The Ultimate Short Circuit High Side MOSFET Tutorial Guide. 10 ways to find the faulty VRM. GPU CPU

[@stephenw4720]

Thanks! Sir!

[@davedave3631]

I’ve watched more tutorials than I can count and this is by far the easiest channel for people like me to learn from. I only wish I didn’t waste so much time on the other channels where content creators just assumed we’d already know what they were talking about. Thank you for taking the time to explain everything.

[@LearnElectronicsRepair]

Thank you and welcome aboard. Also check out the #LER Discord server - I think you might like it Learn Electronics Repair is now on Discord! Come and join the fun and let's fix stuff together, it's free and a nice place to be. discord.gg/vam6YC8vwU

[@Todayisthelife]

I'm as green as it gets.... recently found I have cancer and lost my job so I'm taking this as a hobby. Man am I glad I found you! I have a 3090 to repair. Wish me luck!

[@iaingold7802]

Thanks so much for this video. I think it might be my favourite video on KZbin that I have watched. I am an electronics tech and have not used an ESR meter before... I am going to find one tomorrow. I blew a fuse on my new rx6700 card that I mounted a water cooling block on, using your knowledge and instruction to test the pins I got to deduce that i didn't have a 12v or 3.3v short but decided that what you were explaining would further my knowledge and understanding so watched intently to the end. I think I must have over tightened the screws (for fear of leaving a gap on GPU core due to thick thermal pads on VRMs etc.), I am now hopeful that the fuse was all that is needed. I measured my drivers and they all seem to be the same (with same part numbers at least). I am almost glad I broke my card because I have now found your videos. Thank you for being such a great teacher, i will try to pass on your skills to the students I teach and certainly be sharing with the enthiastic ones your channel

[@hamzalogo]

Watching this video at 3 am trying to sleep, but it had the opposite effect on me, got me so interested and hooked, hats of to you sir, and thank you for the amazing tutorial

[@seeithappen1]

This is much better than learning dry electronics in a school

[@LearnElectronicsRepair]

Cheers

[@kriswillems5661]

This is really good video. Honestly, I am an (fellow?) electronics engineer and have seldom seen such creative approach to finding shorts. It's great summary.

[@LearnElectronicsRepair]

Thank you

[@baghdadiabdellatif1581]

i can't thankyou enough God bless you greetings from north Africa (Algeria)

[@tdtrecordsmusic]

excellent vid. Truly excellent. whenever I come to a situation of having to remove a component like this, I use 2 soldering irons. (Also, puffs of air if needed to blast the solder off.) One iron heats the area and one goes for the pins. Typically, if I can't unsolder it is due to heat sinks, mainly copper pads under components. I wonder if u could have just put an iron directly on the ceramic(black) part of the mosfet. it's dead anyways so no worries about killing it, but this would put the heat where u want it and no where else. Another method of diagnosing is signal injection. Grab the signal generator and put a frequency through stuff. This works amazing. Usually u can read anything. U will have to get used to what happens and learn to read the signals. Resistance shows as signal attentuation. Capacitance shows as signal frequency cutoff, U need to look at bandwidth to check that. Diodes chop ur signal in half. Best part is u can read circuit paths with very low injected signals. Actually U want to use very low currents. 10k or 100k input impedance(resistors). A scope + signal generator can tell u everything u wish to know about circuit layout and the state of the components.

[@LearnElectronicsRepair]

quote Another method of diagnosing is signal injection. Grab the signal generator and put a frequency through stuff. This works amazing. Usually u can read anything. U will have to get used to what happens and learn to read the signals. Resistance shows as signal attentuation. Capacitance shows as signal frequency cutoff Isn't that exactly what I was doing with the ESR meter (which was one of the two most notable successes in this video)?

[@tdtrecordsmusic]

​@@LearnElectronicsRepair Absolutely Yes Indeed it is. However, you are missing the readout which the Oscilloscope provides. I guess I should have mentioned the scope earlier and in a more formal manner. The combo of a scope + signal generator is king. U can also use that combo of signal generator & the gain+overdrive+dynamic microphone but u really have to know what sounds "sound" like. I need a name for that tech. Maybe I should call it the OVERgain mic diagnosis tech? One day I'll make a vid and totally describe and show it. When I do that it will be easier to develop a name. It's quick and dirty but it gets results fast. Its great for quickly discovering shorts + severed traces + faulty caps + crappy diodes. Its really good for detecting signal bleed, like a situation where an inductor/transformer/OR a diode is emanating and leaking signal which is then being picked up by other components or wires. back to the Oscope + signal generator. The waveform which is produced and subsequently read is where all the info is. U have to see the screen and know the input. Can the ESR tool hint that there are parallel paths and tank arrangements ? Does it show the attenuation level, which u can then rework the math and compare to expected results ? With the ESR tool it is only able to make one type of analysis... at one frequency.. One waveform type. So, you can think of the signal gen + Oscope as having a plethora of devices similar to the ESR. Actually, I use this method to determine ESR.

