LIMP tutorial
Contents
1. Mw g Example an 8 inch driver will have Myp of 15 50g A 25g weight should be trialled for the first measurement Points 3 and 4 are as for the volume method above Membrane diameter inches 5 Measure impedance with the added mass Figure 3 4 1 i Magnitude ohms Impedance Phase Resonance shift Added mass method 5 10 20 50 100 200 500 1k Cursor 1015 87 Hz 14 76 Ohm 37 4 deg Frequency Hz File B139 SIN plusMasse lim 2009 04 18 20 42 58 Added Mass rot Figure 3 4 1 Driver impedance with red and without green added mass 6 Get TS parameters Similarly to the previously described volume method click on Analyze but then choose Loudspeaker parameters Added Mass Method Enter the User Input data required 24 Loudspeaker Parameters Added Mass Method Loudspeaker parameters Fs 51 16 H2 Re 5 50 ohms de Le 251 49 uH Membrane diameter cm L 220 61 uH Re 7 47 ohms QE 0 32 Mes 0 56 mes 3 12 Added mass tg Mms 10 96 grams ee ee Rms 1 082545 kg s Estimate TSP by LSE minimization Iw Cms 0 000905 miN Was 6 16 liters Estimate voice coil resistance Re f Sd 80 12 cm 2 Bl 7 198573 Tm Estimate lossy inductor model 1w ETA 0 27 Vo Lpt 82yfimi 88 11 d TT Added Mass Method Added mass 11 00 grams Diameter 10 10 cm Calculate parameters Figure 3 4 2 Estimation of TS parameters using the added mass2. 500 kOhm e Amplifier power P 20 watts Using Ohm s law the maximum output voltage is calculated for a load impedance Z 8 ohms as follows U SQRT P Z U SQRT 20 8 12 65V Gy Vin max Vout amp Max 1 679 12 65 0 1327 17 54dB Thus a voltage divider giving about 18dB of attenuation is required Soundcard 5 R1 i Rx Zin R2 Zin R2 1 i G Rx R1 Rx 2 RI Rx G Rx 3 If the input impedance of the sound card Zin 500 kOhm and R2 10 kOhms R1 is calculated as follows Rx 500 000 10 000 500 000 10 000 9803 92 ohms R1 Rx G Rx 9803 92 0 1327 9803 92 64076 42 ohms gt 68 kOhm Gy 9803 92 1 68 000 9802 92 0 126 17 99dB Additional hints regarding dimensioning of the voltage divider see section 6 The Zener diodes shown in grey are not essential but offer additional protection for the soundcard line in Use them at your discretion 2 2 Working in LIMP LIMP has similar top menu commands to the other programs in the ARTA family S or gt ta M Go PLC majme cer pet Z Fatt 2000 5 A Magnitude ohms Phase g m dej wal o FrequencytHz LOdE R 0dE Impedance Measuremen Figure 2 2 1 LIMP opening screen Open file Copy Use thick pen Calibrate Generator RLC Meter Amplitude Phase Gi a i gt cA mm N RLC Mag MP 7 Z a ie Save file B W color Record Cable compen
3. impedance Z is corrupted due to differences in sensitivity of the two measuring channels a phase value greater than 90 degrees may be detected which can lead to a 180 degree phase reversal Figure 5 3 245 0 Magnitudefahms Impedance shini E 20 0 1128 sf i at LLI 395 0 0 0 pe E OR E M O EN E ve Pe 2 COT IMI 124 5 COE Ss a ee i ee lee 0 0 100 1000 10000 Cursor 1009 3 Hz 32 62 Ohm 89 7 deg Freguency Hz Figure 5 3 Uncalibrated impedance determination for a capacitor of 4 7 uF 250V Figure 5 3 shows the result of a capacitance measurement without calibration Until about 1200Hz the phase runs at nearly 90 degrees and so gives the impression that the component is an inductor After calibration the phase behaves as expected over the entire frequency range Figure 5 4 Magnitude ohms Impedance Phase 7 245 0 30 0 Rr p p H pra ot RLU a a ert SN trimi po A E E LA H Oe a 100 1000 10000 Cursor 242 7 Hz 132 99 Ohm 89 4 deg FrequenojiHzi Figure 5 4 Calibrated impedance determination of a capacitor of 4 7UF 250V Not all LIMP users will have the above problem because it is present only when the voltage V2 across the impedance is higher than the voltage VI across the generator To get around this the sensitivity of the test probe can be changed or the input channels exchanged If the latter is done the reference c
4. Impedances of various measurement setups We can compensate for cable resistance by using the E i Cable Impedance Compensation function To do Inductive impedance at frequency 10044 38Hz this the shorted test lead is measured The reading should be similar to that shown in Figure 2 2 7 Place the cursor at approximately 45 of phase and click R 29 972 mohm the RLC button in the main menu bar The Impedance at cursor window appears showing the L 974 699 nH resistance and the inductance at the cursor position These values are transferred to Cable Impedance Compensation Figure 2 2 8 Note the units If the checkbox Automatically subtract cable impedance from Measured Impedance is activated each measurement is automatically corrected for the impedance of the cables 13 Cable Impedance Compensation Automatically substract cable impedance from measured impedance Cable resistance ohm 0 0299 Cable inductance nH 979 Figure 2 2 8 Cable compensation menu An alternative method for compensating for the test leads is by using the Subtract Overlay function This is a Zero measurement of the test lead stored as an overlay and subtracted from each measurement 3 Determination of TS parameters LIMP supports two methods for determination of TS parameters e Volume method closed box of known volume Figure 3 1 1a e Added mass method Figure 3 1 1b In principle both methods are equivalent
