:.achieve'stable gain independent of the characteristics':of a valve (a thennionically-activated FET to youpg-sterSl). When he tried to patent his negative-feedbackamplifier in 1928-the idea:was ridiculed. Over theyears however this concept has become one of themost important in the field of electronics. Marconi hadmuch the same problem. It seems that people often dis-miss things they don't understand.

' 'Anyway, I digress. To get back to the point, the op-

amp usually:requires a positive and a negative supplywith respect to a common earth. These.supplies areoften noi shown on circuit diagrams, being taken forgranted. Thq common earth (0V line) serves as a refer-eirce point for the voltages that are present in the cir-cuit and as a'return path to the power supply for anycurrents generated by the device's operation.

The main point here is that if the voltage at the + inputincreases with respect to the voltage at the - input, theoutput voltage will be positive-going. Co,nversely if thevoliage at the + input decreases with respect to the volt-hge at the 1 input the output voltage will be negative-going. Thus in normal practice the output corresponds tothe difference between the inputs.

If the op-amp doesn't have any negative feedbackand the + input is at only 0'lmV above the - input, theoutput voltage will be close to that of the positive sup-ply rail. If the + input is lower than the - input by thesame amount, the output voltage will be close to that ofthe negitive supply rail. Thus the gain is equal to theaverage slope, which is typically l0V/0'1mV =100,Q00. This very sensitive property is used in com-parat<ir circuits (it's'used in the ESR meter's buzzeicircuit). But the op-amp is far more useful when.theoutput is restricted to narrower limits.

The Precision Inverting AmplifierThis is one cif the'riiost common opimpapplicationsand is used in the second and third stages of the meter.Circuit operation will hopefully be made clear by therather unusudl representation (due to Tom Hornack)shown in Fig. l.

At (a) the op-amp is arranged to provide a voltagegain of two. The fact that in this case the output isinverted (the gain'is.minus two) is not important. Theheavy negative.feedback,viaresistor Rf forces the out-put to be such that the voltage at the - input is equal tothat at the + input rjvhich is'oV. Remember that the op-

:: amp responds' td the" difrerence between its' inputs. As;,point X is at earth'potential, there is "lV across Rin',(lk0) .and,ther,cirrrdnt flow via Rin, calculated by'Ohm's

Law;.iS lmA. jTherc is no current flow at theinput of the op-amii So this lmA flows via Rf (2k0)whiih thus has 2V'across it.

Notice how Rf and Rin behave like a seesaw as theinput goes from a positive to a negative value, with thepiVot at the null point X. This point is referred to as avirtual earth. There is no current path between point Xand earth, and point X is always at zero voltage withrespect to earth.

The concept of a virtual earth is used as a short-cut

- " = ; ; - ; r ; 9 , ; : ,. : . , i : ; r i l

: I r . ; ; , . . : : ,1 i t . j. ,. . :..:.: . \

(ol {bt DSUd

Fig. 1: Precision inverting op-amp circuit, ia) witn a positive input, (b) with a negative input. Note how Rfand Rin behave like a seesaw as the input goes from positive to negative, with the pivot at the null (virtualearthl point X. The gain of the stage is RflRin, so the output is Vin x Rfl&in.

TELEVISION March 1999 317

ESR METER

Fig.2: Jim Williams'original circuit, the firctattempt at combining an op-amp with a Wienbridge network to form an oscillator.

656t1

Fig. 3: An op-amp Wien bridge oscillator arrange-ment with the output set at 6V p-p (positive peakshownl. At the resonant frequency points a, b and care in phase and the waveforms at the op-amp'sinpuF are a third of that at its outpttt. The ratioBflRin = 2.

when the operation of a current-to-voltage converteris analysed. From Fig. I you can see that, because ofthe virtual earth, Rf appears to be in parallel with RL.So the voltage across Rf appears across the load as theoutput voltage. But although the null point is cortsid-ered to be at earth potential, at a microvolt level it'svery much active.

It can be seen from Fig. I that the stage gain, withinthe limitations of trhe supply, is determined by theratio of Rf to Rin. Incidentally there's a frequencylimit on the gain: with common types of op-amp weare limited to a gain of about x10 at 100kHz. If theresistors in Fig. I are transposed the stage gain will be0.5 - the circuit acts as an attenuator.

OverviewBefore we go further, it would be as well to provide aquick introduction to the meter circuit presented here(see Fig. 5). The first stage consists ofa l00kHz oscil-lator, whose output is fed to the capacitor being test-ed. Put simply, the current flow through the capacitoris sensed then amplified as a voltage. It's finallydetected and measured by the meter movement.

The better the capacitor, the lower its ESR and thehigher the meter indication. It's not quite this simple,because the meter must ignore the other componentsconnected to the capacitor being tested. We'll come tothe solution to this problem later.

The Oscillqtor - HistoryAt the heart of the meter there's a Wien bridge net-work oscillator. This form of oscillator has an inter-esting history which is worth a few paragraphs.

In 1939 William Redington Hewlett (co-founder ofHewlett-Packard) produced his Stanford thesis A NewType Resistance Capacity Oscillator.It made use of aresonant RC network that had been conceived by MaxWien (pronounced Vene) in 1891. The Americaninventor Lee DeForest (yes, we can blame him) had-n't started the ball rolling yet with the creation, inI 906, of the triode valve. So there had in I 89 I been nomeans of obtaining electronic amplification and Maxcouldn't have got his network to oscillate. Thatwouldn't hav'e troubled him, as he was using the net-work for AC bridge measurement. Amazing what peo-ple got up to over 100 years ago, isn't it? I think itwas, once again, something to do with telephones.

But Hewlett had the pentode valve at his disposal.He also had Harold S. Black's pioneering work onnegative feedback to assist him. In addition there wasNyquist's Regenerative;Theory, which described theconditions necessary for. oscillation.

Hewlett showed that the Wien network could bemade to oscillater A crucial problem had to beresolved however, that of stage glin. With a gain ofless than unity there would be no oscillation. With again of greater than unity there would be distortion.With unity gain there,will be what Hewlett wanted, a

,sinewave. He had a flash'of inspiration; the solution'was literally staring him in the fate - the electric light

bulb. :Hewlett's oscillator.was a two-valve affair, with a

6J7 as the oscillator and a 6F6 as the output stage. Hissolution for gain stability was to wire a tungsten bulbbetween the cathode of the 6J7 and earth. The nega-tive feedback was applied between the anode of theoutput valve back to the cathode of the triode oscilla-tor valve. If the output increases for any reason, sodoes the current flowing through the bulb. As itwarrns up, its resistance increases. So does the level ofnegative feedback, thereby stabilising the oscillator'soutput. Hewlett's idea of employing a light bulb wasbrilliant in its simplicity. It survived in the Hp200series audio oscillator during a fifty-year productionrun - into the mid Eighties.

About fifty years after Hewlett built his oscillatorJim Williams, who was working for LinearTechnology Corporation, was sitting in his den onerainy Sunday trying to think of something to do. Hisold HP200 caught his eye. Peering into the back, hesaw the light bulb where it had been placed half a cen-tury ago, and wondered how Hewlett's oscillatorwould perform using a modern op-amp. He went on toknock one up - the original circuit is shown in Fig. 2- and was pleased to find that it had a distortion fig-ure of only 0.0025 pei cent.

Perhaps he could improve on it, by eliminating thebulb? Jim was the first to use a JFET in place of thebulb, but with this device the distortion figure rose toa massive 0.15 per cent. Unfortunately there's notspace to explain why the use of a JFET gives suchinferior results compared to a bulb. In the event Jimdiscarded the JFET in favour of an optically-drivenCdS photocell. This, in conjunction with five op-ampsetc., produced an analyser-limited distortion figure of0'0003 per cent (three parts per million). At one pointduring his quest Jim writes (Analogue Circuit Design,Butterworth-Heinemann) "I could almost hearHewlett 's l i tt le l ight bulb, which worked so well,

i.:l"!t''{

+if

,i$,.**;

#;ffi

3 1 8 March 1999 TELEVISION

#

'i{ii;r-.": : ;

laughing at me". So no apologies for the use of a lightbulb'in this design-

Operotion of the OscilloforFig. 3 shows the Wien,bridge netwoik oscillator asyou probably won't have seen it drawn before. It.illus-trates the situation at the peak of the positive-goinghalf cycle. The positive feedback network consists ofthe series-parallel RC (lead;lag) network: the negativefeedback loop consists of the preset Rf and'bulb Rin.

We'll consider the RC network first. At very highfrequencies the shunt capacitor in the lower arm of thebridge will appear to be a short-circuit and there willbe no signal at the op-amp's + input. At very low fre-quencies the series capacitor will appear to be open-circuit and again there will be no input from the feed-back network. At some point in between there will bemaximum output from the network. The frequency atwhich this occurs is equal to ll(2nRC), which is calledthe resonant frequency (fr) of the bridge network. At

' this point there is no phase shift across the bridge, andthe upper arm of the network has twice the impedanceof the lower arm, giving a transmission lgss of 1/3. Toovercome this loss and achieve the required stage gainof unity, the closed-loop voltage gain (ACL), which isset by thc ratio of Rf to Rin, must be three. The for-mula for the closed-loop gain of a non-invertingamplifieris ACL = M/Rin + 1, so Rf/Rin must be twoin order for ACL to equal three. , ....

