Special Topic1 1 Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices 2 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices Recently, electromagnetic compatibility (EMC) issues have taken on growing importance, especially selfgenerated noise which represent interferes with reception in wireless handsets. Until recently, there was no adequate method of measurement for this issue and subsequently effective suppression methods could not be developed. In order to solve this noise measurement issue, Taiyo Yuden has developed methods by which the adjacent electromagnetic field is measured through use of APD (Amplitude Probability Distribution). These methods enable us to measure the noise causing electromagnetic compatibility issues that until now had proven difficult to measure. In this article, we will present the findings of our research on noise evaluation and counter measures derived by APD measurements in the adjacent electromagnetic fields. Taiyo Yuden Navigator navigator3.indd 2 2009 09.8.3 2:46:12 PM 1 Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices 09.8.3 2:46:12 PM navigator3.indd 3 3 2009 Taiyo Yuden Navigator Special Topic1 Figure 1 Internal noise of cellular phones Frequency spectrum Time waveform -200 -60 -190 -70 -80 -170 Level [dBm] LCD module Level [dBm] -180 -160 -150 -140 -130 877.49 877.51 Frequency [MHz] 877.53 877.55 -200 -60 -190 -70 -180 -80 -170 -160 -150 -140 -130 0.15 0.2 0.25 -90 -110 -120 -130 -150 -160 877.47 877.49 877.51 Frequency [MHz] 877.53 877.55 -60 -190 -70 -180 -80 -170 Level [dBm] Level [dBm] Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices 4 0.1 -100 -200 -160 -150 -140 -130 0 0.05 0.1 0.15 0.2 0.25 0 0.05 0.1 0.15 0.2 0.25 Time [sec] -90 -100 -110 -120 -130 -120 -140 -110 -150 -160 -100 877.45 0.05 -140 -110 -100 877.45 0 Time [sec] -120 Camera module -130 -160 877.47 Level [dBm] Level [dBm] Processor -120 -150 -110 1 -110 -140 -120 -100 877.45 -90 -100 877.47 877.49 877.51 Frequency [MHz] 877.53 877.55 Time [sec] from even low levels of noise, that previously LCD module, processor IC and camera module had not been an issue. of a cellular phone by magnetic field probe. In wireless handsets and mobile devices, such Circuits of DC-DC converters, LCD modules, In the most commonly used frequencies, con- as cellular phones, electromagnetic compat- CPUs etc. are often victims of this issue. The spicuous spectrum peaks cannot be observed ibility issues continue to grow in significance. operating frequencies of these circuits usually and is similar to the heat noise value. That We are seeing more issue such as auto-jamming ranges from several MHz to tens of MHz, but means any obvious noise cannot be detected . which is caused by self-gererated noise, recep- the harmonic noise that may exist, is often up to However, by measuring the time fluctuation tion and sensitivity issues, among others, de- ten orders greater than the frequency, even if it of noise values in the same frequency range, grading device performance-largely due to self- is a faint one. as indicated on the right, we can observe clear generated noise from the device itself. Until now this has not been an issue, but if and fluctuations in the noise waveform. Due to layout considerations, wireless handsets when the frequencies overlap with those of the In this way, the inner noise of the device can be are forced to place multiple components [anten- circuits there is a strong possibility of EMC is- shown by hourly fluctuation, which cannot be nas, tuners] and digital circuits which generate sues occurring. adequately measured by a spectrum analyzer. noise unexpectedly in close proximity to one Noise characteristics from actual cellular But, by measuring the time fluctuation, noise of another, which often come into conflict. phones which could affect sensitivity are shown this kind can be clearly measured. This coupled with the greatly improving sensi- in Figure 1. One method of measuring and evaluating hour- tivity of wireless circuits, such as tuners, often The noise waveform on the left indicates data ly noise fluctuation is through APD Mapping. leads to electromagnetic compatibility issues outputs measured by a spectrum analyzer of the Noise causing EMC issues Taiyo Yuden Navigator