RESEARCH DEPARTJlJ.lENT

S01tlE OBJECTIVE MEASURFJY.!EJ.\fTS OF MOTOR CAR IGNITION INTERFERENCE.

Serial No. 1948/40

Research Department

Work carried out by: D. 11aurice J. G. Spencer

REPORT NO. G.043

Serial No.1948/40 January 1949

Figs. Nos. G.043.1 to G.043.4; G.036.38; Please refer to G.036 and A.023

SOIl/LE OBJECTIVE IIJ1EASUREMENTS OF MOTOR CA..f:t IGNITION INTERFERENCE

The radio interference caused by the ignition systems of some 350:notor cars has been measured. It has been recorded in terms of interfering field strength ilvhich permits the immediate calculation of output signal to noise ratio whatever the transmission system involved, provi-ded that the distance from the receiving aerial to the source of interference and the wanted field strength are known.

1. bt:roduction

During the past few years a considerable amount of theoretical and practical work has been done in connection with radio interference caused by motor Car igni tion systens. R.H.Gcorc;e I D.B.SrJi th, W.E.Bradley in the D.S.A., F.L.E.I.I.StlJ~'"1:;ers in. Holland, S.F.Pearce and G.L.Hamburger in ED.[~land, is only a selected few of the -names of contributurs to this work.

O:")jocti ve measurements of ignition interference have been made from a few tens of kilocycles per second up_ to several hundreds of megacyclos per second. The measurement results have been given in variom; -forms; a pOl)ular E;),;lish and Continental method being the statement of tha:t value of wanted carrier which amplitude modulated to 100';S would give a receiver output signal to noise ratio of 40 db as read on a meter consistine of a diode detector having alms charge time constant and a 160 ms discharge time constant. T::is meter VJQS supposed to have an "input energy against moter reading" c£laracteristic similar to that of a listener's ear.

This method of impulsive noise measurement has at least two disadvantages. Tnc first is that its mc.;ter readings bear little r01at ion to the degree of interference which a given ignition system might cause to a television picture, and -the second disadvantage is that 1/JO do not novv cons idor the characteristics of the above-mentioned detoctor to be as clos8 to those of the ear as they might be (830 RGsearch Roport G.040). Thoro is a third, indiroct-, disadvantago in that the wanted carrier required to produce the 40 c~b output signal to noise ratio is invariably much smaller than the actual peak interferenco pulse volts 1iJhich imping0 themselves upon

- 1 -

- 2 -

the receiving aerial. Thus engineers dealing with interference problems are often unaware of the relatively very high peak voltages ensuing:

This report gives the results of some measurements of the ignition interference over the 90 Nw/ s Jffiil band, of more than 350 motor cars. The results are given in three equivalent forms and are such as enable the .engineer to calculate immedietely for a given carrier the actual output signal to noise ratio in enygiven transmission system provided he under-stands how his particuler transmisSIon system works.

·2. Specification of Impulsive Interference.

Over 0 norrow bend of frequencies, fOI instance 90 to 100 Mc/s, the spectrum of the interference from any individual motor cer is substant-ially unifor.m. The repetition frequency depends on the speed and number of cylinders but the peek VAlue does not. It seems reasonable, therefore, to specify the interference either os a spectrum envelope (envelope in order to avoid the chenge of spacing between spectral lines due to change in pulse repetition frequency, PRF) or as its equi velent impulse, defined by its orea in an .1exci tBtion - time,l plane; for example an impulse of so many.millivolts/metre x milliseconds.

It is shewn in the Appendix that An impulse of area U, and of duration short compared with the reciprocal of the highest frequency of the RF band used for observing it, can be specified either by its area or value, or by the rms voltage which it would produce in a given energy' or utilizatton bandwidth, 2LJr, assuming a given repetition frequency, fr. It can also be defined by the peak voltage which it would produce in a utilization bandwidth, 2 fIt'. This peak voltage figure is theoretically dependent upon the phase characteristic of the band 2&, but in practice, radio receiver circuits are such that'their phase characteristics do not differ sufficiently to preclude the use of this peak voltage definition.