[@mansjacobsson5823]

Thank you for sharing. I'm a total beginner who is just eager to learn about electronics. I just want to be able to fix stuff at home. So I appreciate your time.

[@kriswillems5661]

For those that are wondering why the ESR meter was so good: Because the mosfet is shorted you are in fact measuring between the 12v line and the vrm output. Both the 12v line and the vrm output have capacitors. The other side of all capacitors is connected to ground. So, your ESR meter is measuring the series of the vrm and 12v capacitors (and the ground in in the middle between the both capacitors).

[@LearnElectronicsRepair]

Absolutely correct - and it was a very good demonstration of that fact 😁

[@plinker439]

Where did he put the black probe at 50:25? Exactly what for did he measure the resistance? It's not clear at all.

[@mattbentley8958]

​@plinker439 I believe he was just measuring from each side of the coil. On the good circuit, the esr meter measures the coil resistance. On the shorted circuit, each end of the coil is connected to a capacitor which is connected to ground. Since the esr meter uses a high frequency test signal the capacitors are effectively seen as low resistance connections to ground.

[@marsovac]

Haha i'm not an electrical engineer but a software developer. However after you shown the diagram and identified the fault, I tought "well just desolder the coils one by one and the short will disappear". I continued watching the video and right before the last few methods, where it become complicated, I was thinking to myself "hmm maybe I'm dumb...". And then you left that method for last, and pulled me a prank :D Perhaps it makes sense that they are difficult to desolder, but then it makes sense to put this method as first since it is the most obvious one, and explain the problems and then proceed with others.

[@zeferrum]

I cannot imagine how many hours went into this.. So much for the weekend only hobby!! Thanks so much. . keep up the great work. Martin

[@LearnElectronicsRepair]

A lot! This was one of my most demanding videos to produce.

[@funnyclipcentralz]

Wow wish I found this channel ages ago, you explain everything so clearly and make it easy to understand, best channel I’ve found to learn from. Keep the vids coming cheers

[@ExtremeBogom]

Thanks Richard. Another great video, full of useful information. I've learnt a lot more about mosfets and their relationship with drivers and why there is often two/three mosfets working together. You are a great teacher!

[@LearnElectronicsRepair]

Thanks, glad you enjoyed it 😁

[@technomad900]

You are a legend .. Ive watched many videos on PCB repair .. This covers so much and easy to follow , so useful

[@ppaul6974]

About that welded driver IC, it might be easier to move the coil and other components for that phase one another unpopulated place if is on same power lane, or maybe with a mini grinder to destroy the case of the driver and try to unsolder the metal remains piece by piece the fused one might become easier to be removed but i woldnt put to much hope to come off without a chunk of pcb trace. Thank you for your videos, i wish there was a channel like yours 10-15 years ago it would made my life so much easier! most of your tricks i had to figure it out by myself but still i can find them useful being so well structured as a reminder. Good luck with the repair !

[@LearnElectronicsRepair]

Hi P Paul, yes I totally agree, actually I also suggested this idea on the reply to another comment - to rebuild that phase on another of the 'blank' phases - but I suggested leaving the welded driver in place and just removing the inductor to use in the relocated postion (plus re-route the PWM drive Really most of my 'tricks' are nothing new, they are old school knowledge. Some of which may have been lost, or at least fell into obscurity over the decades but I'm happy to be passing them on to the latest generations of fixers 😉

[@hightttech]

@@LearnElectronicsRepair: Speaking of "old" test methods, do you recall the HP 547A current tracer? Such an AWESOME TOOL. Prior to thermal cameras which, IMO, are the GRAND CHAMPION current tracers and priceless on the bench, I used the 547A to pinpoint shorted and active components. I still have two, but they're mostly retired due to FLIR camera.