5. but the volume method is preferred if the resonance frequency Fs of the speaker is very low The added mass method is much easier as you do not have to construct a box instead you simply add mass to the driver Added mass will be accurate as long as the Fs is not below 25Hz or so as most soundcards start to lose accuracy at very low frequencies Vcone Vr Veox Veone Figure 3 1 1a Volume closed box method Figure 3 1 1b Added mass method 14 3 1 Preparation for TS parameter measurement Before measuring it is worth reading the seminar Loudspeaker Parameters by Neville Thiele and Richard Small that was given in 2008 2 and the guidance issued by a well known driver manufacturer 3 Among other topics the seminar addressed the conditions for the determination of TS parameters signal strength measurement position mounting or clamping conditions and driver break in 3 1 1 Excitation signal strength Remember that TS parameters are measured using small signals Small recommends limiting the signal to a level just sufficient to ensure a clean measurement SB Acoustics in their Technical Note 3 recommend approximately volt for measuring midrange speakers at their resonance frequency The default AES2 2012 standard 4 recommends 0 1 volts for typical measurements As mentioned earlier the energy contents of the available excitation signal types in LIMP PN stepped sine differ Figure 3 1 2 shows impedance
6. curves for both excitation signals with identical gain The red curve is the stepped sine and the black curve is pink noise The more energy rich sinusoidal signal gives a lower apparent resonance frequency Shifting resonance frequency with changing excitation amplitude is a well known phenomenon Cms is not a constant and different strength excitation signals and different measurement methods should not therefore be expected to yield the same results r Magnitude ohms Impedance 538 5 52 0 45 5 39 0 32 5 Avg 0 26 0 19 5 13 0 6 5 0 0 5 10 20 50 100 200 500 Cursor 51 74 Hz 36 00 Ohm Frequency Hz Figure 3 1 2 Impedance with different excitation signals red stepped sine black PN It is important to check if the parameters are consistent to ensure accurate calculation simulation Hi Fi Selbstbau in Cologne offer a useful tool called TSP Check 5 Alternatively the Fs Qts ratio can give a rough indication as to whether results are reliable 15 3 1 2 Driver break in AES2 2012 recommends that loudspeakers must be driven for several minutes before measuring TS parameters to in order to avoid drifting resonance frequency readings 4 SB Acoustics 3 drive their speakers for about 10 minutes with a sinusoidal signal at around 0 8 times the Fs The voltage is chosen so that the speaker operates with maximum deflection The driver should be allowed to cool to room temperature afterwards before measurements are take
7. sional SUL IO UN r du S ya 15 9 L2 DA V N DLCAK Mn nn ns assadadly ndshesGanedie cate eadacauegeietadlyoadsk aati eden 16 SE Pe KORIMONTI Nd pem te i ack a 16 3 1 4 SpE AS UE CMe INE ace ee corte in disloyal atva aes enkaccnueie a eadnns 17 515 Re peAS re eii erce e ta 17 3 2 Speaker models for TS parameterdeterminatioi e esiin eee eee eee eee eee eee enen enen eee 18 3 3 TS parameter determination volume method aa aaaaaaaaanaannnnnnnnnnn nene eee e eee eee eee eee eee nene eee eee 21 3 4 TS parameter determination added mass method aaa aaaaaaaaaaaaanaaaaa anen nene eee reve eee eee eee nenen e neve teze zeese Working with overlays and targets in LIM Pa a aaa aaa e e enen eee eee eee eee nene eee nenen ete te zere ee eee 26 en e ER E E 28 Accuracy of impedance Measurements s nici sh nis sesi sdu a e sd et 31 TG 92 Foreword Because of the increasing scope and complexity of the ARTA Handbook the LIMP Manual is now presented separately to those for the other programs in the ARTA family LIMP is for impedance measurement and determination of Thiele Small TS parameters LIMP can also be used as an LCR meter Available excitation signals include pink noise PN and stepped sine This handbook has been written to help acquaint users with the LIMP part of the ARTA family of programs Note however that it is not a substitute for the original user s manual that comes with the software and the two should be used in parallel w
8. u 1 E f i 4 E p i 4 E a E N p H CI q E E i a F i i Fl i i a K i i Ci 4 K i i i E K i i Ci K i i i 4 pF i E i 4 BF i E i 4 BF Hi E U amp E i 7 amp a R 1 amp Fi N AN E E 1 N N LI J Lf amp LI Y H Li 1 a a i i g I i a i a i i i 0 E E E E C C E CI i s E E E E E E E E 1 3 SS See Fj E i i i i 1 6 FE i i CI i a BO i i KEM Ga i i i 4 KON i i E I i a 8 oF i i i K i Gi a i i H KON i 1 CI S F Ff I i f i 4 KE OJ a E EJ 8 fF F i E d j i 7 K KENI E 4 1 EJ i i 1 8 8 F I 1 CI t i i H KE Ci CI C i i 8 KOR F a E 8 6 fF u Fi E kd i Fi Fi 8 FF H pl H E HE i fe Hot Fl Ta i i i i i 4 4 H i i i 4 HI i ji IO 4 E F a f wt ee ee ee ge e e eps sp a pn pe nn me mr ee on e ps ee ee ee eee I E 1 I p 1 1 a HO Li 1 d p U K EU 1 p 1 1 LI E LI p ki 1 Fi i Eft i E a 8 amp i Fi i H KO FR 4 i 3 Fl i da I i Hi 7 oo amp i i E i i I Fl i i na i i E EE amp i i i E i i i Fl i E f i i EEE i 1 E i a 3 Fl i iG f i i E a Fl K i Hd d 5 TH2821 deviation 6 1 5 8 i EN i occa ka a a i r o y ae 0 01 0 10 1 00 10 00 100 00 Rx in kohm Figure 5 5 Resistances LIMP vs RLC meter 30 6 Accuracy of impedance measurements The perfor
9. 0 10k 20k Cursor 20 0 Hz 6 42 Ohm 10 4 deg Frequency Hz Estimate loudspeaker parameters Closed Box method R 0dB Impedance Measurement Figure 3 3 1 Impedance of driver in closed box black and free air red 6 Analyze driver parameters Go to Analyze and Loudspeaker parameters Closed box method as shown above you will need to enter the User Input parameters required see also section 3 1 Loudspeaker Parameters Closed Box Method a x Impedance data User Input fimax 46 94 ohm ze s Fenl 85 0 H Voice coil Resistance ohms 5 5 imin 6 23 ohm Membrane diameter crn 10 1 over data Amar 53 37 ohm Fso 50 5 Hz Closed box volume lit 5 4 22min 6 18 ohm o Optimization Frequency shift 40 5 t optimal shift is 20 to 50 Estimate TSP by LSE minimization IW Estimate voice coil resistance Re Estimate lossy inductor model lw Le L2 R2 Cancel Figure 3 3 2 Calculating TS parameters Calculate parameters If the input fields are greyed out and will not accept data this means that no overlay has been defined 22 Click Calculate and LIMP will estimate the TS parameters Figure 3 3 3 Loudspeaker Parameters Closed Box Method Loudspeaker Parameters User Input Fs 51 16 Hz Le 251 49 UH L 220 61 UH R2 7 47 ohms Qt O 32 Ges 0 36 Closed box volume flit Ons
