Failure analysis of Selectronic dynamo regulator 1 Failed regulator
Contents
1. The AC is expressed in RMS root mean square which is the computation of this current this means the equivalent thermal effect of a DC current For a pure sinus current the peak peak voltage is expressed as 2 x V2 times the DC value so a 6V RMS makes 17V p p 4 4 1 Car type alternator Exists in some big bikes BMW Goldwing Guzzi The inductor is rotating rotor powered by brushes with lower current than the output this acts like an electromagnetic gain The shape of the polar pieces determines the voltage form sinus trapeze The stator gives the main current 40 to 80Amps is fixed The rectification is made in the alternator with silicon diodes pressed in the radiators in the air flux for cooling The circuits are three phases with full wave rectification this makes a minimum voltage ripple Contrary to car dynamo there is no need to limit the current the designed magnetic saturation limits the current T BE BE BE EE BE E MF E y E 1000 2000 3000 4000 45000 6000 Foo 868000 irimn COGO Zibuth27 2012 Creative Commons CC by nc sa page 16 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 4 4 2 Bike Alternator Generally limited to 200W made with strong magnets placed on the crankshaft Suzuki GSX400 1981 rotor rotating around stator stator fixed on the cover inside the rotor The rectification is made with an external diode bridge it was selenium diodes for2. BSA 350 amp 250 from the fifties back my teen ager BMW R1100 1997 180 Megameters The charge circuit is made of a battery a rectifier if fitted with an alternator and a source dynamo or alternator We are speaking of british bikes therefore this reversed order 4 1 the battery Lead acid for the bikes almost no exception Made of an assembly of cells called 2V because the voltage is around 2V per cell 3 cells in series make a 6volt battery 6 cells make a 12V battery The 2V per cell varies around 2V depending on the acid concentration the charge status the current the TEMPERATURE the technology starter battery buffer battery gel battery and the metals in the alloy pure lead with calcium tin selenium silver The battery of the vintage british bikes is not intended for cranking motor not able to give a current of several times his capacity The capacity is given in Ah ampere hour and is theoretically the current for going from full charge to empty in an hour It appeared soon that the apparent capacity depends on the discharge current we use now a discharge in 20 hours The measure unit in Ah comes from electrochemistry domain related to the amount of coulombs transferred supposed the current is constant which never happens in reality an other unit issued from electricity domain is the watt hour Wh it begins to be stamped on some camera batteries but the power is not more constant The main user
3. with isolation film thermal conductor close to the TIP41 dissipation 2 to 3W QOO Zibuth27 2012 Creative Commons CC by nc sa page 9 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 It is better to use the MBR1545 available at the same Selectronic priced 1 3 This diode is made of two 15A diodes in the same chip same voltage drop same temp coefficient we have then a good operation margin Radiator is MANDATORY use the main radiator care should be taken on the passage of wires The best would be to regulate the true voltage of battery first by moving the sense point of R4 to the output of the diode the black wire but compute first all effcts of this modification Second possibility is to have a sense wire connected to the battery itself then it will be a true compensated regulator those modification will be studied in an other paper As it is the regulator requests to be set 0 5V above battery voltage but the voltage drop to be compensated depends on the current so 7 7V would make an acceptable trade off The test was conducted with a 9 Amps current for a 8 5 Amps max current but the temperature rise is then 191 C above ambient still destructive COGO Zibuth27 2012 Creative Commons CC by nc sa page 10 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 4 Bike charge circuit Just a reminder of basics already supposed well known The bikes from my tests
4. 15 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 4 3 2 Trick for isolation check I often heard this wise guy trick for isolation check put a few watts lamp in series and use the 220V mains NEVER do it Aside the lethal risks specially with a quick and dirty set up this can create a failure on device with no 6V problem or increase an existing leakage Remember when you speak about 220V the reality is a peak peak 622V the destructive spark is sensitive to peak voltage a lamp is sensitive to the average current RMS to be precise In addition the current is variable cold starting current is 15 times higher than steady current a 20W bulb starts at 1 35Amp and runs with 90mA the filament temperature is 3000 C and the temperature coefficient is 0 5 C Sparking risk with a 600V peak if two unsinsulated conductors are separated by a 0 2mm gap there is a destructive spark which never would have caused any trouble on 6V The 220V 20W bulb never replaces a specialized device like high voltage megohm meter with limited current 4 4 The alternator The alternator generates alternating current obviously The battery needs direct current Alternating current is measured in RMS value the equivalent DC current for thermal effect like in a lighting bulb The alternator can be used for the lighting system but the output current varies as the rotation revs up or down 6V at 1000rpm makes 30V at 5000rpm