[@ppaul6974]

Now i see, at that moment have not read all the comments, sure if rebuilding is taken in consideration the failed driver can remain in his place, once the coil removed the 12v rail have no path to the gpu power rail, only problem that might rise if the low side mosfet in the driver decide to conduct with the command pin in the air, but i think its a low probability and if happens probably will burn the fuse on 12v or trigger the psu protection it will affect the gpu no longer. However doing the rebuild requires a bit of patience and attention to details not to forget or wrong place any of the resistors (i wonder if "pre-driver" survived), anyway i'm sure you can do it :). Thank you again for your videos, I find a lot of value in "your old school knowledge" and how you provide it.

[@wantomek]

Another option is to bridge the coils together and inject voltage between DCin and the bridge thus protecting the GPU chip, but very thorough explanation. Cheers!

[@wantomek]

Also - to remove welded parts we often just grinded them down with a dremmel tool with a round head from dental appliances.

[@chrirep]

What a GREAT video! I am a total amateur but I enjoy watching you as you explain it so well that is easy to follow even for the newbies.... YEAH?!? 🤪 Keep up the good work!!

[@naidol]

Excellent. The circuit diagram drawing explaining each component and then how they packaged into chips we see on the pcb was the best part for me (novice). Look forward to video about removing the mosfet. 😊

[@dwightelvey645]

I have yet another method. I've even used it to find the location of a short on internal power plains. I'll try to describe it. This is even safer than the ones you've used. It requires no current through the components. You need a sensitive meter. I've used meter that can measure 1/10 of a micro volt. First locate the thinnest trace. I believe on these boards it would be the 12V side of the transistors. Set the power supply for the maximum current. and some minimal voltage. I usually use 0.4 V. In you case 5 amps should be enough. Attach your power supply to the two remote ends of the 12V that is across where the FETs are attached. The supply should clearly be in current limit. I usually solder wires to make a more stable contact. Now take one lead of the meter to the load side of the coils. Now with the other lead probe along the 12V connections. Where the voltage measured it the point of the short. I've used this method to even find a short that was as high as 1K ohms. It is all about getting enough current to see a voltage drop across the 12V trace. It works like a wheatstone bridge. For internal layers, I will often choose the ground plain to drive the current across. I solder wires at two diagonal corners and drive diagonally. ( assuming a power rail short ). I'll start with one meter attached to the offending rail. the other lead I probe around the ground at various known ground points. I'll lay a string, held in place with small pieces of tape along the 0 volt line ( often a curved line ). I remove the supply and place it on the other diagonal. I then probe for the zero along the string. It may need zeroing in by measuring the other diagonal, again, for the exact component. I used this method to find a 1K short on a burning board with 600 socket and several hundred more resistors of 1K. The idea is that when the reference matches the drop across the high current line, it is like finding the point of contact along a series of resistors. You know that a lead is connected to one of the junctions so the minimum voltage is the same point. As I noted, do not feed the current directly through the components. In your case some current will branch and flow through the coils so you'll need to also move the reference. It is hoped that the coils have more resistance than the traces. Dwight

[@chrisjames2285]

Im just now getting into this kinda thing on the side as a hobbie. It's really hard to understand out of a textbook you make sense out of it to me thanks, great video very informative I liked and subscribed!!

[@christianmusso629]

Extremely interesting and useful video. Loved also the details on those methods that didn't prove up to the task and also not editing out the hiccups that occurred along the way!

[@davidhollfelder9940]

Great diagnostic strategies. Measuring voltage drop across each coil, is great .. similar to finding a parasitic draw on a vehicle by measuring millivolts across each fuse/ckt tree in the fuse box(es).

[@arnaudjacquemin2248]

Thank you. This is what I call expertise! :)

[@kidtask]

Great Video you are the GURU of Electronics

[@ravenhhca]

Your teaching style is superb.

[@philhulse5355]

A true Stokie I am in Stoke too great to see the great side of Stoke on KZbin Great work you do I'll say thee again on another.

[@lavague7648]

Impressive. After watching this and another couple of your videos i can say that i understood most of the vrms and their troubleshooting. What i enjoy most is that you reffered in the physics behind in a very comprehensive way. Congratulations.