10. 2 6 Channel calibration error message The error message suggests possible causes for the excessive channel difference Correct reference channel Figures 2 1 2 2 1 6 and 2 2 4 Same gain for both input channels see ARTA Handbook Section 2 Voltage dividers equal if present Cables and connections OK 12 2 2 2 Compensating for the test leads A general principle is to keep all leads as short as possible If you do that you can probably skip this section Even the shortest copper wire has some resistance 0 017241 Qmm m Figure 2 2 7 shows the impedance response of different copper test leads From the top 5 5m of 2 5 mm 4m of 4 mm 0 9m of 4 mm red ARTA Measuring Box with Rref 20 ohms blue headphone output with Rref 68 ohms Depending on the length and cross section values of 0 03 Q to 0 23 Q are recorded The measured values cannot be accounted for entirely by the resistance of the cables only Contact resistance in switches terminals and banana plugs also contribute as shown by the example of the ARTA measurement box when compared with the headphone version Maqnitudefohms Impedance 0 i E 90 0 45 0 0 0 2 5 mm 5 5m 45 0 4 0 mm 4 0m 90 0 4 0 mm 0 9m an MessBox Avg 15 4 0 mm 0 9m an Phone out E Mi F 10 20 50 100 200 500 1k 2K 5k 10k 20k Cursor 1003 91 Hz 0 03 Ohm 12 7 deg Frequency Hz 2 5mm 2 5 5m 4 0mm 2 4 0m 4 0mrm 2 0 9m 4 0mm 2 0 9m Phone Figure 2 2 7
11. 3 12 Mims 9 20 grams Rms 0 909186 kg s ms 0 001078 min Was 9 72 liters Sd 80 12 cm 2 Bl 6 597049 Tm ETA 0 35 Lof2 83Viin 88 87 dB Membrane diameter cm Optimization Closed Box Method Box volume 5 40 liters Diameter 10 10 cm Figure 3 3 3 Calculated TS parameters 7 Copy Use Copy to copy the data to ASCII files The output is as follows Thiele Small Parameters Fs 79 85Hz Re 5 75Ohms dc Qt 0 63 Qes 0 68 Qms 8 02 Mms 13 47g Rms 0 842902kels Cms 0 000295m N Vas 6 64L Sd 126 68cm Bl 7 555168 Tm ETA 0 48 Lp 2 83V 1m 90 33dB Closed box method Box volume 5 40L Diameter 12 70cm Re 5 50 ohms de Voice coil Resistance ohms 5 5 Estimate TSP by LSE minimization I Estimate voice coil resistance Re Estimate lossy inductor model W Le L2 IR2 ti a I Export Cancel Calculate xpor ance parameters i Copy OK xj 23 3 4 TS parameter determination added mass method 1 Calibrate 2 Specify test mass The mass to be added depends on the diameter aa membrane area of the driver As with the previous method a resonance shift of between 20 and 50 is needed Adding mass of the order of the membrane mass Myp will give an 100 approximate decrease in resonant frequency of 30 If you do not know the Myp of your driver the diagram to the right will give a rough estimate 40
12. LIMP MANUAL 82 Impedance Measurement Thiele Small Parameters TSP RLC Meter Based on the original LIMP Manuals Original tutorial in German by Dr Heinrich Weber Original manuals in English prepared by Dr Ivo Mateljan Weber Mateljan Translation into English of Version 2 40D ARTA 1 80 Christopher J Dunn Hamilton New Zealand September 2014 Contents l 2 gt S Sk a DOC tonto EIM Pore E E T a EEE EE 3 1 3 Pin assignment for cables and connectors seeseesssessssssssssseeeeterrerssssssssssssssescerrereresesssessss 4 Measurement principles and setup in LIMP Ga asni cases usta a 5 De SE U anea RCS eI Mtr tee EAS Tne TE ieee ea RT TeME tere Te BONS eTE ROT EON ert oe eee tae ae ener ene ert oe eet a 5 DLE Impedance measurement using the soundcard headphone ac Ko aaaaa nana aaa aaa aaa aaa nanen nenen seees 5 2 1 2 Impedance measurement using a power amplifier aaa aaaaaaaaaa aaa ne eee eee eee eee eee nene neo eee neve eee eee 7 PRS ummm SCY a eb N AMER ta ec nenee ene ne ner Pree aee oer eem ar Nene Aree eee ent cary reese ema Oar enn tery 8 PAM E Basic setine TD LIMP arc a ated ied acca 9 Ze Compensating for the test leads nana aa aaaaaaaaa aaa eee eee eee eee eee eee eee nenen nenen eee e ee tee eee ee eee eee 13 Determination GE 1 Spar ame LOES ox ica venpoees a S divin Mas aren aan S 14 3 1 Preparation for TS parameter MeasSUreMeNl cc aaa anen ennnnnenen enen eee ee eee ee eee erret 15 e EXCITATION
13. and coils by calculating the resistive inductive or Capacitive component of the impedance Figure 5 1 shows an example of the impedance curve of a coil with a nominal inductance of 0 33mH Magnitudetohme Impedance Phase g 0 fy 30 0 56 4 45 0 r66 0 0 Bf 2 45 0 a 6 90 0 Avg 30 4 20 0 19 2 L 9 6 fl FP 0 0 10 ZU 50 100 2010 S00 1k zk 5k TOK 20k Cursor 3 66 Az 0 33 Ohm 2 3 deg FrequencytHz O33 mA ohne Figure 5 1 Impedance trace of a 0 33mH inductor Impedance at cursor le The operation Analyze followed by RLC Impedance Inductive impedance at frequency 1003 91Hz values at cursor position yields the result shown on the left R 312 388 mohm LIMP tells us that the measured impedance at the cursor position has a resistive component of L 336 919 uH 0 312 ohms and an imaginary inductive component of 0 336mH Capacitors and resistors can be measured in a similar fashion When carrying out RLC measurements it is important to calibrate beforehand This is because even with small differences in sensitivity of the input channels of the sound card e g 0 1dB LIMP may under certain conditions give spurious results because of phase errors Rret Ss r 7 generator Reference resistor Measured SOU mcard VA Impedance npin impedance i ee ee ee ee Fel Figure 5 2 Setup for impedance measurement 28 In the event that the measurement of voltage V1 across the generator and voltage V2 across the