5. 2 charge Tashima 6N4 C 2 9 00 8 50 8 00 7 50 tension 7 00 6 50 6 00 Olts O 20 40 60 80 100 120 140 minutes OPERATION AND CONSTRUCTION DOE HDBK 1084 95 Lead Acid Storage Batteries COS Zibuth27 2012 Creative Commons CC by nc sa Discharge _ Charge Volts per cell specitie gravity Ampere hours returned Ampere hours discharged Specific gravity Full charge S a ram U Q Ww ampere hour Normal discharge Figure 8 Changes in voltage and specific gravity during charge and discharge page 12 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 4 1 2 Temperature Coefficient Flcet AMBIENT mmareraTure gane apo oom 2223x 2 22 50 2a o 202 The temperature coefficient depends on acid concentration therefore from the charge status PowerSonic publishes this chart for his 12V 4 5Ah sealed lead acid battery The corresponding thermal coefficient 0 to 40 C is 0 006V C Temperature C 4 1 3 battery data sheet Figure 23 Temperature Compensated Charge Voltage NP Series Valve Regulated Lead Acid Battery NP4 12 SPECIFICATIONS Another curve corresponding to a unique charge status The corresponding thermal coefficient linearized between 0 and 40 C is 0 004 V C Data Sheet l r I masa E Meme rr ara Yuasa data corr Masstypea Cd E pR eee ERMINAL TYPE
6. 65W soldered bottom side of PCB e one diode 1N4007 1A 1000V field back surge absorption Usually this diode withstand the nominal current of the coil protected I expect then a 1A max current with some possibles peaks The Schottky diode STPS15 allows 15A on infinite radiator and able to work at 175 C but at zero current The assembly without radiator allows the current no more than a few tens of seconds computation to be done according to real parameters of the dynamo Active components from big brands OK The TIP41 transistor is badly assembled and destroyed himself Hyperstatic assembly not any chance of flexibility NOT CONFORM bending radius smaller than two times the lead thickness NOT CONFORM The transistor is fixed on the radiator ans his leads are soldered on the PCB As he is soldered to the bottom side of the PCB he is placed approximately on place before the assembly of the radiator The stiffness of the leads vibrations and temperature increased the cantilever effect and splitted the epoxy from the copper base plate The side of the copper plate are not visible in a healthy transistor COSO Zibuth27 2012 Creative Commons CC by nc sa page 4 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 Broken transistor In fact the broken transistor could have been seen by very careful inspection and by somebody aware of this very uncommon failure the abnormal aspect of the transistor de
7. charging The alternator requests a 12V system as the voltage drop of 1 2V is too important for 6V On the same time was the introduction of the no maintenance batteries CONO Zibuth27 2012 Creative Commons CC by nc sa page 17 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 4 5 2 Bike regulator 4 5 2 1 Rectifier Replaces one function of the brushes was copper copper oxide at the beginning then selenium The reliability was very poor with addition of some leakage current draining the battery Since the seventies the rectifier is made of silicon The voltage on half wave provides a high ripple the 3 phase full wave rectification provides almost no ripple 4 5 2 2 To regulate or nor to regulate Lucas prince of Darkness William Sh 1599 On the beginning regulation was made by switching the coils between day or night because on these times it not mandatory to ride with headlights on during the day Today although it exists in France a law allowing not to ride headlight on for bikes older than 1962 the cops arrest you frequently making believe they don t know this law real stupidity or tactic from law enforcement forces but look carefully on your tyres in any case you better have a copy of this law in your wallet Because of basic electromagnetic laws voltage depends primarily on the rotation speed the only possible choice was to have full charge only at high rpm This is a possibl