At power up the negative feedback is low, becausethe bulb is at its lowest resistance, and.the gain ishigh. As,a rcsult oscillation begins immediately, aud

' the bulb is warmed by the current current flow. Withina fraction of a second the resultant increase in itsresistance reduces the oscillator's output. It settles atthe level at which the bulb's resistance is half that ofthe feedback resistor Rf. So the value.of Rf sets theamplitude,of ,the"output.,Note that the bulb's thermaldelay means that it cannot follow oscillations at rela-tively high frequencies. It.responds to the RMS.cur-

. rent,gnly, and phus lehavps,F a.n ordinary resistor. . i ;g- - b r ' . , i ] ; , t . r , f + ' i , . ' . . + , j r , y . ' , j : " { " .

'

The Bulb ,: ' .: ' .. Although the Wien bridge oscillator is the accepted

statrdard'at frequencies up to say lMHz, the use of abulb for gain control, popular in the USA, has neverfound favour on this side of the Atlantic. I think I.know thc rcason for this., In most textbooks thingsbegin to get a bit vague iwhen it comes to the actualtype of light bulb to use.

It is often said that any low-yoltage, low-currentbulb can be used. This is not so. I have seen the fol-lowing flawed reasoning in some books. Take a l2V,50mA bulb-which has a resistance of l2Vl50mA =240Q. The feedback resistor must be twice this, i.e.480Q or a lk() preset. There's nothing wrong withthis value for the feedback resistor, but it won't workwith such a bulb. The point that's been missed is this:the bulb must be operated at a current level that givesa large change of resistance.

This occurs when the current is only a few mil-.liamperes, and nowhere near bulb incandescence. Whatwe require is a bulb that has a resistance of about 200f1when cold. When the type of bulb normally specified isused, the result is overloading of the op-amp, distor-tion, heavy current drain and dependence on the supplyvoltage for regulation rather than correct bulb opera-tion.

I didn't do what Hewlett did, which was to plot the1/ characteristics of various bulbs carefully. I simply

TELEVISION March 1999

Fig. 1: Pr*ision rectifier circuit, (al with positive input, bl with negativeinput. In (aJ the op-amp's orrtPttt goas as low as required to overcome theforward voltage drop acrosli D'| and still satisfy Ohm's law as far as Bf andR|n are concerned. D2.is off as the voltage at its anode is 2'6V le*s thanthat at iE cathode. In (b) D.l is off, its cathode voftage being 0'6V higherthan its anode voltage. The conduction of D2 limits the positive o,ftptlt 8t0.6V. This limiting factor speeds up the recovery of the op'amp when theinput goes positive again.

I ; : ' r . ; l ; : : i . . ' '

measured the resistauce of pulbs that I thought mightbe'iuitable, and found that the cold resistance of a28Y,24r.A hulb is l70Q-. fitis soemed to be aboutright. When I tried it -,bingo! So when, in this con-nection, you see "any low-voltage, 50mA or so bulb"you can in future read "a28Y,24mA bulb". The oscil-lator,will work a treat.

The Prbcision RectifierThe final stage of the basic meteruses an op-amp as aprecision rectifier..Keeping to the type ofrepresenta-tion we've used before, Fig. 4 shows its method ofoperation.

With a conventional rectifier there's the drawbackthat the signal must rise above the diode's forward-voltage drop before conduction begins. This can beovercome by the use of an op-amp in the circuit. At (a)in Fig. 4 the input is positive and the output reducesthe voltage at the cathode of Dl. This enables theinput to carry on via Rf to the amplifier's output. Asin the case of the inverting amplifier circuit, the out-put is again Vin x Rf/Rin. The diode's forward volt-age drop, which is 0.6V with a silicon diode, is over-come because the op-amp's output goes lower by thisamount, satisfying Ohm's law as far as Rf and Rin areconcerned.

Point X is still held at earth potential by feedbackaction from the output. D2 is off at this time, as thevoltage at its anode is lower than that at its cathode.When the input goes negative however, as shown at

Yflt - 0v

D5UO

Vioa - lY

3 1 9

ESR MbTER

Fig. 5: The basic meter circuit. VBI sets the oscillator's output level. Pin 8 of tCl and tC2 is connected to the +ve suppty, pin1 to the -ve supply.

(b), the op-amp's output rises to the point at, whichD2 conducts. The current then flows via Rin, point XandD2. DI is now off and the output ib zero.-

Bosic Meter GircuitThe circuit of the meter itself is shown in Fig. 5. TheWien bridge oscillator, redrawn, is the same exceptfor the inclusion of a lf,l resistor (R3) between thebulb and the 0V line. Depending,on VRI's setting, thebulb's current ii typically 3.5m4 RMS.'As a resriit, inthe absence of a capacitor under test about lOmVpeak-to-peak at lOOkHz is developed across R3. " ' '

VRI sets the amplitude of the oscillator's ouq)ut. Inthis case the butput is used only for feedback,ind isset at 5V feak+o-peak; There is nothing magicalabout this figure, and with this application no testequipment is required to set it. It's just that'to get ahigher level output you would have to use a highersupply voltage. In fact however the higher thb outputvoltage the better.'

The ESR of the'capacitor being tested forms part of -

a potential divider with thez.lQ resistor R4. The volt-age waveform across this resistor, as a result of thecurrent in the capacitor, is amplified by the rest of themeter circuit. Bear in mind thit with the range of ESRvalues we are measuring an ideal mid-scale figurewould be about 3Q. With low. ESR'value'C (goodcapacitor) the signal across R4 is high, while with apoor capacitor it wlll be low - often, in relation to2.7d2, there can be an effective open-circuit.

Now if, for example, the ESR is 2.7Q, half thesource voltagri across R3 would be passed to the meterand a half-scale reading would be expected. It doesn'tquite work out like. this however, because the sourcevoltage is not independent of the lohd, and we will besetting full-scale deflection with R3 and R4 in paral-lel (test leads shorted).

If the ESR tends to go below the value of R4, itbecomes more effective in increasing the voltageacross R4. As the ESR rises above the value of R4, itbecomes less effective at incrEasing the voltage acrossR4. Hence the non-linear scale, which is ideal withthis application. R3 and R4 are of necessity low invalue, because they compare with the values of ESRin which we are interested..The bonus here is thatbecause of their low values thi effect of associated in-circuit components becomes insignificant.

The design of this littlernetwork is such that the

waveforms across R3 and R4,are virtually in-phaseregardless ofthe value ofthe test capacitor. So we aremeasuring the total in-phase AC resistance to whichDoug Jones refers (see quotation earlier).

You might wonder why the test.signal amflitude isso small. It isn't because rfe warit tg avoid turning onsemiconductor devices - we could go up to a'coipleof hundred millivolts before there'would- be:any wor-ries about thit, It's'similly'i"mattef .bf power con-rsumption. Even our tittle iOmV requires 3.5mA, irndin this case I have (dare I claim cleverly?) used a cur-rent source thdt's already there. A l00mV test sourcewould require a hefty 35mA, quirc a drain onresources. If anything the value of the lQ resistorcould be even lower, so that with respoect to2.7Qitwould more closely approximate a constant-voltagesource.

You may think that to test an electrolytic capacitoreffectively a fair old currentishould be pumpedthrough it. Not so. A healthy l,Q00trrF capacitor willstill present 0.05C1 or so to a couple of millivolts andthus be produce a reading. :1 ': '

The signal across R4 passes through two stages ofamplification each with a gain of ten, and is thendetected for the meter movement. There is furtheriamplification in the detector stage. The output iS inte-gated by C4 to produce a DC output of about l.3Vwith the test leads shorted - this corresponds to zeroE S R . , i : . r . : '

The basic meter circuit uses two dual op-amps. you'will see that the signal path from the oscillatoi in ICIpasses to IC2 then back again. fhil is done to preventthe first, sensitive stage of amplification picking up astrong oscillator signal in the same package.

The Power SupplyThere is no need for a regulated supply, because thebulb stabilises the oscillator and the implification fac-tor of the op-amps is fixed by the ratio of the feedbackand input resistors.

_The power supply arrangement used is shown in Fig.6. IC3a generates split rails from a single supply line.The voltage at its output pin 7 is at half thC iupplyvoltage, because the voltage ar its - input (pin 6) isequal'to the half-voltage level set by R[2 and Rl3 atits + input- (pin 5). This way of using an op-amp is'known as the voltage-follower. There is total negaiivefeedback, and the c-losed-loop gain is unity.

\i. lii

,+,.1.