[@CXensation]

Well done Richard! I was surprised to see how the ESR meter excelled in this case, but still the amps and voltage drop (essentially milli ohms) method seems to be the alltime winner. Yeah it definately seems appropriate with a video showing how to cope with these difficult component removal cases. I'd think of the same approach as in this video, where you show different methods with failures and success. Thank you very much for your efforts

[@LearnElectronicsRepair]

My favourite methods here were the same as yours (plus one other) 1. The ESR meter was really impressive but would only work like this where you have a coil with a shorted MOSFET on one side. 2. The Voltage injection (current injection actually) plus milli-volt meter was the other clear winner and more general purpose option as you *only* see a voltage drop if current is passing between the two points where you put the probes, regardless of what is in-between them, even a length of PCB track. Otherwise it just reads zero as I demonstrated 3. If you are on a tight budget, have a multi-meter but can't afford a bench PSU, then the $1 short circuit finder is a clear winner because A: it worked and B: All you need is a LM317, a 22 ohm resistor and a 4.5V-5V voltage source such as AA batteries, (wall wart) power pack or even just use USB power from your PC kzbin.info/www/bejne/hpCVeHqEp5mkZ9U

[@oussamaassal5588]

i know my comment is so late.. but i suggest you to use a grinding pen with some specific tip to act like a CNC to destroy that mosfet =) after that solder a new mosfet . Greatings from Algeria

[@screen-protector]

in the worst-case scenario, you can grind it to the comfort level and then take your time to finish it off. You can see on the best YT channels that we all face the same problem sooner or later ;). It's never easy though if a component can integrate with the mainboard.

[@alexyurchenko3569]

Thanks, ill be trying to find a short on a ASUS B-550 Motherboard with these methods. 3.3V rail is shorted.

[@LearnElectronicsRepair]

Thank you for your support, it is much appreciated

[@karsaifalo]

Hi Richard, Thank you again for the high quality learning matterial! Thanks to you and this video I was able to identify my shorted driver. Please keep up the great work you do, Yeah? :)

[@bombswabs3041]

I really love watching Mr Carlson's Lab, but his videos are a great way to watch his genius in action. Your videos on the other hand are a wonderful mix of application and explanation. I know very little about electronics repair or theory, but am very intrigued and prefer your style. Please keep making more. I could watch for hours!

[@PaulHigginbothamSr]

Did enjoy your work Mac. Twas fun. Ringing was interesting but seems I hope my oscilloscope enjoys the ringing.

[@EShirako]

Thanks! This was far more enlightening than I'd expected it to be. :)

[@christianward1411]

Now that was interesting!! I know little about electronics repair but I learnt a lot with such a well explained video. Thank you

[@AB-yu2tj]

This is wonderful you're a treasure

[@AMCTechRepairs]

Amazing video! I just wish I'd found it sooner as I damaged a HP motherboard literally the other day when I injected voltage, now I understand why!

[@6690Blacky]

Great experiences! I am happy that I found you. 😀 Good luck man, and be always positive, some things do not go so easy.

[@LearnElectronicsRepair]

Oh and don't I know it... but I am nothing if not patient, persistent and determined when it comes to fault finding. Still doesn't mean I can fix them all of course (but I can usually work out why)

[@reklaw3603]

I have learned so much in this video, your talent to show proper diagnostics is brilliant. Truly well presented! Thanks

[@rentnerzockt]

Hy Richard suuuuper then ever Greetings Olli from the Fleemarked

[@LearnElectronicsRepair]

Hi Olli - glad you enjoyed it

[@KB1UIF]

Hi Rich, Great Video!! I'm not sure about the Driver chip being welded to the board, but I have seen components glued to the PCB board during manufacturing. I remember seeing boards wave soldered, and components were glued on the board prior to this process. I also remember seeing "My Mate Vince" ruin a PS4 power supply board trying to remove a component from the board with hot air, but the component would not leave the board for the same reason. So everyone beware !!

[@tasoscyra2785]

everything about modern switching P.S. in a single video , thank you so much .The voltage injection method from 12v power ril to a phase coil is genius , but ESR method is so far genius , CONGRATS!!!!

[@LearnElectronicsRepair]

Thank You!!!!

[@edwardoware9104]

I was late but have taken time to watch everything to the end and man, you are the best. Full of knowledge 🙌🙌🙌🙌

[@henrybartlett1986]

I’m hooked on these now. Great stuff!

[@jay1st1st]

Hi, i have been using your various methods and found out my MXM GTS880 board has a shorted capacitor on one of the rails. Now i need to change it and see if it works ^_^. I am by no means a pro but i like electronics....and really enjoy your channel ! tnx mate or muchas gracias.