8. in the vintage bikes is the lighting circuit where the current varies highly a cold lamp drains 15 times more current than a hot lamp during the time of heating up and lamp current and lamp duration is very sensitive to the voltage The only constant drain constant resistance is the coil battery ignition system Kettering ignition in USA but ther are lots of magnetos out there A starter battery shows two figures capacity in Ah and the max current for short time eg 60Ah 610A Most values are given at temperature 20 or 25 C A few manufacturers publishes the temperature coefficient as a linear coefficient range of 0 003V C but reality is not that linear straight line is only an approximation in a temperature range CONO Zibuth27 2012 Creative Commons CC by nc sa page 11 19 april 2012 4 1 1 Failure analysis of Selectronic dynamo regulator Discharge curve Measured on new bike battery 6V 4Ah minutes Volts Amp Alh W h Prix A h Prix W h 6 50 6 00 5 50 5 00 4 50 4 00 3 50 3 00 2 50 volts batterie plomb 6V Tashima 6N4 R interne 0 100 6 04 1 285 0 000 0 000 0 000 6 04 1 285 0 214 1 294 36 884 6 04 1 285 0 428 2 587 18 442 5 81 1 235 0 634 3 783 12 456 6 16 1 310 0 853 5 127 9 266 6 04 1 285 1 067 6 421 7 406 5 81 1 235 1 273 7 616 6 208 5 86 1 248 1 481 8 835 5 336 5 86 1 248 1 689 10 055 4 679 5 86 1 248 1 896 11 274 4 166 5 81 1 235 2 102 12 469 3 758 5 81 1 235 2 308 13 664 3 423 5 81 1 235
9. the first vintage bikes now silicium diodes integrated with the regulator since the seventies There were no regulation on the beginning of alternator use simply switching of one or several coils when using lighting The choice was to have the correct voltage on hi rpm only at low rpm lighting was poor Lucas prince of Darkness The voltage is directly dependent of rpm if 6V at 1000rpm it goes to 30V at 5000rpm and the lamps are extremely sensitive to overvoltage inversely proportional to V to the sixteenth power an overvoltage of 16 reduces the lifespan to 10 of normal VOLTAGE IMPLICATIONS The graph shows the effect of changes in the supply voltage on GLS filament and tungsten halogen lamps For example a supply voltage 5 higher than nominal reduces lamp chart from the british Lighting Industries Federation ite by 50 5 E Percentage Watts Lumens Amps 4 5 alternator regulator 4 5 1 Car type Integrated in the alternator generally in a module close to the diodes The diodes request air cooling provided by the air flowing through the alternator This air is already heated up by the coils and by the diodes before reaching the regulator therefore the temperature is not related to the battery temperature which was the case with dynamos when the regulator was laced near the battery so there is no way to implement temperature compensation The voltage is generally regulated to 14 4V which is not a 100
10. we are in the sulfatation zone I sacrificed a part of the lifespan here but at a price of less than 8 I can afford it The manufacturer choosed to stop discharge at that value we can see that we can extract some more energy I went up to 5 7Ah Remember the more deeply you discharge the less the battery withstands a number of charge cycles You better choose the biggest capacity which fits to the compartment and electrodes position it lasts longer The vintage BSA 350 has a battery of 6V 12Ah the dynamo gives 8 5A at 7 5V the maximum charge current is C 1 6 4 2 dynamo regulator Many vintage bikes from the fifties use the Lucas MCR2 regulator adapted to the E3L dynamo diameter 3 inches 7 62cm this is a two coils regulator well manufactured for his time It is made of e one voltage regulation coil e one coil to avoid the current going INTO the dynamo cutout coil The voltage coil is adjustable on a few parameters e the gap of the contacts e the stiffness of the spring As soon the desired voltage is reached the contact opens making a vibrating contact around 100Hz The adjustable spring visible on the picture the vertical parallel lines just on the left of the coil is in fact bimetallic like in many thermostats and acts as the thermal compensation wanted by Sir Lucas Extracted from service manual 10 C 7 9V min7 7 max 8 1 20 C 7 8V min7 6 max 8 0 30 C 7 7V min7 5 max 7 9 40 C 7 6V min7 4 max 7 8 make
11. 2 514 14 859 3 142 5 75 1 223 2 718 16 031 2 907 5 75 1 223 2 922 17 202 2 704 5 75 1 223 3 125 18 373 2 528 5 69 1 210 3 327 19 520 2 374 5 63 1 198 3 527 20 644 2 240 5 69 1 210 3 728 21 791 2 119 5 51 1 173 3 924 22 869 2 013 5 39 1 148 4 311 24 978 1 833 5 10 1 085 4 492 25 901 1 759 4 34 0 923 4 646 26 569 1 701 3 99 0 848 4 787 27 133 1 650 3 69 0 786 4 918 27 617 1 606 3 52 0 749 5 043 28 056 1 567 3 11 0 661 5 153 28 399 1 533 3 05 0 649 5 261 28 728 1 502 2 87 0 611 5 363 29 021 1 473 2 70 0 574 5 459 29 279 1 447 2 52 0 537 5 548 29 505 1 424 2 29 0 487 5 629 29 690 1 403 2 17 0 462 5 706 29 857 1 384 d charge Tashima 6N4 C 3 Olts m fem 0 50 100 150 200 minutes typical charge discharge curve on a bike pilot and regulator don t use normally a gravity meter while riding so I m interested in the volts per cell values We may deduce from this that the floating value correspond approx to 80 charge for 6V battery multiply by 3 for 12V battery multiply by 6 ddp utiliser avant volts date 02 04 12 Ratio 0 0586 vide 6 51 R gt 4 7000 Prix 7 9 tension Volts 0 20 40 60 80 100 120 140 160 180 0 291 6 000 0 282 5 000 4 000 3 000 N N O capa 0 268 2 000 0 265 1 000 0 000 minutes capacit tension voltage aa capacit capacity 5 000 d charge discharge 4 000 3 000 ain 2 000 1 000 0 000 O 20 40 60 80 100120140160180 minutes version 3