320 March 1999 TELEVISION

$iq

- i r '

_ - ; . , 1 , i ,: ; : . i :

Fig. 6: The..iplit-rail generator and buzzer comPdrator circuits'

i ? t

ESR METER-

The meter's total current requirement is only somel0mA,.plus a ,couple of mA for the on indicator D6-Two PF3 batteries in..series are ideal. Long life isassured -. if the oscillator's output is, set-as described .later, the meterls accuracy will be maintained until thesupply drops to-about 5V per battery.

ti tfrl fnt Uetween the batteries was connected to the ,0V rail this split-rail arrangement would be unneces-.sary. It's included to enable a DC adaptor to be usedas

-an alternative power source. An adaptor with an

output from,I2V to 30V can be used. A regulated typeis best, as ripple on the supply c.puld cause problems.

i , , ' ' l i t '

The BuzzerlC3b serves as a comparator forbuzzer operation. The

output froin the meter rectifier circuit, across C4, isappiiea to the + input (pin 3) for comparison with thevbitage,at the - input (pin.2). If the voltage 1t pin 3e*""Jdi that at pin 2; ttre output at pin I goes high (see

comparator ciriuit.description earlier) and the buzzersoun-ds. About lV is developed across the series-con-

inecied diodes.D3 and D4.rWhen the ESR value of thecapacitor being tested is about lO or less, the yoltageacioss C4 rises above this lV reference.

Next MonthIn Part 2 next montf, we will deal with construction,settins uD. use and inductance measurement, and in'additi-on

brovide a bit more information on ESR. Adetailed cbmponents list will be included.

R1(55k

2 2

o l

oa

TELEVISION March 1999321

.7t Suitable'ier scale, ,relpro-ced full size.

I n pu.t I lqt month the design criteria were spccifiedI and a full description of the operation of the circuitf was provided. The meter i3 simple to build and theeffort rcqufued is well worthwhile. , : :::.:

ConstructionFig. 8 shows the board layout. For convenience,'0.lin.matrix stripboard is used. The most common problemqo.l9.".Ts- ilr9 print cuts. If these are made by rwisting ad.itl lit by hanp, as I do, instead of usingthe correcttool, it's all too easy for the cut to be incoirplete orforsome of the print to spread over qo an adjacent track.

If a double-pole on/off switch is used, ihis is a cbnve-nient point for the connection of an on/off indicator. Aflashing LED with a l0kQ series resistor does the jobwe-ll: althoughit takes only a couple of mA, the flashinglight does catch your eye.

It is best to use screened cable for the test leads, inorder to avoid pick-up of unwanted radiation from any

In fhis second pqrt AlonWillcox deqls, wirh construc-fion ond setting up of rhemeter, upgrcides qnd use,ond provides oddirionolbockground informotion

'blerworkirg.TV line output stage in the vicinity. Rather thanuse plugs and sockets, solder the test leads to the ICB:we are often concemed wittr RSR values of less than0.1O, and it doesn't take long for a plug-and-socketconnection to deteriorate and produce resistance.valuesgreater than this. The length ofthe leads is not at all crit-ical. I use a 24in. length of screened audio lead termi-nated by two 8in. lengths of flexible wire.

On- a cold, frosty morning, before the workshop hasreached its normal comfortable, cosy (?) temperature,$9 leter's readings may increase,slightly. Alttrougtrthis_increase may amount to the equivaleniof less thin0JO, the fact that accurate low-rciistance readings areoften required justifies the use of an off-the-board set-zero control rather than a preset type as used in theWizard ESR meter.

This control also comes in handy if you want tosqueeze the last ounce out of the batteries, but thebuzzer can't be relied on when the batteries produce lessthan 5V T9 th" output from the oscillator begins to fall.Also useful in this respect is the ability to alte; the point-er position easily at maximum output, to be able toobserve the operation of the bulb and the correct func-tioning of the oscillator. This is covered in the set-upnotes later.

Probe clips were used for the all-important probes. Thehooks were cut off, the internal springs were removed,then the plastic was cut back to expose more of theprobes. They have proved to be ideal in practice, and aresmall enough to often be able to get under the capacitoron the component side of the PCB. The probes are notpolarity conscious, and it doesn't matter which connec-tion is used for the screen ofthe test lead.

The suggested board size (3 x 2in.) is slightly larger

7 5 4 3 2 I.5

ESR a l 1OO kHz

426April 1999 TELEVISION

ESR METER-

. . r . : : i i . i + r , . . . ._:{:f '

than it needs to be to.dccdmmodate the components.There arc two reasons ifor this: First, so that it will slotinto the case that was selected, and secondly to allowspace for future upgrades. With this in mind the set-zerocontrol is offset to the right, allowing space for a modu-lar preset shaft.

The batteries are securcd with self-adhesive Velcro.It's a simple solution, and the Vrlcro will transfer sev-eral times.

witli theWizard this meterdoes not dif-feientidte between a $hort4ircuit capac.itor and a really good one (with an ESR :arc-und 0.050). Now although we allkiiow ' that a short+ircuit electrolyticcapacitor is quite rare, it ciio'occur andhas caught me and others out. Allot oftime can be wasted when it occurs and ismissed. So I've been testing an additionthat gives an audible indication when ashort-circuit is present. A further refine-ment is an auto power-off facility.

I've not included these extras at pre-sent because I feel it best to present assimple and economic a pdect as possi-ble initially. But space has been left forthese additions.

Gomponenl SourcesI obtained most of the pans used in theprototype from Maplin Electronics andhave included this company's part num-bers in the components list. The buzzerand meter movement were obtainedfrom CPC. For correct operation of theoscillator it's vital that Cl and C2 aregood-quality polystyrene capacitors.

TELEVISION Apr i l 1999

: . ' . , .

Apart from the bulb, the sp&ifrcation 6f ttre other com-ponents is not critical.

The Mefer ScoleThe geometry of the meterls scale (see'Fig. 7) is dictared by the relative values of R3 and R4. If the value ofR4 was'inbreased to say 10O, this would be the mid-scale reading and R3, being low in comparison, wouldhav,e less effecl Say a capacitorwith aerESR of l0O isconnected. Half the signal ai.ross R3

-iriil! .be passed on,

andrtittb'6f ,its curreni will,be' diverteidt6,.reduce R3'svoltage;.rvhich is the result of i6i ibtii.Ent through the. . . 1 . . ! , . . . i r . . ;

' l * i r * , - . ' . " ' :

X - l rocl cut!

Fig.8: Meiel.circuit layovton stipboard.Board size is 3x 2in. lo allowfor upgrudes.See jex].

continued on Page 436

. 4,27

:, l'n

ffi

lnlernal view oJ ihe meter.

ESR METER

bulb' A mid-scale reading of this order may be more Leave the set-zero control at minimum and observeappropriate for testing surface-mounted electrolytic the-operation of the bulb uy ttit.ting on with thecapacitors, which seem to have higher ESR va.luei. ! qrgbei shorted together. vou *iir see the pointerhave no data on this type of capacitor at the time of deflectrrlcrrtrren"lil;;ir";;;ui"iy.;".pdownrothewriting however. stabilised"position.

Anyway, the value of R4 used here,2'7{1, does affect Once the oscillator's output has been set, adjust vR2the source vciltage to an extent that has to be consid- for full-scale oenection w-iirr i;;;.;;, shorted togeth-ered. The current that -flows through the bulb is con- er. ninaity, centre the control knob.trolled_ by the setting of VRl. R3, being only le, doesnot af-fect $e operation of the oscillatoi. Using the Meter

scale calibration with an 0-100 dial ideally follows nxperi-ence with the meter provides the best guide as tothe rule wh;t ESR value to "ip""t oi ;;;; capaciror. Some

Reading = (R3 + R4v(R3 + R4 + ESR) x r00. ffiT*iJ,iff:il#t,::tiT'l#Y"""",T#"J,".TTi';In practice the signar becomes s9 rory at high ESR var- ;#i%:!E:X'r?l3lY"'[119,1*ft,:"d:d#ti":il"#ues that some non-linearity in the cjrcuitis apparent, electrolytic capacitor_.of t00pF oi-over value, youthe result being a-slight dlparnrre from this rbrmura. itgtJ^"Id a reading close to this and cerrainlyBecause of this, it's better to use either fixed resistors below 0.5h.to calibrate the scale or copy Fig. 7. If a standard 90' There * ,orn" exceptions however. I,ll mentionmoYement of a different siif is used, it is easy enough trrem trere as they can iuur" "ooruri;;. A^irih*;lsR-toreduce.olenlargb-thqscalebyphotocopyinE. value is to be eipected yrth ltF, ligh-vorh;;ei;-

letrlnS Un .::', '-The first consideration is. the oscillator's-output level. rtutt-up "o.pon"ni. n"iog 6p"raiJin a pc context, itswith the specified lamp, the oscillator wil wbrk down Fsi iJil;ilp"rr-r 4t an e-xample a new 450v.rypeto JV peak-to-qeak But bear in mind that the oscilla- "tlt ;*;;;Uuce a reading of the brder of:30f,I At thetor's output setring is not importantg ilryt{ . tru--scale , otniq eqo ir rhe.qnge. t9u. Eetrnp"r,oobpr,deflectionissimplyadjustedby-vR2.wit_h.thelestleads:ryqgir;;ortnrriao"*i:prpioJ.ur.adingof0.lesho{9!together'I,owsetting]ofthebircilIator:soutput<iireisiisau}f--:-:.i---:'.l1tr+-.lI�provides a longer battery-life,'iieduging the livel it Table l, from Dubilier, prwides a useful guide, inwhich the bulb ceases t9 function al.1 qery[tor. Stre addition, ir ytu ar" unsure what reading to expect it,stest signal can however be so low at highEsR values "ury "nougtiio *ut"'u "o-f*iro" *iiri",i-iili;;that accuracy suffers. In practice an oscillator output of rr9- yoor?"il-stocked luy-J"r""t oiyti"r.about 5V peak-to-peak is a good com-promise.

