wael_fathe

IT seems that i tought of diode in very "ohmic" way the diode will provide the easiest way till the gate capacitance has less than 0.7 vdc across it , that time diode will shut and rest of chrage will use resistance path hmm seems logical but why dont they use 0.3vdc drop out shotcky, it nice (duh) , it also have lowest dropout, in inverters they use 1n48xx which is not shoctkey .?

sorry for interupting the naturual flow of the forum ,,though our fiend razmasom seems solved his problem congratulations for geting rid 03 error , during my internet search i wanted to increase my knoweldge on mosfet and igbt driver and i read in the internet same as i red here about the gate diode shown , from what i understand that it easies and excelerates the discharge process of gate capacitance but with diode having 0.7 drop how can we insure that the path of 0.7 diode drop is easier than the path that have 10 ohm resistor ,??

that will make a rovolution in servicing this inverters.... i appreciate your work ... a you observed the current in resitor by observing the voltage accross it so at last you show us gate capacitance "in action" studing something is not as nice as seeing it bravo sir you and colomb made too many inventive ways to test all in one solar inverters thank you guys but with new designs as you said voltronic now make cpu responisble for driving the primary fets and no more sg3525 buiness do you have ideas in mind for the these new designs

what triggers my question in the first place is your discussion with britchgreenracing about why drive in primary fets is different than the drive of dc-dc igbts in scecondery in terms of compoenent used for the drive so now with your answer i conclude that you mean diffrent devices used different drives due to different current rating which effects the gate charge and thus component used for drive that is logical now ,,,and i think i observe your theory by downloading diffrent mosfets of different currents and observe the diffrence in gate charge

when we switch fet once a second it make its transition from off to on and pass through that resistive region once with higher frequency it pass throug that region many many times each passing it emit heat out of its body too much frequency= too much heat that is mainly why scientist invented the resonance converter in the resonance converter the mosfet switched only few times the rest of the cycles are done by the resonance components , this way they can achive high frequency with out actually turn on and off fets at the same frequncy of that converter? so in future we may be repair those and find the fet only work on 20khz while the frequency of the converter is 1 mega hz some call them " soft switchs" as most of the switching happen due to resonanct componnets rather than the switching fets

and i always though of a total independence between voltage applied to the gate- source to turn the mosfet on and the amount of the current passing from drain to source and therefor for a mosfet that intended to switch on and off 10a the amount of charge needed to do that is much lesser than same mosfet in same circuit that siwtch on and off 100a viewing the phenomena form gate side we face a "sudden raise of capacitance if the current at drain to be switched on and off changed from 10a to 100a" am i right

amazing ....as soon as low appear on pin 8 of drive transformer d32 becomes forwarde baised due to fact that it has low infront of it and voltage back of it (gate capacitance stored voltage) it speeds up turn off process and skip the gate resistor ... thanks get it

interesting so does the drive requirement of fet or igbt differ according to the current being driven in drain or collector of a fet or an igbt what i know that gates are voltage driven give that spefic voltage to gate and it will turn on fet or igbt ?

yes i totally aware with this phenomena when smps subected to load that is too high such as the capacitor of the bus it turn into a curret source reducing its voltage and pushs in constant current into the load which is a cap ...cap will charge over time , yes true no need for gradual incrase of voltage i think the smps can handle that the author is from third world so mistake or unaccuracy possible , also possible that he refer to clone models with this unsuual design ,

but then if cpu failed to observe what going on the smps could go to insanely high voltages? so it is still possible to make an open loop ucxx design ,,,,that seems tempting .... the current sense resistor only limit the output should the current at secondery become exessive while the voltage regulation is only done by the turn ratio ,,,i think this is done only for the simple reason that there is one load all the time ,,,for other ucxx design the voltage regulation is must as the load keep changing for exampe in devices with fans fans on requires more current then fan off requires less current the voltage will fluctuate at moment fan on so voltage regulation is needed those inverters are great lesson in electronics i have seens all ideas done in them ,,,,soft start voltage dooblers,,,,snch rectifiers ...digital electonics ,,,,,,you know them you conquer electronics please check the video he only hold a paper and describing i wonder if whathe said is truth if what he is saying is not truth then how the smps is supposed to raise the voltage across the caps if it does that in small time the smps will be stressed it can easily break what the method of bring bus voltage "gradually" in the ucxx based soft starter smps?

have seen youtube video where the author explains that some output of the soft start is fed to its input and with new higher input the output goes even higher ......according to this autor its an infinite process but i doubt that there have to be a point where the uc3845 and stop all output but that limit will be so high that a damage will happen if not stopped by cpu much earlier

sir your method is great but i think this method of colomb is easier 1 rmeove all fets and dc-dc igbt(though for igbts its ok to leave them if they not shorted) 2 short u9 colpler to enable sg3525 3 remove cpu board 4 connect the inverter to dc power supply 48vdc 24vdc etc 5 hit the ac on switch on 6 scope for signals at mosfet and dc dc igbts but i think your method is a must if 1 cpu is built in 2 you have not removed the mosfets yet thogh shorted fet or dc-dc igbt will short its drive it seems good to remove all fets if we wanted to scope for good signal am i right or there is more hidden benifits of your method?