navigator3.indd 4 2009 09.8.3 2:46:14 PM Figure 2 APD P1 : Hourly probability exceeding threshold R1 Pk : Hourly probability exceeding threshold Rk Rk : Threshold W( : Term in which the value of noise exceeds threshold n Rn) R0 R1 Rk:Voltage P1 Waveform of noise W (t) Probability W1(R1) W1(R2) W( n Rn) 1 R0 (Formula 1) which is obtained by Wi(Rk); the total amount of time of X(t) exceeding R(k) dividing by T0; whole amount of measuring time. R1 Rk Level of interference 2. APD as an indicator of interference in digital communication APD allows study of issues related to interfer- Noise measurement technology referred to as ence values of digital communication. “APD” is an effective method of analyzing the For example, a strong correlation can be shown time fluctuation of noise, which can cause elec- to exist between the APD noise in the band tromagnetic compatibility issues. frequency of a digital communication system APD correlates the value of interference to the like GSM and the influence of that noise on the digital transmission, like “One-Segment” a system.[*2,3] Japanese digital terrestrial broadcasting service APD is the data graphed from the relation be- That is why measurements by ADP are a reli- tween the Rk (horizontal axis) and variable of able indicator of interference for wireless hand- Since APD also provides hourly statistical val- provability P(rk) (Verticalaxis). sets using digital communications. ues, it can be used to effectively measure time This data indicate the probability the noise This is strongly supported by the reception fluctuation of noise, which is close to heat noise would emerge, how frequently and how strongly sensitivity data of "One-Segment TV" and the (Figure 2) by measuring the time fluctuation interference values measured by APD. of noise. In "One-Segment TV", an OFDM (Orthogonal 1. APD In order to more accurately measure and evalu- Frequency Division Multiplexing) modulation is APD or "Amplitude Probability Distribution", is ate EMC, Taiyo Yuden has developed its own adopted for the wireless communication system. a tool for statistical analysis. measuring device to better apply APD. Our experimental result shows a strong (positive) Now recognized by CISPR16-1-1, APD pro- In high-frequencies commonly used in wireless correlation between the deteriorating value in re- cesses are now increasingly being applied to communications, this instrument allows mea- ception sensitivity and noise measured by APD. accurately measure electromagnetic noise. surement of the time fluctuation of noise and The spectrum band of 473 to 767 MHz was ap- APD stands for the relation, in case that a calculates the APD at an arbitrary amplitude plided, as in digital TV broadcasting channels for mobile TV. in level. [*2] [*1] An APD overview follows below. threshold Rk is set to an amplitude waveform X(t), between Rk and a variable of provability P and time. [*1] 13-62. The spectrum band is so broad that the fre- Taiyo Yuden Navigator navigator3.indd 5 2009 Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices Time T0:Total measurement time APD Noise Measuring Technology APD Curve Pk 5 09.8.3 2:46:14 PM Special Topic1 1 quency from which the noise comes from is "One-segment TV" through hybrid electric cir- converter operates at 1MHz) by wide-range different for each channel. cuits, which can add high-frequency signals. antenna and adding it to the signal through the By measuring the relation between the value of In this situation, if the output of the signal hybrid-circuits. noise in each channel and the sensitivity of re- generator is turned down, the tuner could not The measurement systems was connected, ception, we observed a correlation between the provide a picture at a certain level. noise was measured and APD determined. sensitivity of reception in OFDM modulation We set the level of each channel to the tuner's From the APD method, a certain value of communication systems and the measurement sensitivity level of reception without noise. probability was derived (for example 10-2), and of noise by APD. The receptions deterioration due to noise was a correspoinding threshold was obtained. The measurement system is shown in Figure 3. provided by measuring the reception sensitiv- By