- The-meter has good protection against placing therf an oscilloscope is not available, the following sim- probe, u"ros u .f,.g"d .upu"it-, But this should ofple method can be used. Connect a 1o resistor beiw""n "ourr" u" uuoia"a. rfr" onry ieui Juut" for concern isthe test Ieads and turn the set-zero control VR2 (shown . the .'main smootfring block,, i.". u *an,

^uie",jiiig;:

incorrectly as a preset iri ric, .s* fas1, -g-.tql [ti"r*t: "i;t il*]l'ii-""paciior. we att tnow tr,at nis capacitor.wise. Advancb the sctting of wil slowly. Iust:after the . can abtlnb; ".rrif1f pun.l *n"o " piLr-suppty has,:half way point the oscillator will start up and the meter's faibd l; siilt up. If you are-of the type with a tendencypointer will deflect. c1rry oq.yntil th-e-buzzer iounds ro such acciJent, iiwour,l ue fruiiiito wke a couple:steadily. Ar this point the oscillator's output should be E F"fy ;F;; 6ack-to-babk across the merer,s inpur.,,cloqe to 5v.peak+o-peak. Note that with each adj.ust- rrris stiouta liotect the meter, but it won,t help yourment the pointer will twitch briefly as the bulb se[des, heart - ot yo'ot probes - in the event of such a misfor-giving the impression that the potentiometer is no5r, tune.rt's gLoJpractice t" al*rr*ga irtit capacitor with

continued from page 427

Table 1: Typical ESR values at 100kH2...I

Capacitance value Voltage rating lmpedance*

ltrF2.21tF4.7pF10pF221tF471tF471tF100pF100pF2201tF2201tF4701tY4701tF1,000pF1,000pF2,2001tF2,2001tF

63V50v2 5 V '50v16V35V16V35V16V35V16V25V25V,35V

4Q2.8A2.4A1.9c)1.3c)1.3cro.7a0.scr0.2scr0.25fi0.1 14Q0.114c)0.065fi0.06500.041c)0.036ct0.034f)

50v50v50v

*When new - low-ESR type.

a mains-type bulb - it is always reassurine to have avisual indication that little charge remains. See warningnote later.

InductonceAs the meter operates at l00kHz, any inductance in thecircuit under test becomes significani. The loudspeakerthat reads.8C) on your coqventional ohmmeter app"*sto be nearly open-circuit wittr ttre ESR meter.

This propert_y can be quite useful. When testing a lineoxtput stage for example you might find that there,s ashort-circuit across the hansistor. In this event there arein general three possibilities, (a) the transistor is short-circuit, (b) the line output transformer has a primary-to-secondary short, or (c) the HT line has a short-circuitacross it. If the short-circuit is still present when theESR meter is used, the transistor is almost certainly theculprit: in the other two cases the ESR meter will givean open-circuit reading because of the inductance of thetransformer's primaly winding.. Video head testers operate by measuring, in effect, theinductance of the head. Its impedance falls as the gapdeteriorates with wear. Althou;h this meter operates ata frequency that's inappropriate for a video heid, it can

436Apr i l 1999 TELEVTSTON

'don1t have,a collection of heads in various states ofhealth to be able to confirm any figures, but it seemsthat new heads produce a reading of about 20 whileworn heads produce a reading ofabout lf).

Inductors with millihenry values, of the type used inEW modulator circuits, normally give an open-circuitreading. But ifa shorted turn is present the inductancedrops dramatically and the meter's pointer will deflect.

Anotogue v DigitolIn dealing with the problem of in-circuit ESR measure-ment I've used traditional analogue technology. But, asis often the case, there's another way of doing things.Bob Parker, an engineet,fdown under' who was con-vinced of the importance of such an instrument, hrsttried analogue circuitry. After "a few fairly unsuccess-ful attempts" he opted for a digital approach. His solu-tion'is to use a 7)log processor,with, instead of a,sinewave as a test signal, short current pulses applied tothe capacitor being tested. The resultant voltage pulses,which are.proportionil to the electrolytic's ESR, arecompared to the level existing on a fttmp generator.Time measurement by the 286 processor determinesthe amplitude of the pulses..As far as the power,requirement is concerned, tliere's

a parallel in the form of a remote-control handset. TheLED is pulsed with a high currentfor a short period, theaverage curretrt drawl being low. The brief high cur-rent is supplied by a reservoir of stored power in an .electrolytic capacitor that's wired across the battery'sconnections.

give a irelative indication of the head's and considering the fact that ESR is an in-phase com-poient, I have made the assumption in my calculationsthat ESR amounts to the same thing as an equivalentfixed resistor.

Other uses for the meter emerge the more it is used.For'example, non-electrolytic capacitois can be mea-sured and their capacitance estimatbd. But because of

' the different types of capacitor construction in use, Ican't coine up with any hard and fast rule. A lower limitfor measurement is about 0:lFF. There seems to be lessand less need these days to measure the actual value ofa capacitor. With line output stage tuning capacitorsand timing components 'a conventional capacitancemeter is more appropriate.

The ESR meter is one of those things that, once youhave one, you wonder how you ever managed withoutit. Hats off to whoever came up with the idea - it's notmine. I've just taken this opportunity to share with youthe course I adopted to end up with the solution pre-sented here. I haven't clapped eyes on the Wizard yet -we can't afford one down here in Wales. I'd like toknow how the Wizard designer approached the prob-lem, but no information has come my way.

The idea for my meter was triggered off by theWizard. When I first read about it I was impressed. Itwould test that most troublesome of all components, theelectrolytic capacitor. Not only that but.it would do soin-circuit, ignoring associated components. t liked theidea of a conventiohal meter movement with its easv-'to-interpret

scale, also the buzzer feature for quickchecking. But the meter was shrouded in mystery andits price tag was beyond me. So I decided to have a gomyself.

DevelopmentA clear picture formed in my mind as to how to goabout it. My idea was to supply a low-valuc resistor(l0Q) with a constant-current l(X)kHz sinewave,amptify and rectify the resultant ;voltage waveformacroSs this resistor and feed it to h meter movement. [nthii situation thb meter is to be set at full-scale deflec-tion. The test leads were to be connected across theresistor. Now ifthe capacitor being tested has an ESRofsay 10f,1, the voltage across the resistor would fall byhalf. Thus half-scale deflection would correspond to anESR of 10Q and so on. A very low ESR (good capaci-tor) would produce a near-zero reading while a poorcapacitor would have little effect, the pointer remainingat near full-scale deflection. The result is a meter scalein opposite sense to that of the Wizard.

Within a couple of months of the start of developmentwork on the project I had a working prototype anddecided to write an article about it. Then, at theeleventh hour, I had second thoughts. Something wasnagging me. The meter looked OK, did its job, and oth-ers were happy with it. But after using it for some timeI felt that something was not quite right.

My main concern was that it seemed to be very sensi-tive to the inductance of the test leads. This impedancemeant that even with the test leads shorted the readingcould not be brought down to zero. I had introducedmeasures to offset this, but in this connection a moreserious problem emerged after some further use.

When a capacitor with a very low (near zero) ESRwas being measured, the pointer would vary about thezero point if the distance between the test leads wasaltered. If they were close together, their inductancewould tend to cancel and the reading would decrease. Iwas aware how important it was to be able to differen-tiate between say 0.5O and 0. lfl or between 0'lfl and

Friendly rivalry between the analogue and digitalcamps has existed for a long time. This brings to mindan old story that sums it up well. Two male engineers,

. one specialising in digital design and the other in ana-logue design, are working together in a lab. A nude

.: femalc appeel! at the door, attracting the attention of,, both men. This vision of beauty announces.that every -trlten'ieconds she'will reduce'the distance betrveen hei-r::l'self and the engineers.by one half. The digital engineer,r:lool3S disappginted and cries "that's terrible, she'll. never get here". Xhe analogue engineer smiles and thenrcplies i'that's OK she'll get close enough".

More on ESR,@ob Parker (E/ea ronics Australra, February 1996) putsit this way.

"The electrolyte has an electrigal resistance which,along with the (negligible) resistahce of the connectingleads and aluminium foil, forms the capacitor's equiva-lcnt series resistance. Normally the ESR has a very lowvalue, which stays that way for many years unless therubber seal is defective. Then the electrolyte's watergradually dries out and the ESR creeps up with time.The electro gradually comes to act like a capacitor withits own internal series resistor. . . Heat makes it worse.If an electro is subjected to high temperatures, espe-cially from heat generated internally as a result of largeripple currents, the electrolyte will start to decomposeand the dielectric may deteriorate - the ESR will thenincrease far more rapidly. To make things worse, as theESR increases so does the internal heating produced bythe ripple current. This can lead to an upward spiral inthe capacitor's core temperature, followed by completefailure - sometimes even explosive , . ."