14. e of the boxes are checked the classic method for determining TS parameters is applied see LIMP User Manual 1 section 5 2 2 If TSP by LSE minimization is checked the nonlinear least square error minimization procedure is activated whereby we minimize the squared difference between the measured impedance Zy and modelled impedance Zyr The quality of the optimization can be graphically controlled by using the F3 function key see Figures 3 2 3a and 3 2 3b If Voice Coil Resistance Re is checked LIMP determines the DC resistance of the voice coil from the measured impedance curve This is especially useful if no ohmmeter multimeter is available Please note that if this box is checked any value entered for Re will be ignored by the software Checking Lossy Inductor Model activates a selection of voice coil optimization Zje equivalent circuit models drop down menu Rus Cus Li Glecincal circuit Mechanical circuit al Wideband analogous circuit b Low frequency analogous circuit c Models for Zp Figure 3 2 2 Equivalent circuit of a loudspeaker with various models for Zje Three different models are available Figure 3 2 2c Use of these equivalent circuits in a simulation can improve the goodness of fit of the impedance curve significantly over the entire frequency range 19 Magnitude ohms Impedance pon eas cmt vd P Loudspeaker Parameters Fs 96 94 Hz Re 6 60 ohms dc Qt 0 51 Qes 0 56 Qm
15. eck specification professional soundcards have input impedances of IMOhm or an upstream input buffer 3 Inductive or capacitive effects can be contributed by excessively long measurement cables This also applies to terminal or connector resistance a Use short measurement cables with sufficient cross section 1 5mm_2 b If longer measurement cables must be used place the reference resistor as close to the speaker terminal as possible c Ensure secure connections Use good quality connectors and terminals see note Note test leads as shown on the right are often a source of error in measurements on loudspeakers Because the cables are often just clamped at the crocodile clips they introduce variable contact SS resistance This does not help with reproducibility of measurements If leads of this type are used connections should be resoldered and checked 31 7 I 10 References ARTA Support Internet cited 2014 Sep 21 Available from http www artalabs hr support htm Thiele N Small R Loudspeaker parameters Internet AES 124th Convention 2008 cited 2014 Sep 20 Available from http www aes org tutorials SB Acoustics Technical Notes Internet Measuring Thiele Small Parameters cited 2014 Sep 20 Available from http www sbacoustics com index php technical notes AES2 2012 standard for acoustics Methods of measuring and specifying the performance of loudspeakers for professional app
16. ed Rpc 4 7 0 5591 0 8368 3 14 Ohm manufacturer s specification 3 10 Ohm n b As of version 1 8 LIMP provides an option for the determination of Re from the impedance curve see section 3 2 3 2 Speaker models for TS parameter determination As of version 1 8 LIMP provides in addition to the classic method depicted to the right 8 9 10 additional functionality for the determination of TS parameters Detailed information can be found in Mateljan amp Sokora 9 The following is a brief summary In the Loudspeaker Parameters window the User Input parameters must be determined and entered These will vary depending on the method chosen i e sealed volume or added mass see also sections 3 1 4 and 3 1 5 Loudspeaker Parameters Closed Box Method Impedance data User Input Zimax 0 00 ohm Feoi 5 0Hz Voice coil Resistance ohms 5 5 Zimin 0 00 ohm Membrane diameter cm 10 1 Closed box volume flit 10 Optimization Estimate TSP by LSE minimization V Estimate voice coil resistance Re Estimate lossy inductor model Le L2 R2 Lalla Le L2 IR2IIK Calculate Figure 3 2 1 Loudspeaker Parameters window Closed Box method shown 18 The Optimization area contains three check boxes and a drop down menu Use the check boxes to select the type of optimization required e Estimate TSP by LSE minimization e Estimate Voice Coil Resistance Re e Estimate lossy inductor model If non
17. hannel set for ARTA in the measurement setup must also be changed In order to obtain correct capacitance and inductance values the cursor should be set to a frequency at which the impedance is less than 100 Ohms This ensures that the measurements stay within the range of about 1 tolerance for further details see section 6 The following examples demonstrate the performance of LIMP as an RLC meter compared with a 4 wire RLC meter TH2821 and a mid range RLC meter PEAKTECH PT2165 Measurements were taken with a mid range sound card EMU Tracker Pre via the headphone jack without a voltage divider see also section 6 All values with the exception of two small inductors were within 1 The inductor measurements were notable in that the PEAKTECH PT2165 gave almost exactly the same results as LIMP even for the inductors that were outliers with the TH2812 0 18mH and 0 33mH 29 Nominal value uF LIMP PT2165 TH2821 A TH2821 Smooth bipolar 4 675 4 674 4 678 0 06 Smooth bipolar 8 745 8 739 8 755 0 11 Smooth bipolar 34 328 34 300 34 390 0 18 Table 5 1 Capacitor values LIMP vs RLC meters Nominal value mH LIMP PT2165 TH2821 A TH2821 Table 5 2 Inductor values LIMP vs RLC meters ann n SS DNS DaD aSa Aad v m D Aa SRS A ee p e M i i i tiG f i E i Fl i iG t i i i k Fl i iG i i k i i tiG i i b Fl i iG i i k Fl i tiG i i Fl fi i i r f Fi LI i i i i i f i E p