12. External link is made by wires 3mm dia 16AWG copper 1 29mm dia section 1 31mm2 specified 105 C Good quality wire registered UL Underwriter s Labs possibly made of PER reticulated polyethylen withstands better than PVC thus allowing higher working temperature Copper wire with metallic finish bright grey poor solderability The wires are fed through the radiator in a 4mm dia hole NO bonding NOT CONFORM All wires are wounded by the radiator not deburred on one side fortunately on this sample on inner side where the wire moves less Because the wires are NOT going through the PCB the angle at the radiator varies and the contact with the radiator may be hard Two wires are oversized they will only feed an amp or less The main current flows in a section of 1 31mm_2 slightly overloaded not a big concern considered his short lenght Because the hole in PCB is too small the wires are butt soldered and can t get through RECOMMENDATION bond the wires to the PCB and to the radiator Ideally the wire should have a stress relaxing loop Anyway the holes HAVE to be big enough or redrilled e Correct solder except big wires e leaded solder illegal for sale since 2006 The solder flux residues are cleaned on the PCB except for added elements probably added in a second phase of manufacturing big wires and TO220 transistors This may induce long term corrosion problem NOT CONFORM but probably not so im
13. Failure analysis of Selectronic dynamo regulator version 3 2 1 Failed regulator Selectronic N 1 for Dynamo 6V GOW LUCAS cccccccccceeessssssesssssnnnaseeeeeeeeeeeeeaeaeaeeseees 2 LECCEN NION er r E A A T E E E E N A E TAS 2 Desc PC O a E E E E E E T ome oeam 2 CSCOP Oa E E EEEE E EAE EE EEN EEEN E EE EEEE E EER 2 LPE CLUE WO AEC a E EA 2 LoP O eE E EEE EE E E E E E E E 2 Deis VV OS VT SMOLIN US EEEE T EET S E E E T S E A E S A E T 3 e Ee Mla OMVC DiE AARAA A E E E A E T EE E EE EEE E E E neeans 4 L3sEzplanation ofthe SUL CSSCS 04 cosscacqosssdossavievavacecatmsstenenocgensdossavievevacecs ee ESE AE EEE SEAE EE 5 1 3 6 Mounting of other components sssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssssseeeessssses 5 EELEE al Tie AS EE e E E E E OE N N EAEE T 6 1 5 Conclusions on the assembly of the reQulator cc sssssssssesssccceceeseeceeeeceeeeeeeeeeeeeeeeceeeeeeeeeeeeeseeeeeseeeeees 6 2R SOE VEO IN A E EE EE eerste E ies EE E E E nama E east 7 ZA Measuring values OIC O MIP OM CIS sss cnnccasonaunsn pumenasstecnantss aust sridayanniapubenusies ential awstea cide vanisedusnasaueiecaMleadiostaCees 7 De SEU Y AU OMS E E E E AE E E OE E EE O E E ETETE 7 BoD WNBA CO COUN a ategs navecetneueceanantmctontsaananessnccscatntanccanantuclentiganavedanocecetntanceacantmsantiaanavsanteceae E N 8 28 VEC IC Al 1651 e E E E EN EE E N E E E EN 8 FECONDO O erer er E EEE E EEE 9 3L Recon ticom of schenmiliOS sess
14. OS Zibuth27 2012 Creative Commons CC by nc sa lt m EJS amp 1O JE 318 a ag Rie gjs i 5 3 3 E i installation Can be installed and operated in any onentaton except permanently inverted Handles Batienes must not be suspended by their handles where fitted Vent valves o k D 2 8 VRLA Batienes release hydrogen gas which can form explosive mixtures in air Do not place inside a sealed YUASAS VRLA batienes must be recycled at the end of ile in accordance with loca and national laws and regulations ALL DATA IS SUBJECT TO CHANGE WITHOUT NOTICE Issue No V 2 Issue Date Mar 2011 YUASA page 13 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 The tested Tashima battery has the voltage at 23 C going from 6 04 to 5 51V when the battery always connected to the same load resistor 4 7 ohms has his charge status goeing from full charged to full discharged The emf should be normally measured any load charging circuit disconnected since a certain time say half an hour after disconnecting the load This is a boring measurement I simply computed a value of internal resistance at begin of measurements from the difference loaded unloaded Deducted from the internal resistance I extracted an instant emf no settling time of 30 mins in the curve discharge at C 3 the red curve fem emf in french The discharge once the 4Ah drained the voltage drops rapidly AND
15. able QOO Zibuth27 2012 Creative Commons CC by nc sa page 6 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 2 Regulator N 2 Received in working condition Same model than the previous one 2 1 Measuring values of components Same resistors on same place thus no adjustment with different Zener diode And we have then a different regulation voltage regulation voltage 8 77V regulator 01 8 26V Zener voltage 5 79V regulator 01 5 48V 2 2 Observations e The leads of the TIP41 have a good bending radius OK e the silicon isolation film is wrong placed under the TIP41 bad thermal transfer and electrical isolation NOK e Only one wire goes through the PCB the yellow one Bad solder wetting NOK e the PCB is not parallel to the radiator due to the stress given by the wires and the TIP41 this one is OK PCB amp radiator are not parallel because of the stress on a wire bending radius of TIP41 almost correct isolating film wrong not This wire is not in his hole then misplaced totally under the transistor Misplacement of film unexplained ratches on PCB bad solder no wetting of wire amp not passing through the PCB ere Zibuth27 2012 Creative Commons CC by nc sa page 7 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 wound on all wires 2 2 1 Back to correct Simply back to standards with the recommendations link wire f