If U is given in (peak mV/m)x(ms), !Jf in kc/s £md fr in kp/s, the rms millivolts, V~ms, are given by

Vrms == 2U ~frtsf

and the peak millivolts by

Vpeak :::: 3.62 u6f •• b ••••• O.O' ••• O •••• ·" ................ .

These two formulae ere immediately applicable to practical cases of which three examples will now be given.

1

2

- 3.-

3. Uses of the Spocification of Impulsive __ Interference.

3.1. A 90 Mcls A.M. Sound Receiver.

Suppose that a 90 MC/s receiving dipole were situated at such ~ distance from a road (60 feet away and 25 feet high) that the average motor car gave an impulse at the aerial of U = 0.004 (peak mV/m)x(ms). The way in which this may be measured will be dascribed later. Now ~uppose the motor car to have four cylinders and t" 'De travelling at about 30 mph. The PRF of the interfering impulses w ill be about 50 pIs ;:: 0.05 kp/s.

Whatever the A..TVl receiver overall bandwidth, provided it is at least:: 5 kc/s, the effective utilization band will be that of the listener's ear, namely about ~ 5 kc/s IH 2!:Jf == 10 kC/so Now suppose themceiver to be tuned to a wanted carrier of 1 rms mV {m modulated to a depth of 40%. The audio signal to noise retio will be, simply

lfanted carrier modulation level (mV!l!!;) Vrms (mV/m)

or, applying equation 1

0.4 x 1 20 10€lO ___ _

2 x 0.00~0.05 x 5 = 40 db.

3.2. A 90 MC/S ~ 30und Receiver.

Suppose the same conditions as above apply in this case. Assume the FM system to employ a poak deviation of ~ 75 kcls and a 50 ~s pre and de-emphasis. First we shall see whether t his particular signal to noise ratio is better or worse than that obtaining at the FM/AM improvement threshold. The RF input carrier to impulse ratio at improvement threshold is ,from Research Report G. 036 , equation, 6' and A.ppendix 7 equation 4.

..lL U

= ,; 1 + n 2

eO'/n Fl(n) 4n !::f 3

where n is the coupling parameter of t he receiver IF bandpass circuits

8' = arctan n Fl (nl is a function of n shawn in G.036.7 TJ is the carrier amjJlitude at improvement threshold.

... 4 .....

For virtually all receiver IF circuits 1

3.3. A 45 Mc/s AM Television Rece.iver.

According to Research Report A023 Table IH, the noise field due to ignition interference is four times greater at 45 Mc/s tharu at 90 Mc/s so that we must take our new value of Uas U = 0.004x4 :::: .016 (peak mY/m) x (ms). If we assume the t el~vision receiver to have a utilizationbandwid'lih of 2M' :::: 5000 kc/s,the output peak signal to noise ratio for 100% modulation (peak white) will be-;fOr a 1 rms mY/m wanted carrier,

wanted carrier modulation level (mY/m) Vpeak (mV/m)

or applying equation 2

20 10 1 x ~-2-~ 3.62 x 0.016 x 2500

= - 40 db.

3.4. Summary of Examples.

To calculate a signal to noise ratio obtaining at a given wanted signal field strength, calculate the .AM cases from equations 1 or 2 depending upon the type of AM transmission considered. If a ~ 75kc/s 50 ~s FM transmission is being consldered, calculate the peak carrier strength required to obtain the improvement threshold from equation 4. Divide the given wanted poak carrier by the value of ~ thus obtained and call this ratio Po/Po From Figure 46 (G~036.38 attached) obtain the FM/AM improvement at that value of Po/p and thus obtain the FM signal to noise ratio from the previously calculated AM signal to noise ratio. G.036.38 Fig.46 does no't apply to peAk signal to noise ratios~ only rms aural ratios. The application of FM to television vision has not yet been seriously cpnsidered an,d has therefore not been investigated. If FM systems other than ~ 75 kC/s 50 ~s ones are to be considered, a study of the addendum to G.036 will shew how to deal with them.