18. ion results using parameter sets from Figures 3 2 3a and 3 2 3b 3 3 TS parameter determination volume method Determination of TS parameters using the volume method is carried out as follows 1 Calibration section 2 2 1 2 Set test volume A well sealed test box of known volume with an aperture over which the i inverted driver can be placed is needed A rough estimate of the required Ti volume can be made using the table left The volume of the enclosure needs to be such as to cause a 20 50 shift in the resonance of the loudspeaker Parameter inputs are controlled in LIMP Figure 3 3 2 20 For example a 17cm driver will need a box of around 12L Bear in mind J5 that when entering the parameters into LIMP the exact volume to 30 include the space taken up by the inverted cone will be needed Figure 3 1 1a 30 3 Measure the driver s impedance in free air 4 Save the free air impedance using Overlay then Set yellow trace 5 Repeat the impedance measurement in the test housing closed box Figure 3 1 1a 21 IN PR 5 4Ltr 3 lim Limp File Overlay Edit View Record Setup BA Help E lal Dm yr if bb Loudspeaker parameters Added mass method Loudspeaker parameters Closed box method Gen Pink PN Fstart Hz 20 RLC impedance values at cursor position 3 5 Magnitude ohms Impedance Phase Resonance shift Volume method Fe z Set Fri gt E E 8 4 0 0 20 50 100 20
19. ith each other Further information can be found on the ARTA website which contains the most up to date releases and application notes Although we aim to continue to supplement and update this handbook in line with the continuing development of the software we would ask for your patience with any discrepancies that may arise from time to time Suggestions for improvements and corrections are always welcome 1 Introduction to LIMP 1 1 Installation requirements To use the ARTA suite of programs you will need e Operating system Windows 98 ME 2000 XP VISTA Windows 7 Windows 8 e Processor Pentium 400MHz or higher memory 128k e Soundcard full duplex Installation is very simple Copy the files to a directory and unzip them That s it All registry entries are automatically saved at first start up 1 2 Equipment The following is a brief summary of the equipment required accompanied by some basic directions and cross referenced to more detailed information elsewhere Soundcards There are three types of soundcard e Standard onboard soundcard found typically on a computer motherboard e Plug in cards for PCI or ISA bus e Soundcards connected via USB or firewire Essentially all three types are suitable for use with LIMP if they have an output channel Line Out and two input channels Line In Note however that onboard sound cards fitted to laptop computers often have a single mono channel only identified as the micropho
20. l Right 7 Frequency increment 1 24 octave vi FFT size 32768 Reference Resistor 10 59 Min integration time ms 200 Averaging Linear Frequency range Hz Transient time ms 100 Max averages 100 20000 High cut off ee 1 Intra burst pause ms 100 Asynchronous averaging W Low cut off 2 Sampling rate Mute switch off transients W Default cancel OK Figure 2 2 4 Measurement Setup General measurement parameters are defined under Measurement Config Measurement config e Reference channel default is the right input channel e Reference Resistance see sections 2 1 1 and 2 1 2 the exact value must be determined by measurement e Upper frequency limit see below e Lower frequency limit see below Reference channel Right Reference Resistor 10 59 Frequency range Hz r The frequency limits can also be controlled via the top menu High cut off 20000 a Fstart Hz 10 Fstop Hz 20000 4 Low cut off 2 Sampling rate 47100 Stepped sine mode In stepped sine mode the parameters for excitation with the stepped sine signal are defined Frequency increment 1 24 octave All parameters for this part of the program are set out in Min integration lime ins 200 ARTA Handbook Section 9 1 The default settings are suitable for usual impedance measurements Transient time ms 100 Mute switch off transients mutes the clicks at the end of each sine wave packet Intra bu
21. lications Drive units AES 2012 TSP checken einfach gemacht Internet Available from https hifi selbstbau de grundlagen mainmenu 35 verschiedenes mainmenu 70 199 tsp checken einfach gemacht Vance Dickason The Loudspeaker Design Cookbook 7th ed Peterborough NH Amateur Audio Press 2006 Anderson BE Derivation of moving coil loudspeaker parameters using plane wave tube techniques Master of Science Brigham Young University 2003 D Appolito JA Testing loudspeakers Peterborough N H Audio Amateur Press Distribution agents Old Colony Sound Lab 1998 Mateljan I Sikora M Estimation of loudspeaker driver parameters Croatia Acoustical Society of Croatia 2012 IEC 60268 5 Sound system equipment Part 5 Loudspeakers Geneva Switzerland International Electrotechnical Commission 2003 32
22. lors and gnd style B W background color Lise thick pen Substract overlay Add overlay Product selection or testing requires assessment against prespecified tolerances and limits We can do this in LIMP by importing text files with txt or zma suffixes These can be generated using a text editor or MS Excel as follows freq mag phase 91 1 0 0 0 0 Table 4 1 ZMA data representing tolerance limits fs 104 91 1 22 0 0 0 12 9 Hz see vertical red lines in Figure 4 1 116 9 0 0 0 0 116 9 22 0 0 0 Maqgnitude ohms Impedance PR ir p 19 9 17 8 15 7 13 6 115 g 4 20 50 100 200 500 1K 2K 5K Avg 0 10K 20K Cursor 10 00 Hz 3 64 Ohm Frequency Hz fs und f 3s Figure 4 1 Tolerance limits on a loudspeaker impedance curve Impedance profile slopes can also be used to set tolerance limits Figure 4 1 26 Figure 4 2 shows the measurement of the resonance frequencies of 32 small woofers From a previous batch the resonance frequency fs was believed to be 104 12 9 Hz Using Load target curve the tolerance values from Table 4 1 were imported as a ZMA file i Magnitudefohms Impedance Phase 40 0 45 0 0 0 45 0 90 0 Avg Exp 20 50 100 200 5001 Cursor 10 00 Hz 3 64 Ohm 5 0 deg FrequencyiHzi Ts Mittelwert 3s 104 12 9 HZ Figure 4 2 Loudspeaker selection tolerance limits Fs 3s 21 5 Using LIMP as an RLC meter LIMP determines the value of resistors capacitors