16. and armature together shortly to a battery with correct polarity Take care that the dynamo works then as an electric motor For the wise guys don t expect to have invented a cranking device the usable torque is way from enough The Lucas regulator has two voltage coils connected to the dynamo output making 13 ohms resistance which drains 0 57 Amps an electronic regulator drains only a few milliamps 4 3 The dynamo 4 3 1 Lucas Dynamo Made of a fixed inductor Field coil 2 8 ohms and a rotor called armature with brushes 0 50hms rotor brushes for rectifying and collecting current It is an electric machine able to work as a motor if current flows TO the dynamo The E3 series have a diameter of 3 inches 76 2mm dynamo Lucas E3L the dynamo is not designed with axial symetry only one field coil rotor brush side I finally got the electrical parameters of the E3L dynamo J Gardner revised by Hancock Field coil 2 8 ohms rotor resistance armature 0 55 ohms including brushes max output voltage 7 85 V at 8 5 Amps This is intended for charging the battery at 7V through a regulator resistance of 0 1 ohms voltage drop 0 85V This not a full charge only between float and cycle charge around 70 of full charge but is at full load of 60W It is not possible to measure correctly these resistance values with a normal multimeter even an electronic one grace Zibuth27 2012 Creative Commons CC by nc sa page
17. ations of battery manufacturers which reduce in fact their lifespan The electromagnetic regulators show a rapid wear out and are replaced by electronics One electronic regulator made by Selectronic in France uses good quality components but the design and one poor manufacturing phase give a poor reliability numerous claims on the Web The correction of some design errors is though easy and cheap and it is recommended to adjust their voltage Zibuth27 april 2012 COGO Zibuth27 2012 Creative Commons CC by nc sa page 19 19 april 2012
18. could think that the field coil will drain no more than 6W provided that the electromagnetic gain is 10 The dynamo will provide 10A if it was designed for 6V operation and 8A if designed for 7 5V operation I have not yet a dynamo for test at the begin of this paper and I work on estimates current voltages I take 10A before more precise informations 1 3 Composition 1 3 1 Printed circuit board Single side goldened FR4 epoxy with solder mask ref 3801 D L Selectronic well manufactured Some insulation distances could have been slightly bigger considered that there is no real estate problem anyway the solder mask protects Gol finish is OK because the quantity of of solder is big enough to dilute to an acceptable level the gold tin intermetallics A double side PCB would ensure better holding of feed through components The circuit was cleaned from flux residues after soldering of components then a second phase of manufacturing connecting wires was done without cleaning No general protection with varnish 1 3 2 Pinout Red battery ground Black battery White field Yellow dynamo out A diode links wires yellow and black avoiding current from battery to the dynamo which uses the dynamo as a motor All charge current flows through it CIOSO Zibuth27 2012 Creative Commons CC by nc sa page 2 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 1 3 3 Wires wire mounting
19. e cause for the nickname Lucas prince of Darkness As electronics was not well developed at these times to affordable price Lucas as many others used at first hand the pilot s hand switch later on used Zener diodes zerro current at low voltage in silicon leakage current is in the microamps range at the Zener voltage the current becomes suddenly limited only by external resistors the zener limits therefor the voltage and dissipates the excess in heat big radiator needed Lucas alternators are monophase the magnets travels on the same time in front of the corresponding polar pieces BATTER T MUC e E J MEAG IGHT COAT BOL im a Fig 23A 12 volt circuit for coil ignition ine Fig 11 Charging circuit in the pilot light position SS p First type a third of alternator is used for charging the battery Second type a Zener diodes limits voltage and heats up one Two thirds are used for lighting no regulation at all third of coils is added for lighting In the case of magneto ignition one third is used for charge tho thirds added for lighting Full wave rectifier Third type two Zener diodes must be paired for equalizing heating else on ee ee heats up a lot and destroys himself the second following soon 12 75 wl oach roms cen on one wove steovsy hus It is a half wave rectifier and half of the power is usable with high ae ripple of voltage u The advantage of the