4. )-'uthod of Measurement of Motor Car Ignition Interference.

The FM/1?J1i improvement threshold, equation 3, is characterised aurally in a FM receiver by a spectral and time functton change in the receiver response (output). Above this threshold, that is, for carriers greater than that . required to obtain it, the impulsive interference sounds like.a succession of .clicks and consists of bipo~ar pulses having a frequency spectrum

- 6 -

proportional to frequency. Below the improvement threshold the inter-ference consists of a mixture of clicks and an ever increasing pro ...

. portion of pops which are unipolar pulses having uniform frequency spectra. As the carrier level is decreased from the value at improve-ment threshold, the percentage of pops to total of pops and clicks increases, rapidly at first and then becomes asymptotic to 50 when the carrier is very small, Po/p ----> O. The aural determination of improvement threshold is thus quite easy and can be used as a peak IF output voltage measurement instead of the more clumsy methods employing valve voltmeters or oscilloscopes.

The measurement of car ignition interference used was thus as shewn in Fig.l. As a motor car came past the aerial the standard signal generator attenuator was rapidly adjusted to such a value that the pops in the interference were just eliminated. This carrier level was referredw field strength at the dipole aerial, and recorded.

Fig. 2 shews a statistical record of the results. Fig.3 shews the same results in a more useful form. Fig.4 shews the actual inter-ference in general terms. Fig.4 was obtained from Fig.3 by means of equations 1 and 2 after U was determined from equation 4. The receiver used for the measurements was a B4B.C. type HR/8 FM/.AM receiver having n = ~~ and 6f = 80 kC/so

The values of ~ used in equation 4 were the values given as abscissae in Fig.3 after multiplication by /~. The measurements were done at the junction of the Brighton and Dorking roads near Kingswood in Surrey.

In contradistinction to Research Report A023 Tables 11 and V, there was no noticeable difference in ignition interference whether measured on a vertical or a horizontal dipole, indicating more or less circular polarization. This disagreement between two sets of measurements, both undertaken on a fairly large number of vehicles, might be worth investigating.

DM/MP (H. L. Kirke)

/

- 7 -

APPENDIX 1.

Relation of aE Impulse to its Spectrum.

Consider a rectangular impulse of duration short compared with that of the upper frequency limit of the band in which the pulse is to be examined. Let the pulse be of height A mV/m and of duration T ms. Such a pu.lse will hav-e a spectrum

T -jwt ~(jw) = ~Ae dt •••••••••••••• po •••••••••••••• 5

If U is the area AT· of the pulse, we may write

A _°wT ~(jw) 7 jw (l-e J ) •••••• flC .............. oo ••• e> ••• 6

If wT «1 we find that

~(jw) = U "b ••••••••••••••• OOO .......... · •• 7

If the pulse, U, be repeated f r times a .millisecond, tJo.e mean square value of it, measured in a frequency band 2!J f ko/s is

(lO •• !!1 ••••• 0 ............ " ••••••••• 8

V nus ;::::: 2 U ~ f r" I1f ... e ••••••• 0 •• " ••• 0 0 • III ••• 0 •• ., • • •• 1

Now tlte ratio of peak to rms values of the output response, for a full bandwidth of 2!Jf and a PRF of fr, is given in Research Report G.036 equation 8, as

-st/n Crest Factor =' 2e /11Tti:

-,/ f--r-&~F""'l-\""") n) ... 0 00 Cl ........... . 9

The factor before ~6fJfr in equation 9 remains within three quarters of 8 decibel of 1.91 for 1

ISSUE

e · o' ~e I

, I

90 Mc/s DIPOLE

q FM ~ RECEIVER

STANDARD ~-~----'

SIGNAL \ 4- ..n. GENERATOR

FIG, f. METHOD OF ME~SUR~MENT OF Cf:..R IGNITION lNTERFERENCE.