23. mance limits of LIMP as an RLC meter is determined by the input impedance of the soundcard and the measurement setup Measurements will be optimized if the soundcard has a high input impedance and will work without an upstream voltage divider because the parallel resistance of the voltage divider in this situation produces interference The limits of the setup are shown by noisy measurements e g Figure 5 5 right panel 100 kOhm For this reason the ARTA Measurement Box see ARTA Application Note 1 represents a compromise between accuracy and ease of use If the resistor values in the application note are used errors of measurement up to about 100 ohms should be less than 1 Note If the soundcard has an input impedance gt 500 ohms then the resistor values of the voltage divider RI R2 and R3 R4 in the Measurement Box can be increased by a factor of 10 Measurement errors up to 1000 ohms should then be less than 1 If the test setup is correct impedance measurements with LIMP should be subject to errors of less than 1 If this is not so one of the following is usually responsible 1 The sensitivities of the soundcard input channels are different 2 The input impedance of the soundcard is too low 10 20 kOhm 3 The cables between the power amplifier and speaker are too long These problems can be addressed as follows 1 Calibration of the sound card see section 2 2 1 2 Use of a soundcard with high input impedance ch
24. method 7 Copy Export the data using Copy to Clipboard or Export in CSV file If you intend to apply statistical analysis to the data use the CSV option as it gives you immediate access to full Excel functionality 25 4 Working with overlays and targets in LIMP As of Release 1 8 the LIMP Overlay menu provides for targets and overlays The Set as Overlay function allows the loading of overlays only Reactivation of the function loads the current curve as an overlay and deletes the previous overlay automatically This restriction is in place because the overlay function in LIMP is intended primarily for the calculation of TS parameters The Set as target function allows the use to load numerous targets although they are all shown in the same color Overlay Edit View Rec Set as overlay curve Load overlay curve Delete overlay curve Set as target curve Load target curve Delete target curves Two additional simple arithmetic functions are included in the Edit View Record Setup Edit menu Add Overlay and Subtract Overlay An example of an application for Subtract overlay has already been covered 1 e compensation for test leads as described earlier This new functionality opens up possibilities for among other KA things quality assurance measures Examples include the selection Jf of loudspeaker drivers auditioning PA hire equipment etc Some examples and usage tips follow Copy Co
25. n Other sources suggest a break in period of several hours Conversely Vance Dickason suggests that break in is required only for the early detection of hidden defects in the driver 6 3 1 3 Speaker mounting Small recommends that the driver under measurement be firmly clamped in position Figure 3 1 5 In his presentation he demonstrated the effects of unwanted unintentional influences on a mass spring system Figure 3 1 3 shows the effect of mounting the driver on a firm MDF red or soft foam blue surface horizontally or vertically The driver in the left panel has an apparent membrane mass of 11g while that on the right is 43g An additional resonance is also produced by mounting on the foam substrate Kagnitude ohmej impedants Kagnitude ohmej Impedanca 3 LI 0 Byg 10 a 50 100 10 a 50 100 eo Ne Hr 7 25 Orn Faguen Hz Drew Ne Hr 6 21 ghm Faguen Haj eo Hakonas hate Uer bgs ibbiy Seins rna B122 hete Unei pis Scheuinshort roti Figure 3 1 3 Influence of the mounting surface on speaker measurements red foam blue MDF Driver Axis Horizontal Figure 3 1 4 TS parameter measuring positions Driver mounting is a subject discussed frequently on forums and various arrangements have been suggested Regardless of the opinions expressed the fact remains that any driver mounted vertically Figure 3 1 4 will have its voice coil position shifted by the force of gravity F mg where e 9 81 mi s Small and othe
26. ne input Mic In Amplifier A power amplifier with linear frequency response and power 5 10 watts is adequate The output impedance should be lt 0 05 Ohms An inexpensive solution that meets these requirements and is small and easily portable is the Thomann t amp PM40C see also ARTA Handbook Section 5 4 ARTA Measuring Box The ARTA Measuring Box is not absolutely necessary but it does make life a lot easier When switching between acoustic and electric measurements the annoyance of having to swap multiple cables is replaced by the simple flick of a switch see Application Note 1 1 Multimeter A multimeter for use with STEPS is not strictly necessary but it is nevertheless indispensable for the calibration of the measurement equipment to be used Besides a good meter is a useful tool for all manner of other measurements If you do not have a multimeter you should ideally opt for a true RMS type There are plenty of options on the market for well below 100 Cables Several cables are required all of which should be of good quality Poor connections inadequate shielding etc can interfere with measurements see also ARTA Handbook Section 6 1 1 Keep all cables as short as possible 1 3 Pin assignment for cables and connectors Unbalanced Balanced Ze STEREO JACK Sleeve earth GROUND SHIELD Pin 1 earth GROUND SHIELD Tip Pin 2 Ring Pin 3 Figure 1 3 1 Pin assignment for c