[@manuforti1314]

Another excellent video. I have had 2 board where there is no way the hot air is removing a mosfet. Both cases have required a dremel to grind most of the mosfet away then flood with solder to float the remaining parts off. Not easy but doable. I watched a guy do it on a video, he was german but the auto subtitles and with a lot of pausing the video I could see what he was doing.

[@jonnymakers9560]

Excellent video. Thank you for the awesome explanation. I agree it takes a lot of heat to remove those FETs, they have thermal pads underneath perhaps the video manufacturers know this and make it impossible to repair.

[@castlecodersltd]

I definitely did learn something today. Thanks for this knowledge. ☺

[@katnik211]

Another great video. Thanks for sharing. I have an idea how to remove a bad mosfet welded on the board... Just grind it out using a dremel. Using low rpm with a little patience i will be (probably) removed. I believe it worth a try.

[@randomthingschannel2692]

Dang. I just got a lot more aware of many things!!!!! Thank you.

[@AVOIDAVOIDVOID]

Very nice and thorough. Your effort is appreciated.

[@jimmyramkisoen191278]

Very good to understand how vrm works. Thank you for your time en dedication!

[@LearnElectronicsRepair]

You're welcome JImmy

[@boblewis5558]

Love the use of your experience and basic electronic knowledge to use HF rather than DC to find the problem by using an ESR meter - maybe a custom unit with selectable frequencies might be a good tool to build. 👍👍👍 With regard to the inductive pickup a couple of things occured to me as I watched: 1. Usually, or at least often, the internal layout of the coils is of a coil mounted with its winding horizontal, i.e. parallel to the board. I noticed you only tried your little test coil with it orthogonal to the suspect coil. Not the best orientation to maximise the magnetic coupling! 2. Given that the sort of coils used are often flat wound with relatively few turns, and any output to be measured would be small, a larger turns ratio might help. So using a more customised coil with many turns of fine wire on a high permeability graphite rod orientated such that the coil being tested and the test coil are axially aligned might help too? 3. I've never been a fan of using the "what gets hot" method used frequently by Louis Rossman for example, since it "could" potentially exacerbate problems. But your millivolt method of checking voltage across the coil is a FAR better solution. 4. Again though, you are using DC. Might this be another case of better results by using HF to check? An HF feed (rather than DC) on the 12v rail, which can ONLY pass through a shorted MOSFET, would show a higher (any) voltage across the coil. None of the o/c MOSFETS can pass that signal hence no voltage would be present. Another case for the ESR meter as the HF source or a simple 555 circuit with say an auto sweep function as a test tool? 5. Number 4. above might even be a means of improving the potential magnetic output for the coil pickup test. You've shown some great ideas! And a few to add to the "toolbox". Even after a 47 year career as a graduate electronic engineer with over 14 years of in-field troubleshooting and repair of "real" (big 19" PDP and VAX rack) computer systems, never too old or experienced to learn new tricks! Keep 'em coming! 👍👍👍

[@davidhollfelder9940]

When you used the MESR meter, you could try using a guitar pickup coil/inductive pickup (these are very sensitive), with a scope, to find the current/signal flowing thru the coil of the shorted phase, compared to the others.

[@DeadCatX2]

I wonder if you could pick up the magnetic field in the coils using a hall effect sensor as your pickup

[@andymouse]

Hi, great work. I have the same meter as you I've had it for years I even have the same grubby yellow bumper on it ! have you forgotten that this meter has a low ohms (40R) setting that gives you extra resolution and allows you to calibrate out the resistance of the leads ? I ask because I've never seen you use it but you change meters here for a similar setting but does NOT allow you to 'cal out' the resistance of the leads, just curious because you have to look closely at the writing on the Fluke 79 to see the function (40R) and the 'CAL' is referenced on the back. One test you didn't mention was using alcohol on the chips and seeing which one evaporated the alcohol first and as you were putting current through anyway it might have given you another win ! New subscriber and enjoying your work....cheers.

[@LearnElectronicsRepair]

Actually i was not aware of tha tfeature on my Fluke 79, even I owned it since 1993. I should check that out.

[@temyraverdana6421]

Thank Richard, an amazing useful lesson!