comparing the thresholds which have the A signal generator produces digital TV broad- ity of the tuner to the generated noise received same value of provability, we can see the rela- cast signals, was connected to the tuner of from the source of the black box (DC-DC tion between the two in the quantity of noise. For this measurement, the thresholds of -55 -75 -60 34ch (599.142857MHz) -65 -85 -70 -90 -75 -95 -80 -100 -85 -105 -90 -110 -95 -115 -100 -120 -105 13 17 21 25 29 30 37 41 45 49 53 57 61 probability were 10 -2, 10 -3, 10 -4 and 10 -5 and graphed for each channel. Receive sensitivity [dBm] -70 -80 Noise Level [dBm] Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices 6 Figure 4 Measurement of interference by receiver sensitivity and APD Moreover, this graph made without adding the noise shows the noise of the measurement sysNoise from DUT (10-2~10-5) System noise on APD measurement (10-2~10-5) RS with noise RS without noise RS:Receive sensitivity tem (thermal noise). Figure 4 shows the comparisons of those graphs at each channel. This figure demonstrates that the reception of high channel whose noise level is evaluated by APD as high deteriorates significantly. In contrast, the re- CHANNEL [CH] ception at low channel is less deteriorated. Figure 3 Receiver sensitivity in "One-seg" and APD LEADER ISDB-T LG3802 Digital TV signal generator Tuner for One-seg TV broadcasting PC Hybrid circuit Shield box DUT Antenna APD measurement machine Taiyo Yuden Navigator navigator3.indd 6 2009 09.8.3 2:46:15 PM can be used as a tool to evaluate EMC from the taneously map the electric and magnetic noise by APD can be an index of influence which has device itself. of DC-DC converters that previously proved an impact on reception of one-segment tuner. To do this Taiyo Yuden has developed an APD too difficult due to large time fluctuation in the mapping measurement system. electric and magnetic fields (as shown in Fig- 3. Mapping Measurement of adjacent By use of this system we can now accurately ure 6). electromagnetic field by use of APD and simultaneously determine the APDs in Differences in mapping results between the EMC issues are caused by adjacent electro- both electric and magnetic fields. This is ac- spectrum analyzer and APD approaches may magnetical field interference as the source of complished by use of a special sensor, designed occur. the noise. by Taiyo Yuden, featuring two outputs, thereby A distribution of a digital IC magnetic field, as For this reason the mapping of the electromag- enabling detection of both electric and magnetic in Figure 7, appears in the lower region of the netic field in devices is an effective predictor of fields and then segregating the data. (Figure 5) IC when using a spectrum analyzer, but in the issues or validation of countermeasures. After the sensor automatically scans the fields upper region when using APD. ”Conventional mapping is accomplished by us- APDs are calculated and mapped in both electric This difference is attributable to the fact that ing a spectrum analyzer to obtain values. and magnetic fields, color-coded and displayed. several signals have different time fluctuations, However, in digital communication systems the This system is capable of analyzing spaces from but the same frequency in the IC. correlation between the spectrum analyzer val- approximately 0.5mm-1.0mm, can analyze We confirmed this by measuring indistinct sig- ues and the interference is too weak to be used electric components; multilayer ceramic capaci- nals in a spectrum with a large time fluctuation as an effective predictor of EMC issues. tors, ferrite beads, and can also analyze traces flow to the upper region of the IC terminals, While mapping of the adjacent electromagnetic on the electronic substrates to determine EMC while distinct signals in a spectrum with a small field by using ADP related to an value could issues. hourly fluctuation flow to the lower regions of effect the digital communication systems, this This powerful new system enables us to simul- the IC. Figure 5 APD measurement system in Electric field/Magnetic field Divider Lo. oscillator CH1 Electric wave form in time domain=CH1+CH2 Magnetic wave form in time domain =CH1-CH2 Down-converter 30MHz – 2.7GHz 70MHz