Both Bob (Dick Smith Electronics) and Ray Porter, inan article on ESR meter design in this magazine a fewyears back, mention the use of fixed resistors to assistwith meter calibration. Armed with this information.

TELEVISION Apr i l 1999 437

ESR METER

zero. This.was a design weakness, and I was not happy.At this time Bob Parker's K7204 ESR meter became

available. I had one with which I could make a com-parison, and noticed the same sensitivity to inductancethat I was experiencing. Both meters monitor the volrage across the capacitor, whereas the Wizard measuresthe current through the capacitor - this was clear fromits scale.

I went on to build a little circuit to do just that and, loand behold, this over-sensitivity disappeared. The effec_tive lead impedance dropped dramitically, and thatwhich remained could be iimply compensated for bymeans of the FSD preset. So it seemed that the Wizardyay yas the right way. All rhis can be explained usingthe relevant maths, but I'm not about to brush up on mythcory and extend this article longer than it is aieady.'. Reluctantly, I-accepted that I had to scrap my originalideaand starr all-over aggl Thir was by nbw becoilingan obsession with me - I had to come up with the besisolution, and it had to be .the simplest.

Certaiir things had to go. After s-ome use I realised thatqlugs and sockets for the test leads were out ofthe ques-tion. Their increasing contact resistance made low-o-hmsreadings unreliable. I had also been determined that themeter should operate with a single pp3 batterv.

To this end I had built a DC-DC converter io providea negative supply line. This provided operation .iown to7V. But it used 7mA and geatly increased the circuit

complexity. To do away with it meant that I needed ahigher supply voltage. The use of two pp3s in series toobtain this may seem to be a backwards step, but isn,treally. The meter now takes half the current used by thefirst prototype, and will regulate down to less than 5Vper battery. In,fact the T:t9llgadinC remains unchangedover the supply range l0-30V with the oscillator set at5V peak-to-peak.

Now,- almost a year after building my first prororypeand quite a few versions later, I have presented'this _;yfinal (?) solution!

WorningIt seems that forgetting to discharge the main smooth_ing block or HT reservoir capacitor before measure_ment is a more frequent occurrence than had beenexpected. When this happens R4 at least will blow. Asreplacing components, on stripboard is a messy job, Istrongly recommend that protection is built in. Aimallboard with two 1N4007 diodes wired back-to-back anda lA (N25) circuit protector in series with one of thetest leads can be mounted on the back of the metermovement.

ThonksI have to thank Martin Pickering for his constructivecomments after testing an earlier, voltage-sensing proto-t)?e meter.

Components listItem

R 1 , R 2R3R4R5, R7, R9R6, Rg, RlsR 1 0R l 1R12, R13, R14R 1 6VR1VR2

*Value for current,economy

C1, C2 47OpF,1% polystyreneC3, C4, C7i 0.1!F miniitur'e ri;sin dippedc5, c6 22 iF ,16V

[c1, lc2, lcaD1,D2, D3, D4, D5D6

LP1

M 1

SW1

BuzzerStripboardCaseKnobBatteriesProbe clips

Value/type

3kQ,1yo1 Q2.701oka100kf)

- 91kO5.6kfi56kO2.7K1or 10kfl*500Q cermet preset1 0kCt linear potentiometer

TL082CN ,1N4148Flashing LED plus cl ip

Miniature wire-ended28Y,24mAlamp

100pA meter movement

DPST switch

Miniature alarm

ABS box type BM22

Two PP3s plus clips

Order code

M3KM 1 RM2R7M l O KMlOOKM91KM5K6M56KM2K7 or M10KWR39JM71

BX53RA49VH09

RA71oL80OY96 and W40

BT44

CPC code PM11125

R D 1 7

CPC code LS-M3JP47cc83YXOlHF28HF21

Order codes are Maplin,s unless otherwise indicated.

438Apr i l 1999 TELEVTSTON

OHMS

-.oix7 5 4 3

compa re

ESR at lOO kHz

The equivolent series resistqnce (ESRI of qn electrolyticcopocifor is o relicrble indicqtion of its condifion.Alon Willcox's firsr ESR mefer".detign, published in thesepqges some five yeors a9o, wqs very populor. This Mork 2version incorporoles severql improvements, in porficulor qsimpler oscillotor design ond single-botfery operotion

simplified equivalent circuitof an electrolytic capacitoqsee Fig. l, is usually shown

when providing a briefexplanationof what ESR is all about. In addi-tion to the ideal capacitor Xc, asecond component is present. It hasa significant effect on the capaci-tor's performance, and is referredto as the equivalent series resist-ance (ESR). Electrolytic capacitorsare used mainly for decoupling and,to a lesser extent, for signal cou-pling. This being so it's importantthat, for optimum performance, theimpedance (AC resistance) of thecapacitor is as low as possible.

An electrolytic capacitor's ESRis mainly determined by the condi-tion of the electrolyte (paste) thatseparates its foils. The electrolyteincreases the component's capaci-

Fig. 1: Simplified equivalent circuitof an electrolytic capacitor. Xcrepresents an ideal capacitor. ltsreactance moves towards zeroohms as the frequency increases.Xc = l/(Zt fcl. The value of theequivalent series resistance ESR isdetermined mainly by the conditionof the electrolyte.

tance but, when it deteriorates, itincreases the impedance present.The lower the ESR, the better thecapacitor!

Electrolytic capacitors used inswitch-mode (chopper) power sup-plies and those mounted close toheatsinks tend to run hot. Heat isinclined to dry out the electrolyteand, in time, a capacitor may devel-op a high ESR. This will itselfintroduce a power loss - and moreheat! The effect offailure ofthistype in a power supply can be cata-strophic. For example, if the capac-itor is in the HT monitoring sectionof a TV set's power supply the HTvoltage might rise, damaging theline output transistor and maybe thefield output IC.

This type of problem is moresignificant when a set is started upfrom cold, as the condition of afaulty electrolytic capacitor isworse when cold. Thus cold checksare by far the best approach to test-ing! Problems with electrolyticcapacitors, particularly those usedin switch-mode power supplies, arecaused not so much by a change ofcapacitance value as by an increasein the component's ESR. Thusremoval of a suspect capacitor tocheck it with a conventional capac-itance meter is largely a waste oftime- Furthermore a faulty capaci-tor may be overlooked because itscapacitance value has hardly

changed. An ESR meter getsaround this problem by using a testfrequency (or pulse rate) that's highenough for the capacitive reactanceto be almost zero ohms, leavingjust the ESR as the measurement.

Meosuring ESRESR obviously cannot be measureddirectly, using a conventional ohm-meter, so a means has to be foundto 'get to'the ESR that's hiddeninside the capacitor. A number ofready-made meter designs and kitsare now available for measuring theESR of an electrolytic capacitor in-circuit, using cold checks. Theyachieve success and simplicity invarying degrees. Use of a suitablemeter enables cold checks to bemade without the risk of any fur-ther damage occurring. Strictlyspeaking it is the impedance that'sbeing measured but, over a particu-lar range oflest parameters, it canbe shown that this presents more orIess the same value.

It is all too easy to over-compli-cate things when it comes to ESR.There are those who argue thatESR meters don't measure the ESRprecisely. But, in the real world,how precise does the reading needto be? It really doesn't matter. Theservice engineerjust needs toknow, as quickly as possible, whichcapacitor is causing the trouble. AnESR meter does just that! Oncetechnicians have become accus-tomed to using an ESR meter, theywonder how they ever managedwithout one. Although the ESRvaries somewhat with frequency,we can in practice regard it asbeing a constant in-phase compo-nent, and calibrate our meter usingfixed resistors. This is useful, as themeter will also serve well as a low-ohm meter.

So. if an ESR meter doesn'tmeasure capacitance, what readingscan we expect from good and badcapacitors using such a meter?How do you know whether acapacitor is OK or not? The curveshown in Fig. 2 gives a practicalidea of the sort of ESR readingsthat should be obtained rvith goodcapacitors of different values.There is no hard-and-fast rule horv-ever - it 's not an exact science. Allit needs is a bit of getting used to.This doesn't take long: just meas-ure the ESR of a f'ew nerv capacl-tors. Try l, 10, 47, t00, 470 and1,000pF. You wil l f ind that valuesof 47pF and above measure quitelow. 0'5O or less. with the buzzercoming on (the buzzer turn-onpoint can be varied, see later). The

7 6 December 2004 TELEVISION

ES8(o)impo.rtant point is just how lowcapacito-is with values of 47 uF andabove measure i At 47018 and ovr.,the reading is close to zero'ohms.If, in practice, a 1,0004F capacitorproduces a reading as high as O.SSI,rt's no good! i:

Anqtd$i,i'ditiin"l ,d t s P l g ) B . r i , . - . . ,Digital..methods bf measurementand lisiildy are'generally regardedas pro,vldrng &ore accurate results.F9r mpny applications rhis is true.but itjs nqr:,45pessarily so wi*r gSnmeasureineriflThere ire two wavsof interfacing the,capacitor beinltested' eiffi the m6tir:ftre capaciicircan be connected in parallel-withthe test sign4l.g"ourqe, as shown inFis. 3(a);Wiii"i;?:ii6b', t#'rie. irbi.'