[@dougiee6589]

Brilliant methods and well covered I like the sure fire method to save time and not use too fancy smancy testing and yeah thanks for breaking down the gpu circuitry with drawings and nice going mentioning the tester settings while poking around hehe - my confidence is growing to solve my dead MSI gtx 1060 which died under the same circumstance of incorrectly cabling a riser to it in my mining rig damnit

[@LearnElectronicsRepair]

To be honest the fancy smancy stuff was mostly to add some entertainment, and to satisfy myself that it would not work. I think the ESR meter technique surprised a lot of people though. And the voltage injection plus using the millivolt meter to determine which inductor the current flowed through technique was pretty impressive also.

[@davidernstzen1915]

Hi Richard How about a repair to esab inverter welder circuit board Thank you David

[@gianfrancosaladino6320]

this is a great video thanks for sharing your knowledge greetings from venezuela

[@garypoplin4599]

A variation of your ‘voltage injection/ millivoltmeter’ method would be to short the output of each coil to ground and see which one drops the voltage on the power supply: the coil connected to the shorted MOSFET will drop the voltage drastically. Same idea except you’re using only the power supply display without the DMM. 58:46 - You can buy voltage and you can buy resistance; but, you can’t buy current! Current is created by dropping a voltage across a resistance. Just as there is no fluid flow without a pressure difference, there can be no current flow without a voltage difference (drop). Voltage drop is much more efficient at locating faults than resistance readings and you’ve proven that, once again, here. Voltage drop is your friend! Also, the danger of meter damage by inadvertently subjecting the ohmmeter to a live circuit voltage is eliminated.

[@deemauk2591]

Brilliant views, thank you. May be a silly Q, but can you use a Dremel to remove the majority of the package and take it off in bits?

[@CulinarySpy]

Thanks what I was thinking! It's easier to heat small chunks and spread flux, low melting point solder in there too once it's cracked into bits....

[@neebwill]

You are awesome! Love your videos and have learned a fair amount along the way. This one was particularly useful for me.

[@skysurfhf]

Thanks for the video, best for finding short circuit ... Greatings from Portugal 🇵🇹🌟

[@MrMaxeemum]

Excellently explained again. Love these videos.

[@peterdavey2767]

WOW another great video

[@tunkunrunk]

apart from replacing bad caps , nobody ever dared to try fixing graphic cards . Most of us just threw them to trash bin . I've never seen lot of people successfully fixing dead graphic cards . faulty ones are sold on eBay for a few bucks

[@ramzanahmed9237]

You sir are brilliant thank you very much

[@rahulmechatronics]

excellent tutorial

[@larrywildman4381]

It would be fan to see you removing the mosfet with a dremel ... I did it sometimes ... often the damage is between the layers of the pcb. And even in these extreme cases, sometimes you could repair the board (but you need a lot of time and patience .... inject some current, find the hot spot, remove a bit, see if the short has gone, repeat until success.)

[@bmck9097]

Thank you. Excellent video!

[@gbondbasic]

enjoyed this

[@LearnElectronicsRepair]

Thanks Gordon

[@davidhollfelder9940]

Can you grind the bad mosfet with a dremel away enough to then desolder it?

[@dkmmhdk]

To do ring-testing you should probably have a capacitor across the coil. I've testes (TV line output) transformers with a function generator set to small with pulses with a longer pause feeding a winding of the transformer through a resistor, without the resistor the low output impedance of the FG would dampen the the circuit too much, a capacitor across the winding and the scope probe also across the winding. a good transformer should show a series of dampened oscillations, a shorted transformer only a few oscillations (rings)

[@LearnElectronicsRepair]

Maikael Hansen Yeah I thought it was a wild chance, but fun to try. Thanks for the info

[@LeoparkNote4x]

Channel yang sangat bagus, very good channel because take a time to describing in detail for very low knowledge people like me.

[@mpadlite2925]

Thanks for another very interesting and instructive video!!! :) Best regards

[@joekellman2754]

Another great video many thanks. I think a good topic for another video would be on laptop batteries or any type of modern lithium ion batteries that use logic to control the way they charge. as someone who refurbishes a lot of laptops, dead batteries are always an issue, but many times the cells are still good. Any info on this subject i suspect would be much appreciated.

[@knutbjornlarsen4435]

Yes, that would draw a large audience of people frustrated with devices that are still good, and the battery life seems should be longer. Most of us don't know how a battery charges and what built in protection effects charging and battery life.