IF CH2 30MHz – 2.7GHz 70MHz IF A/D converter PC A/D converter Electric APD Magnetic APD Synchronize Loop (sensing part) Output A H E Output B A=Ie-Im Ie = (A+B)/2 Im = -(A-B)/2 B=Ie+Im Ie Ie Ie : Electric field current output Im Im : Magnetic field current output Taiyo Yuden Navigator navigator3.indd 7 2009 1 Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices This verifies that noise level which is evaluated 7 09.8.3 2:46:15 PM Special Topic1 Figure6 Case sample of measuring DC-DC converter 1 Measurement by APD Hourly waveform -20 -20 -40 -40 CH1-CH2 Time [dBm] CH1-CH2 Time [dBm] Hourly waveform -60 -80 -100 -120 -60 -80 -100 -120 0 10 20 30 40 0 10 20 0-44 [msec] 30 40 0-44 [msec] Measurement by spectrum Spectrum waveform Spectrum waveform 0 CH1-CH2 FFT [dBm] 0 CH1-CH2 FFT [dBm] Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices 8 Figure7 Difference of the distribution measurement between APD and Spectrum -50 -100 -150 580.8 580.9 581.0 581.1 478.85-479.43 [MHz] Taiyo Yuden Navigator navigator3.indd 8 581.2 -50 -100 -150 580.8 580.9 581.0 581.1 581.2 478.85-479.43 [MHz] 2009 09.8.3 2:46:18 PM Figure8 Substrates of the DC-DC converter used for the experiment TPS62040(1.25 MHz Step-Down converter: Texas Instruments) 150 mm Ground 4.7μH 2.8 ohm 50 mm 22μF 22μF Input Vin:5V 200 k-ohm 610 k-ohm Output Vout:2.02 V 1 22μF Without slits ▼The magnetic field distribution Without slit -85 The magnetic The electric field distribution field distribution Noise Level [dBm] -90 -95 ▼The electric field distribution -100 -105 -110 13 17 21 25 29 30 37 41 45 49 53 57 61 737.142857MHz (57 CH) CHANNEL [CH] ▼The magnetic field distribution With slit With slits -85 The magnetic The electric field distribution field distribution Noise Level [dBm] -90 -95 ▼The electric field distribution -100 -105 -110 13 17 21 25 29 30 37 41 45 49 53 57 61 737.142857MHz (57 CH) CHANNEL [CH] Countermeasures for noise deteriorating One seg TV's reception tained an electronic circuit composed of a DC/ Our comparison therefore included both a DC converter operated at 1.25MHz and pas- conventional substrate, without slits, and a sive components; a wound inductor, two ce- substrate with slits at the ground of the substrate A test was conducted for evaluation of coun- ramic capacitors, and for voltage control. used and power supply pattern as in Figure 9. termeasures using the APD measuring system two resistors. By setting antennas adjacent to each substrate, in the One seg TV frequency range. Operating test conditions of the circuit: input noise of each the channels was received and We chose a DC/DC converter as our example voltage of 5V, output voltage-2v, and current of measured by the APD system, so its values since it is a commonly used component and about 700mA flowing in a load-resistor of 2.8Ω. could then be evaluated the interference value source of noise in One seg TV. Noise causing EMC issues is greatly influ- to the One seg TVs. The experimental substrate, Figure 8, con- enced by the circuit design on the substrate. In the first half of the channel there is little dif- Taiyo Yuden Navigator navigator3.indd 9 2009 Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices Figure9, Figure10 Noise and its distribution with slits and without slits 9 09.8.3 2:46:18 PM Special Topic1 Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices 10 From this we can determine substrates with slits For each of the channels, the comparison results half noise generated from the substrate contain- enable capacitors for input and output to oper- showed noise generated in each substrate in ing slits was lower. ate more effectively as decoupling elements in APD of probability 10-2 as shown in Figure 12. In the frequency of channel 57, the noise value higher frequency ranges. On the whole, the value of noise received by difference is great, about 8 db, and the resulting By evaluating the substrate with slits, we were the antenna adjacent to low ESL capacitor was mapping distribution of noise in the electro- able to evaluate the variation of results and ef- lower than that of the conventional capacitor, magnetic field over the top surface of the entire fectiveness of countermeasures attributable to while the values for the “Capacitive Filter” were