\_ If a digital ESR readout i; ii;ad.the capacitor must shunt the testsignal.,With an deal,capabito4 theresultrwill be:z0ro,ohms; an d.z&te.,.: : tdisplay:.-with a. badcapacitor therewill be d hieh rcading;with linles_h_untlng away of. the test-signal.We are interbsted in ESR viues of3Q or less,rbut a digital metergives readings far higher than this.It's not a major problem, but cangrve nse'to superfluous informationaS far aS ESRtmeasUrements are r_ jconcemed; A inore. important factoris ftat the:bqraltdmetnoa results ina signifiidrftlniiease in,tne read::^ :ings obaiiied&ecauiff 'of,ejri:i:Ssive

isensitifi.df0lttiri[dluctiriie,ofithe,,l$! le3+,r ffis".ii,ii' be.overcome tyfi tting.twollehds.tor6actlprobe:;It

-

provides.cancellation"toi,laige - .extent,;but:thissolution.is a U'it .. ,

rmsy in.r'se and'to conshuct. The""cries method ofiinterfacing does ,;.,not sufferrfrom,this.ilrawback; sothe use of;conventional test leads ,becomes' acceptablel ll don't knowyhy th i se f f ec toccu rs - I j us t . : ,found out the hard way. . .

Although we have 6ecome usedto_digital readouts nowadays, forESR measurement there is-littledoubt that a moving-coil meter isthe.best type of display. It gives araplo, ei$y-to-rnterpret indicationof the condition of the capacitor.After some experience of using it,oxe gets to know where the pointershould approximately be with goodcapacitors of different values.Indeed a meter scale becomesalmost unnecessary. I know ofpeople who have used this type ofmeter quite satisfactorily without

I t F *

,Qgq$anr;gl9se

qf9:3lgt.'Rsample. Value

cfiosen foroptimum

meter scale

ever having taken the trbuble to fita scale dedicated to ESR measure-ment.

To repeat: it's not an exact sci-ence but, with some experience,

Layoui o( Non willcox's Ma* 2 ESR meF,r on stripboard. n" n4z incorporatesseverol.improyements, including a simpler ""riil.i";-i;;b" i,ii;igb_botteryoperalion.

ptb?d anp

Fig' 4: The basic ESR meter circuiL new version. lcl and IC2 require positive and -negative supplies at pins g and 4 respeaiiely.Further circuitry is required, see nert month, to generate tne sim-riit iujpfu and provide a buzzer comparator.

one soon gets to knoy whfglr- - lly capacitance will provide negli_, capacitor is causing_ the qo.uble by .- gible opposition ficjto it.knowing roughly where the ESR'': :t '-meter's pointer should be with a Waveform: Analysis of a square_good one. one argument that's wave has been used, and l'vL trieasometimes put forward against the ,this myself. The results are a bituse of a moving-coil movement is unpredictable however. I had a lotthat the movement will be dam- ofiroubre trying to preserve theaged should the meter fall off the waveform intaci at the mv level tobench. I have a solution to this give a useful, predictable indica-problem. If you are inclined to be tion. It's not *:orth the effort. Aglumsy, attach a piece_of string sinewave is so easy to generate andbetween the merer and the bench, gnplify that I cannot sJe any.lustiIong enough so that the meter ncition for incorporating square_comes to-a halt just before it hits wave analysis into the dEsien of anthe deck but not so short that you ESR metei.

' ' do r i t t ge tab i t o fa jo l t t ose rvLas '117 ' l ' * "1 * . ( . .:a'ieminder. , ' -

An improved ESR. meterThr meter presented here is similarto the one described in theMarch/April 1999 issues ofTelevision. Since then the use of anESR meter to check quickly forfaulty capacitors has Lecome wellestablished. The original circuitworks well and has stood the testof time, but feedback from thehade has prompted me to makesome improvements. These includesingle-battery operation andimproved temperature stability.The new circuit remains stabledown to 6.2Y, and consumes aboutl3mA. Because of the improvedtemperature stability of thl newoscillator circuit, there is no needfor an externally-available set-zerocontrol.

The oscilloior cirrcuitAn IIF oscillator or pulse generatorwith a stable output over a reason-able supply voltage range is theheart of every ESR meter. ICla inthe new circuit, see Fig. 4, is thesinewave oscillator in this design.The simplest way of generating- a

sinewave.is to incorporate in theoscillator's positive-feedback patha network originated by Max Wienin 1891. It ensures that the feed-back is an in-phase component atonly one frequency, which is fixedby the rRC values used. At veryhigh frequencies C2 presents i low-impedance path, while at lower fre-quencies Cl becomes an effectiveopen-circuit. At some point inbetween there will be maximumoutput from the network Cl, Rl,C2,R2, which is refened to as aWien bridge. The RC values usedhere'result in oscillation at aboutlfi)kHz.

There are two simple methodsof stabilising the output level withan op-amp Wien-bridge oscillator.The method traditionally used is toinclude a bulb in the nelative-feedback path (between-pins I and2here). This was the approachused in my 1999 design. Therehas been some misunderstandineabout the bulb, because of ttre Oif-ferent specifications used in USliterature. I went into the manerin some detail in the 1999 articles.The correct specification is28Y,24mA.

The principle behind the use ofa tungsten bulb is that its resist-ance increases with the currentthat flows through it. If a bulb'sresistance is measured, the smallcurrent from the ohmmeter willincrease its resistance, Whenemployed correctly in this applica-tion the bulb does not come anv-where near incandescence. I decid-ed to use the lamp then because ofits elegant simplicity and extreme-Iy low distortion figure (0.0025Vo).With the new design I wanted toreduce the operating voltage of the

l . )ecemhcr , )OO4

TFI F\ / lq l r . ) \ l

TesFsignol porumeteruAmplitude: To test for ESR or foractual capacitance there are no con-.straints on how low the test sisnalcan be as far as the capaciior iiconcerned. Taking into accountpower consumption, and the prob-lems associated with lowJevel sig-nals of the order of microvolts,noise considerations etc., a level ofabout 5mV peak+o-peak seems tobe a good compromise.

Frequency: The time periodshould be short enough to zero outthe capacitive reactance. 100kHz isa popular choice. At about this fre-quency the ESR and impedanceconverge with the type of capaci-tors in which we are interested.Apart from that consideration,lOOkHz is about the top frequencyat which a predictable gain of up toten times can be obtained withreadily-available, low-cost opera-tional-amplifier ICs.

Remember that if the rate-of-change of the voltage is fast enough,

7 B

whole mcter so tha t a s ing le PP3battery could be used as the porvcrsource wh i le s t i l l ach iev ing verylorv po"vcr consumption. Theproblem '"vi th the use of a bulb istha t in these cond i t ions the bu lb 'stemperature is not suff iciently faraway from the ambient tempera-ture to ensufe good stabi l i ty. sothis t ime I decided to use thebrute-force method of diode stabi l-i sa t ion (D l , D2) .

The idea here is that when theoutput from the osci l lator r isesabove the conduction point of thediodes the negative feedbackincreases, the output sett l ing at anampli tude which depends on thecharacterist ics of the diodes. Inthis case the net result is asinewave source signal across R5

, with an ampli tude of about 6mVL peak-to-p"uk. Diod. stabi l isat ion

introduces greater distort ion, butthis is not important here. Tomaintain osci l lat ion, the value ofR3 must be over trvice that of R4.A preset resistor, adjusted to justsustain osci l lat ion ( lorvest distor-t ion), is usually used in the R3posit ion. I decided to use a f ixedvalue that 's a ferv ohms on thehigh side, to ensure rel iable osci l-la t ion regard less o f d is to r t ion .

Interfoce with rhe copocitorbeing tesfedThe interface with the capacitorbeing tested is the crucial part ofthe meter. It took me a long time toget this r ight. See Fig. -5. The wave-form across R5 (the source resistor)is not an ideal, constant-voltagesource, because of the need to

'--, .nclude the sample resistor (R6),rvhose value must be comparable tothe ESR values in rvhich we areinterested. The ESR of the capaci-tor being tested forms part of apotential divider rvi th R6. Thus i f ,for exanrple. a good 1,0007rFcapac i to r rv i th an ESR o f about0 ' lO is connected io r tes t , R6 isc { ' fbc t i vc ly in para l l c l rv i th R-5 . Th isIncans tha t the supp ly -s igna l sourcei \ l cs \ h r ' t ' ausr ' . rv l rc r r t r c i lp : rc i lo I i sbe ing tcs tcc l . t l l c cons t i rn t -cur ren tsourcc to l l5 i s sh l r rc t l rv i t l r thct :S I l u r rd [ {6 in p l ra l l c l .

' [ ' hc vo l taqc r i ,u r ' c l i r rn r i l cvc l

o p c t l a c r o s s 1 1 6 l t s : r r c s u l t o l l h e, l r i l , ' i l 1 l l t t , ' t t . - l r l l r e , l r P i r r ' i t r r t 'l r c i r r g { c s t c t l i s l r n r l l l i t i c c l u n t l t h c nr l c t c c t c c i b y t l r c r c . l o l t l r c c i r c u i t .l l t h c t r S I { o l t l r e c r r p a c i t o r b c i r r gt e s t e ( l i s c q u a l t o t h c v a l u c t r l l { 6 .l u r l l o l t h c s t r p p l y w a v c l i r r r n r v i l l

b c passc t l o r r . - l ' l r c

supp l y r vavc .