20. e it was available in my drawer Mounted obviously with short wires so the PCB and the transistor on the radiator are now mechanically decoupled RECOMMENDATION dismount the TIP41 place linking wire between transistor and PCB of course after checking the good health of the transistor and shortening the leads to 5mm because the solder has NOT to go up to the body the copper diameter could be reduced to 1mm considering the short lenght of 2cm max 1 5 Conclusions on the assembly of the regulator Not knowing the schematics not yet reconstructed but knowing how a classic regulator should be I ca say that the components are from good quality some are overspecified no need to bother to compute exactly the necessary power or precision not a problem only somewhat pricier except schottky diode Most of the PCB is well made soldered and cleaned But the mechanical design is BAD some points not conform to IPC 610 standard class toys The mounting of the wires and specifically of the transistor TIP41 is BAD This failure is the consequence of truth In my humble opinion it is necessary to e change the assembly of the TIP41 and add wires to link the transistor to the PCB change the transistor if any suspicion of stress e redrill the wire holes in the PCB and resolder them correctly e glue the transit of wire to the PCB This analysis was sent to Selectronic on March 2012 in hope I had a dream the future deliveries will be reli
21. ep and bright black color for the normal part of the transistor grey and finely grainy aspect of the broken part 1 3 5 Explanation of the stresses The stress is in that direction effort because the vertical part in this drawing of the lead was slightly too long The solder spot in this position gives a strong application point to the force And the lengths are just short enough to avoid bending It is possible than if the lead is too short when the screw is torqued the copper plate can be bended restricting the thermal transfer or the lift of the epoxy may occur on the other side Film isolant isolating sheet silicon semelle plate of transistors soulevement lift up effort stress In this assembly the destructing effects of this design are dependent of the torque applied to the silicon sheet because of the elasticity of the silicon sheet and of the nylon screw insulator In addition to that the solder spots are located i the worst place exactly in the axis of the two PCB fixation screws Because it is now obvious that there is no electric relation with the remaining of the circuit and the schematics seems simple I simply replaced this transistor only It works 1 3 6 Mounting of other components Diode STPS15 the bending of the leads do not start at a minimum distance from the body NOT CONFORM risk of moisture penetration and internal corrosion In addition the mounting without radiator does not all
22. in a I A R E EA EA 9 PARCE ON ES O e a E e E E AEE A E EE EEE EE E E E R E 9 APEE o o EP Mo e E E O E E E ee eee 9 ADE e Wa C CUI e E A E EE E E E E T E E 11 A HLS DaO e a A TA 11 ALLDATA 12 4 1 2 ent ACEI eG OC MPIC TO IL srsoeideiriininarsinoie onnie ee oiae iiaae eiaa 13 r E EE e CL E EE AEE E AEE T A 13 AEA EE DEEI ET Ea E E E A E E E E e OEE E E E E E E 14 Aa TNS A AN AIM Oli ae E EE E E N E E edanetecnscrentendes 15 Aea daS D TO E E S E E E AEE E E E A 15 Ad VIC HOT 1G EA a E E E T E AT T A E E E TE E 16 AA WING a O r E E E E E E E E E A 16 AAC e an OE e A AE E E 16 AAL DOADO e a a E E E A EAN EERE AAE EN AIE A EE EA E AIA AET 17 ARSE NaCO FFE Ol EE eT E EE AE P EE A E A TA A 17 RSE OE A a E IEE A PE E E E E VAI O IEI E VOR IA TSE E A EI ener onceeees 17 nA A EO A S to ETA A E E EE 18 ALERO E E E EE O E E E E E 18 4 5 2 2 To regulate or nor to regulate Lucas prince of Darkness William Sh 1599 ee 18 Deg UMC A a E E E E E E EE E A E A 19 SCO O eaea e E E E A E E E E E A E 19 COGO Zibuth27 2012 Creative Commons CC by nc sa page 1 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 1 Failed regulator Selectronic N 1 for Dynamo 6V 60W LUCAS I was given this failed regulator for repair and comments 1 1 General view Extarnal view of radiator PCB view component side PCB view bottom side 1 2 Function It is a dynamo regulator for 60W british dynamo positive to earth one
23. ow the maximum current of 15A there are specified by the manufacturer only with correct radiator Jevevcere Zibuth27 2012 Creative Commons CC by nc sa page 5 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 diode BD136 One lead is bended too close to the body we can see a black line showing separation body lead It is preferable to replace it An example of more tolerant mounting of the transistor This one was not a new one that explains the irregularities This transistor was only used for checking the hypothesis of single failure 1 4 Electrical measurements Regulated voltage measured on lab power supply not on dynamo circuit 8 26V It is consistent with values I already observed on some Zener regulators for alternators I don t say it is correct It seems to me that is a boiling voltage for the battery and is probably not good for long term For example in a 12V car circuit the voltage is limited to 14 4V 7 2V on 6V circuit For reference purpose the reference diode voltage is 5 48V The regulator shows no linear behavior maybe for a few millivolts only This proves the insensibility to individual characteristics of the transistors allowing the replacement by any other in the power range For replacing the dead TIP41 not knowing the exact need I just want a type withstanding at least the 6A of the TIP41 I finally decided for a MJE3055 NPN 100V 15A 90W just becaus