27. onnecting cables 2 Measurement principles and setup in LIMP 2 1 Setup The principles underlying measurement in LIMP are illustrated in Figure 2 1 1 Rref E 7 generator Reference resistor Measured iar b V1 impedance iipelanee i SS i e Figure 2 1 1 Impedance measurement circuit Voltage VI Line in right channel through reference resistor Rref and speaker and V2 Line in left channel via the speaker gives the impedance Z Rref x V2 H VI O V2 H In the original LIMP User Manual reference is made to two test setups e Impedance measurement using the headphone jack of the sound card Figure 2 1 2 e Impedance measurement with power amplifier Figure 2 1 6 2 1 1 Impedance measurement using the soundcard headphone jack The easiest way to perform an impedance measurement with minimal additional equipment is to use the headphone output of the sound card All you need is a reference resistor Rref and a small amount of cable The measurement setup is shown in Figure 2 1 2 According to the LIMP user manual the reference resistor Rref should have a value of 100 ohms but this can vary between 33 and 100 ohms depending on the sound card Loudspeaker soundcard Figure 2 1 2 Impedance measurement using the soundcard headphone output The high value of the reference resistor is dictated by the fact that soundcard headphone outputs are usually not designed for connection to speakers Figu
28. re 2 1 3 compares the impedance curves of a standard headphone and a speaker The much reduced relative impedance of the speaker shows clearly how the headphone output could suddenly be asked to deliver far more power than it was ever designed for Maaqnitude ohms Impedance al N era te P it ool LL in eee E a E oA A wo AA A s AAU A oo APA HHN mT so T Sheer TTT 0 0 20 50 100 200 500 1k 2k 5K 10K 20K Cursor 10 00 Hz 34 33 Ohm FrequencytHz Avg 0 Figure 2 1 3 Typical impedance curves for headphones black and a speaker red Figure 2 1 4 shows the specifications for the headphone outputs of two common soundcards e The ESI UGM96 delivers a maximum output voltage of 4 dBV 1 0 104 4 20 1 584V e The RME Fireface UC provides 19 dBu 0 775 104 19 20 6 907V Line amp Kopfhorer Ausgang Type Stereo 6 3mm Klinkenbuchse Max Ausgangspegel 4dBV THD N 0 0035 A gewichtet Ausgangsleistung 100mW Max 32 ohm Impedanz 32 600 ohm Figure 2 1 4 Soundcard headphone output specs in German ESI UGM96 Klinkenbuchse jack Ausgangspegel output level A gewichtet A weighted Ausgangsleistung output Impedanz impedance aL o LINE IN HEADPHON LOUDSPEAKER Figure 2 1 5 Practical implementation of Figure 2 1 2 Figure 2 1 5 shows a practical implementation of the circuit illustrated in Figure 2 1 2 Such a set up is not necessarily essential but it does help
29. rs 2 7 therefore recommend mounting on the horizontal axis Figure 16 3 1 4 in arrangements similar to those shown in Figure 3 1 5 Those who wish to perform repeated measurements should consider the construction of permanent rigid and sufficiently heavy jigs for driver mounting Figure 3 1 5 Jigs for TS parameter measurement See the ARTA Hardware and Tools Manual currently available in German only 1 for more information 3 1 4 Sp measurement For TS parameter calculation the effective membrane area Sp or effective diameter Dp must be known Because the membrane surround also participates in driver resonance a proportion of this must be factored in It is usual to include one third to one half of the speaker surround Dp inkl 0 33 bis 0 5 Sicke 3 1 5 Rp measurement The DC resistance Re of the voice coil is needed for the calculation of TS parameters If you are not sure if your multimeter measures DC resistance accurately try this trick It works even with very basic multimeters Ry Uis DE URy e Puta known resistance RV e g 8 2Ohm 0 25W in series with the driver e Connect a 1 5V battery 17 e Use the multimeter to measure the voltage across the resistor RV Upy and the voltage Ujs across the driver e Calculate the DC resistance of the voice coil Rpc as shown in the formula above Example 4 Ohm woofer e Known Ry 4 7 Ohm e Measured Upy 0 8368V Uis 0 5591 V e Calculat
30. rst pause ms 100 Mute switch off transients W 10 FFT mode pink noise excitation In FFT mode parameters for pink noise excitation can be u defined FFT size 32768 e FFT size values for FFT resolution Averaging Linear e Averaging type of averaging none linear exponential Max averages 100 e Max averages maximum number of averages e Asynchronous averaging asynchronous averaging Asynchronous averaging x on off Before measuring check the output level so as not to overdrive the input channels The two excitation signals are very different so when the signal type is changed levels should be verified All necessary settings can be accessed via the menu item Generator Setup Generator Setup Generator Type Pink FN vi Sine freq H2 1000 Output level ode Pink cut off Hz 100 Input level monitor Generator e Type of excitation PN or stepped sine e Output level 0 to 20dB e Frequency of sine excitation e Cut off frequency of pink noise signal Input Level Monitor Click on Test If the display shows red or yellow the level should be reduced Soundcard calibration The soundcard is calibrated using Calibrate Input Channels The process is very simple If using the ARTA Measuring Box move the switch SW1 of the Box to the Impedance Measurement position and SW2 to the Imp Cal position Click on CAL in the top menu bar