I r l r r n i s r r o t i r r t l cpcn r l c r r t o l l l r eI oac l l t o r vcvc r - . I l t h c I :S I t i s l c ss

than the value of R6, as rn rnostcases i t is, the lvavefonn voltageacross R6 increases.

As the ESR r ises above thevalue of R6, the latter becomesless effect ive. The result of al l thisis a non-l inear scale, expanded atthe lolver range and somewhatlogged ou t . Th is i s idea l fo r thepresent app l i ca t ion , because Insome cases i t i s impor tan t to bcab le to d is t ingu ish benveen a vcrylow va lue , c lose to zero , and oncof about 0 -5Q.

I have achieved low current con-sumpt ion and c i rcu i t s imp l ic i t y bytusing the feedback current andcomprornising somervhat on theideal. constant-voltage source.

Nexf monthIn the conc lud ing ins ta lment nex tr r ro r r th I ' l l comple tc the c i rcu i tdescript ion, deal rvi th sonre practi-cal points and protection ntethocls,provide a detai led contporre nts l istand a str ipboard layor-rt . I

Internol viewof the newESR meter.

fhe test leodconnections are ol

the bonom, withproteclion diodes

bet,tveen them.

6mV p-p under0 6V P-P anncr2nr oPen-c i rcu i t

( lCla p in 2) : : ; : : : ' , " condir ionsi Y " " " " 'i / Ii t iI l R 4 1 o o ( ) I E s n

Constant-voltagesource

(from oscfeedback loop)

Fig.5: The method of test capaci tor inter face used in th is nreter . The source vol tage is not a t rue

constant voltage because of the need for R6, whose value nrust be cotrtparable to the ESR of the

capacitors in which we are interested.Current economy and circuit simplicity are achieved by using feedback current front the

oscillator circuit as the source fed to the capacitor under test. The outltttt is inversely 'logged'

out in relation to ESR.

l l l l \ ' l \ l r r ' ,

This second, conclWittcox's lotest ESR meter, with o suggested stripboqrd loyout ond

component specificotions

I n Part I last month I dealt with

I gSn and its measurement, andI described the significant fea-

tures of my latest ESR meterdesign. I'll start this month with adescription of the basic circuitryused.

Circuir descriptionFig. 4 (see page 78 last month)shows the basic ESR meter circuit.The important features of the oscil-lator and test interface sectionswere covered last month. There isno need for a regulated Power suP-ply. Just to remind you, the amPli-tude of the HF output waveformobtained from the oscillator (ICla)is set by the characteristics, whichare virtually temperature-stable, ofthe two diodes Dl and D2 in thenegative-feedback Path betweenpins I and 2. The amplification anddetection levels provided bY theother stages are set bY the ratio oflhe source lnd feedback resistorvalues used. This is standard opera-tional-amplif icr practice. The wayin which opcrational amPlif ierswork rvas cove red in some detail inmy articles in the March/APril

1999 issues, so only a briefdescrip-tion is given here.

The frequency of the Wien-bridge oscillator is approximatelyequal to ll(zfiRq when resistorsRl and R2 and capacitors Cl andC2have equal values. At this fre-quency there is no Phase shiftacross the bridge and thus maxi-mum positive feedback. At reso-nance, the upPer section of the net-work (Rl, Cl) has twice the imPed-ance of the lower section (R2, C2),so there's a transmission loss of1/3. To sustain oscillation, the over-all gain (ALC) must be greater thanunitv. The transmission lossthrough the network is offset by thegain determined bY the ratio I +(R3/R4). In this circuit the ALCwould be more than the three timesrequired were it not for diodes Dl

. and D2, which override the effectof R3 as mentioned above.

The following two stflges ofamplification (IC2a and IClb) arestraightforward, rvith no need forconection circuitry. Capacitor C3 inthe feed to the detector stage (IC2b)is included to remove anY DC com-oonent and also to rctlucc sensitivity

to low frequencies (mains hum orwhatever). Rl2 sets the gain in thedetector stage. Because ofthe highintrinsic gain of an operationalamplifier, the forward voltage droPacross detector diodes D3 and D4 isovercome and detection at even mVlevel is not a problem.

SpliFroil genetator ondbuzzerThat is all there is to the basic cir-cuit. But the operational amplihersrequire positive and negative sup-plies. This requirement is providedby IC3b, see Fig. 6, which is config-ured as a voltage-follower. There is100 per cent negative feedback (pins7-6), so the output voltage must set-tle at half the supply voltage, set bythe equal ratio of Rl4 and Rl5. Iwas pleased to find that the circuitremains stable with outputs as lorvas +3'lV. This lower supply voltagerange is quite consistent.

IC3a is configured as a voltagecomparator. When the ESR rcadingis less than 0'5Q, the output fromthe detector goes higher than theforrvard voltage drop across D8.The output at pin I of IC3 thcrcfore

1 5 8 J r rua ry 2005 T i l EV IS ION

:oes high, activating the buzzer..-fhe ESR level at which the buzzer

operates is set by the overall gain.This point can'easily be changed byaltering the value of Rl2. If itsvalue is increased, the overall gainrises and the buzzer will operate ata higher ESR level.

The value of l00Q for Rl8 ischosen to limit the current from anexternal 12V source to a recharge-able battery to the trickle level. Theoverall consumption is so low that,if you are one of those who remem-ber to switch off battery-poweredequipment when it's not in use, anordinary battery is OK and will lastfor quite a long time.

Although the basic meter circuitis happy with a supply between 6-30Y the buzzer and LED won't be.The value of 2'7kf,J for R19, whichis in series with the power-on indi-

. rtor LED, gives good brightness\ver a supply range of 6-9V with

only a feg mA drawn.

Prqcficol pointsAs the meter operates at 100kHz,any inductance in a circuit beingchecked will produce a high-impedance reading. If an EW coilproduces a reading, it has shortedturns. Another use of the meter iswhere the line output transistorappears to be short-circuit. A quickcheck with the meter will isolate it- if the short is elsewhere, in mostcases the impedance of the line out-put transformer will be in the wayand will result in a high reading. Ifthere's a low reading, the transistoris most often the culprit.

ESR meter users have come upwith new applications. The non-polarised, high-voltage capacitorsused in the line output stage (tuning

FromRl2lC4

f I Prlntcuts(24)

etc.) can be tested. I've not takenthe trouble to design a new meterscale for this application, as I haveno practical experience of suchchecks. It seems to me that if sucha capacitor gives any reading at allother than short-circuit it will beOK. This type of capacitor does notchange value.

I would like to think that thepresent article just about sumseverything up with resPect to the

l c l , lczpin I

l c l , l c2pln 4

+ve supply

Meter +ve'

Top - Fig. 6: Thesplit-rail generator

and buzzer com-parator circuits.

Centre- Fig.7:Suggested layout

on stripboard,

Left - Layout ofthe prototype

meter on strip-board

Bottom - Fig.8:The ESB meter

scale, shown fullsize l58mml.

t 5 t ! 2

o

ESR at 100kHzCr

'Connecl to track slde of board

TELEVISION January 2005 1 5 9

Parts listItem

R l , 2R3R4R5R6R 7 , 9 , 1 1 , 1 6 , 1 7R8, 10R 1 2R 1 3R 1 4 , 1 5R 1 8R 1 9

...\\

Valueftype

3ka220A100a

1 A2.7A10kc2

100ka*68kQ

3.9kA56kO

*100Q*2.7kf2

CPC order code

REMFR4 fo l lowed by the value

Al l0 .5W, 1% meta l f i lmVR1c't,2c3,4c5, 6 ,7

1OkQ cermet preset

470pF low-loss high-stabil ity*r0. 1 pF cer-amic multi layer221tF, 16V

RE01881

cA02068cAo2098cAo1613

sc1N4't48sc1 N4004sc1N4002scTL802.sc00023

P M 1 1 1 1 9sw-z201lz1S00654EN55030.1N00772.PC00046PC000668T02187PWo0037.

D 1 , 2 , 3 , 4 , 9 , 9D 5 , 6D7l c 1 , 2 , 3LED

1N41481 N40041 N4002TLO82CN3mm Superbr ight

M 1S 1

100pA moving-coi lMin iature toggle swi tch

Buzzer 5V DCCase ABS boxTest leads 2mm plug to probesVeroboardSpot face cutter for VeroboardPP3 battery clip leadHigh-current protection choke

See textSee text

See text

quick in-circuit location of faultyelectrolytic capacitors. If anyonecontemplates the design of a PCBfor the project, it is important thatseparate operational-amplifi er chipsare used for the oscillator and thefirst amplifier stage - to avoidinterference between the oscillatorand the sensitive first amplifierstage. Fig. 7 shows a stripboardlayout for the meter's circuitry. Fig.8 shows the meter scale.