sir you read to your pi pico not from pi PICO i think its typo? quite a legendary work proving all this for the rest of us in a nutshell measuremnt problems are either resistors open op amps haywire voltages unstable and especially -12vdc as it seems the 12+ willl cause more serious problems if it falls segnifcantly or totally collapse nice work sir

that nice and promising asnwer i ordered 1200vdc 30a super fast diode from china i wil try use them in my next repair project and willl report back the results in this forum for the benfit of all of us whho knows maybe some else get cheated and recive igbts with out diodes ....i bought the part with the d at the end as indication of diode but they only send the part with no d at the end of part number which means no diode installed ,,

then i suppose that they will need heat -sinking too ?

then i suppose that they will need heat -sinking too ?

so is that ever possible to remove the synchronised igbts from dc to dc and install general purpopse diodes i know it is not right for normal operation, but can that work for testing purposed ?

u miss understood me i am talking about the hbridge igbts i got 100 pces they have no internal diodes i ordered more 200 pces of 1200 30amp diode can i fit them acorss emiiter collector of an igbt rated 40t60

you misunderstand me i am talking about the h-bridge output igbts i have 100 pces 40t60 they have no diode can i install external one if i find a good way to keep it cool?

yes it has a seperate ac charger .....i scoped between emitter and gate , no wavform at all , i did it at the power on moment and get nothing , with all igbts installed the device work and provides 220ac those unit have mystery ? ------------ u said Also if you do have scope and you can test on IGBT gate and emitter, verify that your voltage swings are within specs, eg - 5.4v during pulse switch off, and 12-15v during pulse switch on period. ------------------------ this unit do not turn its igbts to -5.6 this discussion happen with colomb at the other forum user confirm that near by the igbts there are only 2 zeners one both 16vdc so switch on is probably 15 volt switch off 0vdc

inverter come in with output igbts blown very severly due to customer fed mains ac into the output i replaced all drive componnets that tests bad after that the inverter come on with error 06 i rescanned the area and found bad zener that relflects a short in one copler tl350 with the zener and small opened resistor in serious with zener replaced the inverter works and produce nice 230ac now can tlp 250 replace t350 it actually did but will it last?? second question the zener in circled in red near the smps is replaced with 15vdc zener randomly /....what is the value of the zener i am afraid that it is not the right value ? and igbt will work stressed since it is seems to be across the vcc and ground of the copler by the way i even loaded the 3k inverter with 100watt lamp all igbt are not heat sinked it it work fine igbt only warm another mystery with the igbt removed i turn the inverter on i get no pulses between gate emitter at alll ,,with the igbts installed i get 220 which confirm the presence of the pulses can i calll this phenomenon "phantom ..pulses"" reaally need help i did repair ir but the joy is in understanding

i think this is not a universal measurment that applied on all inverters since there could be a difrrence in resistance value and the gate capacitance of the mosfet ?

oh yes got it ....it boostap circuit....capacitor shorts baise one 2 transistor they work power pin 7 after that cap charges and open circuit both transisors are off again but at that time auxiliary winding brings the juice again and ka3842 works again amazing stuff that capacitor is failure point for sure with out it inverter may not work thanks colomb have little comment for the bidrriectional converter battery and bus are locked by transfer ratio the inverter power dirction is most natural need and demand however beyond capctor of 500vdc the thoery is little diffrent the cpu sense the need and demand and adjust the phse to allow sine wave created by cpu to get out of it or the one at input to get in the cpu dont force the battery to accpet power or to deliver power the power dirction is natural however the cpu complete this power dirction once it reaches to 500vdc capactor from there its export power if he "felt " that power dirction from battery to line and import power if he feels that power dirction into the battery

we never been in situation of some thing that giving and taking at the same time in electrobnics that why it so hard to grasp we acustomer to input that taking or out that giving thanks for explanations but i am embarking on the task of creating forum brief https://forums.aeva.asn.a ...some thing like guide lines because i read and forget all you say and have to back to the piece of info that i forget then i find my self shocked that i dont know where i read it so i start page by page and combed and filtered keep some briefs that will help me construct ideas during trouble shooting time after all mr colomb its easier to search a lake than to search an ocean i combed forum page by page and eventually reach a post where you explained how cpu shut down the main smps.....i wanted to documet the right information to me it is accptable if the cpu shutdown the inverter while the main smps works by solar or ac the cpu can shutdown lcd and the dc to dc but as you explained that cpu can shutdown the main smps by u8 copler but the moment it did so that will cut the power for cpu itself and prevent it from making an orderly shutdown >??? or maybe output caps have enough power to keep cpu work till the inevitable "black out"?? cant see the on off switch on your schematics ? if it is really exists can you show me how actually cpu senses the off on status and make a judgemnt of black out or not to black out with solar power or ac present it can be accptable as those are ored to the main sps and they provide nice power for main sps primary till the cput sense the switch and decided to black it out but with only battery power exist and from what i know the swicth brings power to main sps here we have situation that cpu uses u8 to shutdown some thing that is already being shutdown if cpu make it first he sleeps nicely? if power goes first the cpu may be sleep with courrpted firmware? due to harsh shutdown? thanks

few posts back to refered to the fact that you can actually bring up the dc to dc with 3vdc on battery ...no objection since its fixed ration then it will just spit what ever input it gets to that input x8 but i wonder how you will give the sg3525 its 12 and -12? and this new test i would imagine that it requires cpu board to be out ?