substrate by APD measurement is shown in different decoupling elements. even lower. Figure 10. As shown in Figure 11, the measured APDs From this, we can confirm the results of coun- In the substrate without slits, we can see that act as an indicator of the value of influence upon termeasures used to the value of the noise mea- the distribution of electric and magnetic fields the sensitivity of reception in One segment TV. sured with the APD system. are dispersed widely throughout the ground and We measured three different types of capacitors As demonstrated above, we now have methods power supply. There is one particular concen- utilized as decoupling capacitors; a convention- to effectively evaluate EMC issues and counter- trated field on the input side of 5V. al multilayer ceramic capacitor, a LW reversal measures, making it possible to tailor solutions In the substrate with slits, we can see the dis- capacitor with low ESL, and a “Capacitive filter” to effectively counteract these issues. tribution occurs mainly around the IC, and less functioning as a low-pass filter that has a feature diffusion across the substrate can be seen. of four terminal geometry. Figure11,Figure12 Decoupling elements and noise Noise level(APD 10-2) -70 -75 Noise floor A)JMK212:Conventional type -80 B)AWK107:Low ESL type A) Conventional multilayer ceramic capacitor JMK212 C) Decoupling filter -85 Noise Level [dBm] 1 ference in noise demonstrated, but in the latter B) Low ESL type multilayer ceramic capacitor AWK212 -90 -95 -100 -105 -110 13 navigator3.indd 10 21 25 29 30 37 41 45 49 53 57 61 CHANNEL [CH] C) Decoupling filter Taiyo Yuden Navigator 17 2009 09.8.3 2:46:19 PM develop even more effective solutions and Taiyo Yuden has developed an accurate and measurement methods for electromagnetic effective system to evaluate electromagnetic noise, which will leads Taiyo Yuden to noise through innovative use of APD and advance and develop techniques to further mapping systems. utilization of APD. Based on the findings presented herein, it The reference list has been demonstrated that APD can be [*1] Kazama, Hori & Tsutagaya, “Suggestion for new measurement method of electromagnetic noise for EMC countermeasure of wireless applications”, Nikkei electronics no.919 pp.115-122 [*2] K. Gotoh, Y. Matsumoto, and S. Ishigami, “Performance evaluation of multicarrier wireless systems affected by a disturbance using amplitude probability distribution,” in Proc. EMC Europe Workshop, Paris, June, 2007. [*3] S, Kazama and H. Tsutagaya, “ADJACENT ELECTROMAGNETIC FIELD APD MEASUREMENT FOR ANALYZING AUTOJAMMING ISSUE ON WIRELESS COMMUNICATION SYSTEM“, 2007 IEEE Int. Symp. On EMC Recordings (CD-rom) [*4] Kazama; “Break with try and error, suggestion for new method of measurement technology of high frequency EMI”, Nikkei electronics no.809 pp.105114 effectively used in evaluation and analysis of EMC issues and countermeasures as in our tests illustrating the deteriorating sensitivity of reception in a One seg TV signal. Going forward, Taiyo Yuden confidently supports the use of APD systems as an effective method through which noise issues can be quantified and solved by prposing proper countermeasues and components. By collaborating with customers through ADP evaluation technologies, we can 1 Special Topic 1 Utilizing Amplitude Probability Distribution Mapping Measurements in the assessment and suppression of electromagnetic compatibility issues in digital wireless devices Coming Trends in the Years Ahead Reference: Taiyo Yuden's products for suppressing noise Line up Products for suppressing noise SMD type Serial connection type Ferrite beads Parallel connection type Capacitors Common mode choke coils Leaded type For DC For AC Ferrite beads BK series, FBM series Applicable to noise in any type of equipment because of affluent lineup Multilayer ceramic capacitors, three-terminal type, LWDC series For power supplies with low impedance and PCs with low ESL Common mode Choke coil CM01series For high-speed differential signal [Mbps -Gbps], suppressing noise in MHz to GHz band Common mode Choke coils for DC line Common mode Choke coils for AC line TLF series Taiyo Yuden Navigator navigator3.indd 11 2009 11 09.8.3 2:46:20 PM
© Copyright 2026 Paperzz