Don't be tempted to use plugsand sockets for the test leads - youwould in time get problems in thelow-ohms range. Soldered connec-tions should bc used throughout. Iuse 2mm test leads with the plugscut off. Make surc that you fi le theprobes to give sharp points. Thecoating that's on then.r has a signifi-cant resistance.

The case spccificd in the partslist is a bit on the dccp side. To

achieve a slimmer appearance, Ibought another case, type EN55029(too slim), and combined thehalves. This might sound extrava-gant, but you still end up with twocases and they cost only about f2.

Protection mefhodsIn a letter in the August issue thisyear Jim Littler suggested wiringan inductor across the test lead ter-minals to protect the meter shouldit be connected to a charged capaci-tor. I can see no problem with this,and followed up with a letter in theScptember issue. If a value some-rvhat lower than the l5(i7rH rccom-rnended there is used, producing areading of say 30Q, this rcadingrvil l be present each tirre the nteteris srv i tched on and you rv i l l knowthlt it is working correctly. It wil lnot al'fect the use of thc ntcter. Welr re only in terested in vulucs that

are much lower.If this method of protection is

used with a digital meter, the dis-play will settle at a fixed reading.This will show that all is well withthe meter and will also eliminatesuperfluous readings. In the case ofa moving-coil display, it will dou-ble-up as a power-on indicator.

The use of a circuit protector inseries with the test leads has beensuggested. The problem is that itwould tend to blow too easily andrequire frequent replacement.

To reitcrate, diode protection(D5, D6) should always be incluclcd.

Any contments about high-cLrr-rent choke protection, which seenrsto be a r"rniqr.re idea, and on ESRmeasurcntcr)t in general would bewelconrc.

You can reach me by email atalancsr(lr hotrnai l.com

I th ink that covers everyth inq. t

r 6 0 J ,u rua ry 2005 TELEVIS ION

Table 1: T.yRical ESR values at,1Q0(Hz.r. ,",... t

Capacilaicb vatie Voltage rating tmpedih'ce*:

l t tF2-21tF4.71tF ' '

.10ttF ,221tF471tF471tF100pF100pF2201tF22O1tF47OpY

. a7O1tF1,000pF1,000pF2,2001tF2,2O0pF : .

50v50v50v63V' 5 0 v

25V'�50v16V35V.16V35V16V35V16V25V25V',35V

4rl2.8cr2.4A1.gfi1.3cl1.3fi0.7ct0.5c)0.2scr0.25cl0.1 14cl0.1 14ct0.065cl0.06sfi0.041rt0.036rl0.034cl

*When new - low-ESR tYPe.

Anologue ponel meters It - . #

Io n o e r ss t * i l r e r i c .

FAnuetc z z 9 8 2 7 8

?*,rrlrrrrlrrrrlr tult t r t lr, or ' r r ' 4

^ 1 t , t t l ' l ' l ' l r r r 7 , ,

a

a

a

a

a

Attrocfive md modern dextignClxfce of printed scdes4 popdqr sizes"Eosy-frf'scdes

Chss eO

A modern and attractive range of quality dc moMng coil meters availble atvery competitlve prices.

There are 4 popular sizes which can be either surface or flush mountedbehind the panel through a rectangular cutout. The meters utilise a wellproven and robust, 90 degree deflection, pivot and jewel moving coilmovemenl.

Meters are supplied in sensitivities of 100 microamps, 100-0-100microamps (centre zero), 1 milliamps or as 4-20 milliamps (supressedzero) for signal process indication.

Pre-printed scales are available seperately graduated and numbered 0-10,0-30, 0-100 and 100-0-100.

{ _ f i +

Fixing studs on all meters (Dimension G)are M3 x 12mm

Dimensions in mm

M o d e l A B c D E F H I J W S c a | eLength

CV-15X 57 48 43 24 40 35 1.5 14.5 28.5 50x23.5 40

CV-16X 68 58 45 31.5 48 48 1.5 14.5 28.5 61.5x28 50

cv-18x 80 67 50 38.5 64 48 5 14.5 28.5 72 x35 63

CV-20X 100 83 50 51.5 80 64 1.5 14.5 28.5 92x45 78.5

AVA1ASIE FRort NATKta,lA! OlglRlBUtORs OR DIRECT FaOl ANO*S SUEJECT rO A mOQ OF 25 PC$ ANALOGUE\CVl Revision 4 03/06/04

Onders electronics plc Bayham Place, London NWl OEU UKTel: +44 (0)20 7380 8167 Fax: +44 (0)207874 1908 wwwmeters.anders.co.uk

Crder Code Size Range Rec. Price €:

1+ 5+ 10+ 25+ 50+

METERS

3V15X1MA CV-15X 0-1mA

:V15X100UA CV-15X 0-100F4

:Vl5XGZ100UA CV-15X 100-0-100!A

3Vi5XSZ20MA CV-15X 1004-1004

3V'OXIMA CV-16X 0-1mA

3Vi6X100UA CV-16X 0-100ttA

3V16XCZ100UA CV-16X 1004-100A4

M6XSZ20MA CV-16X 4-2OmA

;V18X1MA CV-18X 0-1mA

3V18X100UA CV-18X 0-100rA

;Vl8XCZ100UA CV-18X 100-0-100FA

3V18XSZ20MA CV-18X 4-20mA

3V20X1MA CV-20X 0-1mA

CV20X100UA CV-20X 0-100!A

3V20XCZ100UA CV-20X 1004-100pA

3V20XSZ20MA CV-20X 4-20mA

METER SCALES

S/CVISXB CV-15X Blank

s/cv15x10 cv-15x 0-10s/cv15x30 cv-15x 0-30s/cvr5x100 cv- 5X 0-100

s/cv15xczt00 cv 5X 100-0-100

9/CV16XB CV 6X

s/cv16x10 cv 6X Blank

s/cv16x30 CV 6X 0-30

s/cv16x100 cv 6X 0-100

s/cvl6xczl00 cv 6X 100-0-100

s/cv18xB cv- 8X Blank

s/cvl8x10 CV 8X 0-10

s/cv18x30 cv- 8X 0-30

s/cv18x100 cv- 8X 0-100

s/cvl8xcz100 cv 8X 100-0-100

S/CV20XB CV-20X Blank

s/cv20x10 cv-20x 0{0s/cv20x30 cv-20x 0-30

s/cv20x100 cv-20x 0-100

s/cv20xcz100 cv-20x 100-0-100

SD80/0-100UA I MULTICOMP I Panel Meters I Electrical I Farnell

2OO7 /08/08

SDSOIO- IOOUA - MULTTCOMP - METER, 81X81MM O- IOOUA

Page I of I

4S 60

Manufacturer: MULTICOM P

Order Code: 4433932

Manufacturer Part No: SD80/0-

lOOUA

Rol15 :

Descript ion

,Ava i l ab i l i t y

Availabil ity: 1

Price For: 1

Minimum Order Quantity: 1

Order Mul t ip le : 1

U*it ?rice: 99"5$

Image is for illustrative purposesonly.Please refer to product description

Technical Specif icat ions

r*l Product Rang;e

Catalogue Page:

5 8 s / U K 2

ffirech::ical Oata $heet {i.95 kBi

Product Attributes

Weight (kg) : 0.12Approximate weight in primary packaging

Tariff No.: 90303399

Country of Origin: TW Taiwan

Country in which last significant manufacturing process was carried out

aaa

aaa

M E T E R , 8 1 X B 1 M M O - l O O U ADepth , ex te rna l : 36 .5mmDiameter , pane l cu t -o u t : 6 3 . 5 m mLength / He igh t ,ex te rna l :81mmMeter MovementF S D : 1 0 0 p APi tch :34mmResista nce, coi l : 1 500RWidth , ex te rna l :81mm

_ _Qtv.._ . -^:":O r i r e

:i L - 4

. t - n

. L O - 2 4

: 2 5 - 4 9

List Price

€9 .5B

f 9 .38

89. t9

€8 .96

E7 .37

f i0 ie rL*qc

p: i , /acy $l :atcme nt I Si te r f lep j Te rr rs cf Prrchase I Te rmS l f Acce r :

Cc;:iTrlght ii-; 2307 l:rer*:er Ferne il UK Llmite d ALL R:Gli15Copy!-rght in t i l r whoie arrd every part of l | i is !1 i r t is i le *e l rn;s ic Pre r : ier ?anet l i )4. Larni f . : l

i i cen$ed , t r ans fe r i - ed , cop ie i o . r ep : ' odu lec i i n who le oT i n 3a ! - l i n a l y manne r o r : ' : t t : l ' : ; i r ' ;acccrcance y, , l th the terms of i re O'^ner 's agreer:ef l | , wi tno, , r t ln* pr t i ) t wrt '

Fainel l is a d:v is ion of pren' r ;er f* rnel l UK Limited. Regist* : ' *d in fnciard and Waies f i . : C{. }8(Lecd: , i . .g l : 2QQ.

Prcd ucer R cg lstration l'l u *r be r : \'! f.f. I AK t), 4117.

http://uk.farnell.com/jsplElectricaliPanel+MeteTsAvIULTICOMP/SD80/0-100UA/disp... 08108/2007