24. portant here because of the presence of solder mask and no high impedance circuitry No wire goes rhrough the PCB because the copper diameter is bigger than the hole So the wires are butt soldered A traction stress is not welcome here NOT CONFORM The red wire is very bad soldered no wetting of the copper and an excess of solder flowed through the PCB hole a touch of soldering iron showed that it not a wire going through the PCB NOT CONFORM CIOSO Zibuth27 2012 Creative Commons CC by nc sa page 3 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 flowing of solder not even correct on other side Solder to be made again after drilling a correct hole and total insertion of wire The transistor screwed on the radiator has a too sharp bending radius and cracks appeared in the electrode NOT CONFORM The electrodes which survived to assembly may break eventually under vibrations remember this is made for british vintage bikes and they have a reputation of high level of vibrations 1 3 4 Components e dated 3 2003 or 1993 e Resistors ALL in 1 tolerance OK but not necessary for all to be seen in schematics analysis e STPS15 diode Schottky ST Semiconductor 45V 15A 175 C NO radiator e BD136 power transistor Philips PNP 45V 1 5A 10W 150 C no radiator e two C547 BC547 transistor NPN 60V 100mA 0 5W Philips e power transistor on radiator TIP41C Fairchild NPN 60V 6A
25. re analysis of Selectronic dynamo regulator version 3 2 adding a sense wire The other coil cut off opposes the magnetic field created by the main current to the field created by a voltage coil as soon as the current wants to go from the battery to the dynamo the current is stopped This coil is replaced by a diode mandatory a Schottky diode for lowering the voltage drop In the car world there is a third coil for current regulation There is a spike voltage limitation on the vibrating contact made with a pellet resistor the black disc with green paint stripe firmly pressed against the armature for heat dissipation value 50 ohms This contact wears out even covered with platinum or tungsten specifically because of the functional vibrations 100 openings per second making a million operation in less than 3 hours Obviously there is no longer any spare part available An electromagnetic regulator when the voltage is not reached shortens rotor and stator called here field and armature the residual magnetic field creates a small voltage in rotating armature which in turn creates a higher voltage the dynamo after this autostart loop reaches finally the regulation voltage It happens that after some position in the earth magnetic field or exposure to leakage of alternative magnetic fields motors or transformers the residual field in dynamo becomes too low for the autostart feature it is necessary to connect temporary field
26. rential amp T1 T2 The output voltage is then shifted down from 0 2 to 0 5V depending on current we are far away from the 3 requested by Yuasa In the case the transition voltage is too high place additional resistor parallel to R3 in the case voltage is too low place resistor on R4 3 2 1 D1 diode STPS15 As expected 81 3 4 components D1 diode heats up a lot from 21 C to 130 C in a minute amp half no possibility to take into account the time constant so the temp was significantly higher At this temp direct voltage drop is 0 5V at 9A dissipation is 4 5W you add 7 2 C for junction temp making 137 C The PCB will degrade at these temps The data sheet shows that if the STPS15 is placed on PCB but on the copper side with 1 3 cm of copper the brutal change of slope in the curve the thermal resistance is in the range of 45 W then 193 C above ambiant Largely over the max 218 C as 175 C is allowed but 175 at ZERO CURRENT Fig 10 Thermal resistance junction to ambient versus copper surface under tab epoxy printed circuit board FR4 Cu 35um D PAK 80 70 Selectronic made a major design failure in the design of the 60 s 50 regulator which may only be used at a fraction of the max wat et EE E S E E EE a E dynamo current 30 20 10 o 0 5 10 15 20 25 30 35 40 Rth j a C W FE sew em EEEE This diode NEEDS a radiator In fact it can be mounted on the main radiator