RESEARCH OEPT

BBC SOME OJECTIVe: MEASUREMENTS OR'N M.L~ GR,EOPOR4T 3 ,1 1------1 OF' MOTOR CAR INTERFERENCE, cfi-D W-

I OS/1/0C AP'O +SHE.ETS. ~-

ISSUE I-

19 -\0-48

; r i:~

r~ E~

~i IJ' E1

, r{ 11

r'~ it hf+fhHH It

I~

ft If ~F if If f~

h~li!:t lim , J~I'l if ;T ~ .tm!

if ~ f~">lLlJljC:"" it ff :1tr+fn

~Ii $f, ~

ti G:~ :f: ~ 1 tt gfl 1-: 2. 2 .. 5 3

tt ;+ H

It!-

I,"" I .....

.~ I~t It!

p. ~mtt;~ +1' :1' ~h r;t.:H~ ~.at: I:l+ ~. m· ctt'Ft.

ST~E.NG'TH IN 1':",S. mV/,.,... Ri:QI? TO OBTAIN IMPROVE,w\E.NT "HRf.S~OLD,

FIG. 2. ~lEL.D STRENGT~S REQUIRED 10 MA.INTA\N t=M/AM lMPROVEMENI THRE.SHOLD FOR 300 CARS.

BBC SOME OBJ'ECT1VE MEASUREMENTSI---+-rt'H ~--:--I OF MOTOR CAR INTERFE.RENCE. t----~

RE?ORT

G.043·2. 4 SI-IE.ETS

19'10-4-8

90

80

,..... ~ o '70 n-O z

S GO l-V

~ o tt 0. 50

20

10

o o Z 3 ~ 5 G

FIELD STRE.NQTH IN ""'.m.5. mV/m .

t='IG.3. ~IELD STRENGTHS REQU'RED TO PROTECT A

TRANSMISSION ' r=ROM GIVE.N PE~CENTp..GES OF C~.

BBC SOME OBJECTIVE .MEASUREMENTS~~ ~---IOF MOTOR CAR lNTERF"ERENCE G.043'3.

4SHE.E.TS .

BBC ~~OF

'0

\ ~ti.:

I

+ j

It JX; itl

it

'ctml' • iHit

fItJ tte1

...

t

" ;

ri tt +

• ! im . rl:i:1

, " fro J

.f + tW . t~

. .. t$l tfi . . ftJ ,

t

HL. it:' in " Itl:tl .4F £,1'1. + EH ::Hr l .. ''';:; +-,... :t;:h~·,,~ ...... tt t m It~+:tU It

ffi~ g ftl , mllF

m ft

; . : ~i

l.j:

. . rP" 11+ W!' i.!;j 1 11 .. I tt t .f u'l±l:± ' ~ !±EH! .

~ ID lS~ , ·tgIrlftH

tli •

!ID ~t! : ~.~ . 1+1

rH !++ t1@ t ~ mt . ..

-+ . . . .-, cj:j

~t

t . ,

• t

..

o $.tt;p.!t ctH· '111- :Wf o 10 20 210 40 50 00 70 PERCENTAGE. OF C,tl..~S E'XCEEDtNG INTE.RFERENCE VALUES

F\G 4.

, tit, .

+ +-'-' ~

, rr itt~ . r$1 1:1 H .. ,

~ :.ttti fH

80 90 SHOWN AS OI

OJ CD ()

-~ fT1 ;0 ,., fTl

~ Z n rn

party on any (o r m w' "0 t mission of t he CO'por~ '01'

0 ~ fT1 r ~ < fTI

~ C JJ J> .- -to Cl)

il 1-P r ~ 0

5 iD rn -20

;l]

~ 0 (fI

Z a. rr

-30

. . •

1

1.9 _

.!i_ Vl I Ul ~ c: -.J m

·001 ·0\ PO/p 0'1 I

p= NOISE. "TO C~RRIER R.b..TIO AT IF O\J\PUT. Po: VALUE O~ P ~"T ~M/p,.M \N\PROYEMENi THRESHOLD