[@Dutch_off_grid_homesteading]

Heya I like the mili ohm meter that was so clear where the short was. but olso the 1 dollar that you bild is very clear and as you can see there are more way's to rome

[@antonypalmer5804]

brilliant video with lots of information.

[@phillip557]

I like your channel. Please keep making content and don't give up. I am really enjoying your channel. Thank you.

[@LearnElectronicsRepair]

Thank you, I will

[@bluefishactcl1464]

Very good one !!!!

[@ruudb72]

Execlent video. Thanks!

[@tonict2302]

Awesome video

[@SzymekCRX]

I've been wondering why should we be so gentle while injecting voltage via 12V rail to limit it to like 0.8V if there were probably already multiple attempts to run that card with a real ATX PSU connectd inside a PC :)

[@christianmusso629]

Well, as he said an ATX PSU will instantly shut off and won't allow you turn on the machine again until you'd unplug the PSU from mains 🤔

[@surgingcircuits6955]

Excellent video. TY

[@Broxie]

Thanks Richard. Never tried it but stuck like that if I still need the part off I would probably try to grind the chip away.

[@issanesheiwat1396]

Hi Richard, I love the videos but I have a question about installing the mining adapter the wrong way. I totally agree how sending 3.3v to the pcie lanes can fry a gpu. However, my understanding is that these pcie lines have capacitors in series which would effectively act as a DC block when the 3.3v is in contact. Perhaps the 3.3v fries the capacitor into a short and then leads the 3.3v to the gpu? Im just trying to understand the scenario better. Thanks and I appreciate your input

[@LearnElectronicsRepair]

The capacitors are only one one side of the PCB, each PCIe lane consists of four tracks and two of them are not protected. Nor is REFCLK+/- Check this video out, it really explains this in full detail, plus tells you all you need to know (to fix stuff) about how PCIe works. kzbin.info/www/bejne/Y6XJlZ2sm6yEo9E

[@Jookia]

Around 48:36 you show using an ESR meter, but I'm not entirely sure how this works. You're measuring the inductor's resistance and finding a lower resistance because there's a path through the shorted mosfet and through bypass capacitors that are connected to ground on both the load and 12v rails?

[@kriswillems5661]

Because the mosfet is shorted you are in fact measuring between the 12v line and the vrm output. Both the 12v line and the vrm output both have capacitors. The other side of all capacitors is connected to ground. So, your ESR meter is measuring the series of both capacitors (and the ground in in the middle between the both capacitors).

[@LearnElectronicsRepair]

@@kriswillems5661 Exactly. I'll make a short video explaining this today, plus try one other method that someone mentioned in live chat during the premiere last night 🙂

[@rfarrer7685]

Excellent video, very clear explanations and a series of methods you would only achieve from a high level of experience. Thanks for sharing. Ive subscribed 👍

[@juliusvalentinas]

Best video on short finding ! But the card is scrap because the mosfet welded itself?

[@frankreiserm.s.8039]

Great video.

[@enriconapolitano4321]

Thanks Richard.

[@davidlingaard9435]

Thank you for this video sir

[@AKhan-nu9mw]

Very Good, Sir

[@vlastimilhrazdil6757]

Maybe a topic of a future video, but it would be interesting to remove a whole phase from VRM. It will be hard to change a duty cycle and put a bit of stress on other phases or to limit power, but it should be possible, but I am not sure if all control chip will work if one phases isn't connected...

[@LearnElectronicsRepair]

Actually I'm interested in leaving the welded MOSFET driver in place, removing the coil, and rebuild another phase using the coil and new Driver on one of the four 'spare' locations on this VRM. Then connect the correct PWM pin to the new phase?

[@vlastimilhrazdil6757]

@@LearnElectronicsRepairI understand, it is much easier to repair it, but maybe on some older cards that don't have really a value... some experimenting, if it is even possible with little information we have...

[@kriswillems5661]

Be aware that component selection is critical in synchronous convertors and VRMs. Small timing differences can have very bad consequences.You would probably need a driver of the same type.

[@vlastimilhrazdil6757]

@@kriswillems5661 But it's interesting ! If I won't forget, I will surely try this in something, what it will do !

[@xxycom8963]

@Vlastimil Hrazdil , I remember NorthridgeFix doing what you are suggesting, just to test that the video card was working, but obviously with reduced efficiency.

[@zenos9302]

impressive video :)

[@shanesdiy]

Excellent video, thank you for sharing your knowledge!