27. rom TIP41 to PCB output wires correctly soldered in a bigger hole 3 Regulator back to standards slight melting of wire insulator because the the difficult wetting of the wire Transistor linked to PCB via a wire The output wire is glued to the radiator 2 3 Electrical tests Obviously no change in in regulated voltage Test under load 9Amps as the dynamo should give 10A Negative margin for a 10A system the diode should withstand 15A As expected on the first sample 1 3 4 the diode D1 blew up after one minute and half The diode is replaced by a MBR1545 available from the same provider Selectronic price 1 3 As the diodes are from the same chip no problem of parallelization 30A total The diode is placed on the radiator and the temperature rises now to 35 40 C at 9A NOW the regulator becomes reliable CONO Zibuth27 2012 Creative Commons CC by nc sa page 8 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 3 Electronics modification For qualified people only at your own risks 3 1 Reconstruction of schematics onic Dynamo 6V regulator REVISION DRAWN BY Select RMZ zibuth27 gru t 3 2 Remarks on design Very classic schematics of regulator differential amplifier power amplification open loop very high gain The dynamo voltage is controlled by compare the reference voltage zener diode to a part bridge R3 R4 of the output in an diffe
28. s a thermal coefficient of 0 01V C which means the temperature is somewhat compensated The voltage regulation coil has additional 4 wounds of thick wire diameter 2mm the main dynamo current flowing through This compensates the voltage drop in a typical cabling from dynamo to battery compensated regulator Good manufacturing quality The Selectronic regulator and probably other electronics regulators do not have this compensation feature which needs a current sensor and a 10 Amps sensor is very visible the use of a diode as sensor is not an easy task A good solution is a sense wire directly connected to the A output from the regulator or better to the battery terminal The Selectronic regulator miss two important features wiring resistance and temperature compensations Regulation voltage Selectronic 8 77 or 8 26 Sir Lucas 7 9Vtyp 10 C and 7 6Vtyp 40 C One cannot claim that Selectronic has a higher voltage in order to compensate the voltage drop in the circuitry as it is always too high low or high current Lucas regulator has a anti ohmic behaviour his output voltage is higher on high current so considering the average cabling circuitry the battery voltage remains constant at same temperature of course Ideally the schematics of the Selectronic regulator has to be modified by adding temperature compensation and COSO Zibuth27 2012 Creative Commons CC by nc sa page 14 19 april 2012 Failu
29. system is the halving of the load split 60 50 between each ZENER as compared with the previous full wave system Rectifier Doubte Diode Model 2DV 406 Lucas Part No 49181 COGO Zibuth27 2012 Creative Commons CC by nc sa page 18 19 april 2012 Failure analysis of Selectronic dynamo regulator version 3 2 4 5 2 3 The Japanese All Japanese use tri phase alternator and the rectifier is in the same module All Japanese since the seventies use the same regulation scheme with thyristors The thyristor is a memory power switch once triggered with a small grid current it remains ON up to next zeroing of voltage or current Thats perfect as the alternator produces alternating current which regularly switches OFF the thyristors The thyristor is obviously not mounted on the battery side once triggered it will only switch off when the battery is emptied to zero or when it disappears in a puff of smoke stinky The thyristor is triggered by a zener diode as soon as the voltage becomes too high and switches off next voltage inversion The alternator saturates and has no damage from this short circuit The schematics is the same for all Japanese bike manufacturers only the amount of thyristors may vary one or three three is the better standard schematics for Japanese rectifier regulator RECTIFIER REDRESSEUR RECULATOR REGULATEUR 5 Conclusions The regulators from British vintage bikes do not respect the specific
30. te sex h C oait tO Bet Tis ecg Komi other manufacturers I tried Exide Bosch I found OPERATING TEMPERATURE RANGE only PowerSonic to have such usable docs floating voltage is the voltage with no current in or out of the battery CASE MATERIAL I use this document as standard value for bike b CHARGE VOLTAGE ener Float chergevotage ate zr tate ve The values can be transposed to other members of th Fat Charge vonage emperane conectan ane NP family hermetic battery valve protected gas Crete reoat eae a a recombination VRLA The cyclic charge corresponds to bike use charges discharge and requests pr cision of 3 soit a E E E E 0 072V l ment on 6V battery makes 0 21V 7 47Vmax et a A E 04Vmi 3RD PARTY CERTIACATIONS BY 7 04Vmin 1809001 Quality Management Systems Ze the Selectronic regulator with measured voltage of ISO 14001 Environmental Management Systems EN 18001 OHSAS Management Systems A 8 26 and 8 77V is off specs Yuasa even with Schottky UNDERWAITERS LABORATORIES Inc bi diode between dynamo and battery STANDARDS The Lucas regulator with 7 8V typic is still off specs the thermal coefficient is 0 004V C in accordance with Wikipedia USA in cycled use 7 02V at 40 C and 7 5V at 0 C The regulators Lucas and worse Selectronic are off Yuasa specs which induces a reduced lifespan But using a battery manufactured in 1950 impossible to get one in working condition C
Download Pdf Manuals
Related Search