31. s 5 37 Mms 2 25 grams ki Rms 0 255123 kg s Cms 0 001197 m N Vas 1 72 liters Sd 31 97 cm 2 BI 4 011855 Tm ETA 0 27 Lp 2 83V 1m 87 22 dB fs Qt 190 08 5 10 20 50 100 200 500 ik 2k 5k 10k 20k Cursor 20515 50 Hz 24 79 Ohm Frequency Hz L2R 000 Figure 3 2 3a TS parameters and goodness of fit of the L2R model all optimization parameters inactive E Tr dei ii Loudspeaker Parameters PRA Fs 97 66 Hz Re 7 07 ohms dc 56 0 Qt 0 53 49 0 Qes 0 59 Qms 5 20 a Mms 1 97 grams 35 0 kini Rms 0 243228 kg s m Cms 0 001369 m N Vas 1 97 liters 21 0 Sd 31 97 cm 2 140 i BI 3 902635 Tm ETA 0 31 mi s Lp 2 83V 1m 87 55 dB 0 0 fs Qt 184 26 5 10 20 50 100 200 500 4k 2k 5k 10K 20k Cursor 20515 50 Hz 24 79 Ohm Frequency Hz L3R 111 Figure 3 2 3b TSP and goodness of fit in the model L3R all optimization parameters active As we can see the results differ slightly depending on the model selected The fs Qt ratio differs by about 3 and Vas by about 13 What does this mean in terms of speaker design Figure 3 2 3c shows AJH simulations using the parameter sets from Figure 3 2 3a Model L2R 000 black and Figure 3 2 3b model L3R 111 red The difference between the two can be seen quite clearly If it s important for the design you have to decide by yourself 20 AN SPL dB cr e fae BS BU rir k 30 50 100 ei 500 KI Frequen Hz Figure 3 2 3c Simulat
32. sation Setup measurement View magnitude Figure 2 2 2 LIMP toolbar 2 2 1 Basic settings in LIMP Before starting some basic settings have to be entered 1 Soundcard and input and output channels in Soundcard Setup 2 Measurement parameters in Measurement Setup enter Rref 3 Soundcard level in Generator Setup 4 System calibration in Calibrate Input Channels n b Under Soundcard Setup Figure 2 2 3 you can see whether the soundcard has been detected Selection of the menu item Sound Card Driver shows the card currently being seen by the software Select the driver use an ASIO driver if possible and the input and output channels Soundcard Setup Soundcard driver BEHRINGER USB AUDIO Control Panel Input channels 1 2 Wave Format Output channels i 2 16 bit gt Figure 2 2 3 Soundcard setup in LIMP Depending on the soundcard drivers available Windows can open either the Windows mixer or the ASIO Control Panel will open For LIMP no further settings are required here n b In Measurement Setup as a rule you should only need to enter the value of the reference resistor All other settings can be left as their defaults There are three setup areas for LIMP e Measurement Config left e Stepped Sine Mode middle e FFT Mode pink noise excitation right Measurement Setup Measurement config Stepped sine mode FFT mode pink noise excitation Reference channe
33. to open the Calibrate Input Channels dialogue and calibrate the system 11 Calibrate Input Channels Generate Seg length 32768 Sampling rate 48000 Output volume dB 12dB Generate Input Level Monitor L Calibrate Status Connect left and right input Not calibrated channel to signal generator output 1 Number of averages EN i Figure 2 2 5 Measurement Setup The Generate button generates the signal so that the levels can be Status checked The Calibrate button then calibrates the two input channels relative to each other When calibration is complete the Calibrated for status window shows Not Calibrated or Calibrated for with the TE s Seg length 32768 sequence length sampling frequency and measured channel difference shown Fs 44100 Hz Note The calibration is valid only for the selected settings If you Channel diff 0 11dB change the sampling rate or the length of the FFT sequence you must recalibrate Uncalibrate The channel difference is compensated for by the software If the measured difference is gt 2dB you will get an error message Figure 2 2 6 d x Something is wrong channel difference larger than 2dB Check whether left and right input channels are connected on generator output Check whether input level controls are at same position Check whether input probes have same gain Check whether cables and connectors are OK Figure 2
34. to prevent errors that might result in damage to the soundcard 2 1 2 Impedance measurement using a power amplifier Because the headphone output has only a limited current capability it can be replaced by a power amplifier although the higher power levels that are needed for acoustic measurements are not necessarily required for impedance measurements Under this arrangement Figure 2 1 6 the reference resistor must have a lower impedance The recommendation from the original LIMP user manual is Rref 27 ohms lett out Power amplifier Vollage probe Loudspeaker left in Soundcard Figure 2 1 6 Impedance measurement using a power amplifier When using a power amplifier be aware that the output voltage can be significantly higher than the output of the headphone jack These higher voltages can overload or even damage the soundcard The use of a voltage divider voltage probe is therefore recommended For example Hi Z Instrumenteneingang Type unsymmetrisch 6 3mm Klinke Max Eingangspegel 4 5dBV max THD N 0 003 A gewichtet Impedanz 500 kOhm Figure 2 1 7 Soundcard line in specifications ESI UGM96 Instrumenteneingang input specs unsymmetrische 6 3mm Klinke unbalanced 6 3mm Jack Max Eingangspegel max input level A gewichtet A weighted Impedanz impedance e Maximum soundcard input voltage Vin max 4 5dBV 1 0 104 4 5 20 1 679V RMS e Soundcard input impedance Zw
Download Pdf Manuals
Related Search