EEVblog 1452 - Fluke PM3370B Combiscope REPAIR

[@EEVblog]

A follow up is on my 2nd channel: kzbin.info/www/bejne/m2qblIyEhbqrmck

[@lolilollolilol7773]

Dave could you do a teardown of some Blitzwolf nomade Bluetooth speakers ? From what I've seen on other videos, the quality inside of these consumer products looks very remarkable for the price. It could make for an interesting video.

[@johnbonham7515]

He barely touched the PTC with that plastic stick, and it moved about 3 mm, and on the back side of the circuit board, its solder is duller than the other solder points, and I think you can almost see a hairline crack... Nice vid as usual.

[@AngDavies]

0:48 for reference i think

[@orti1990]

Yeah, that doesn't look right.

[@greeceuranusputin]

I saw that move too but it didn't register.

Dave you have a huge hint in here!

[@ourplesoop]

Yep at 3:00 that joint is clearly bad

[@bakuhost]

I had similiar problem when repairing amplifier, it would randomly oscillate. I've found suspected section where the oscilations were created but couldn't find anything because it was super random. So i grabbed hot air gun with smallest nozzle and heated elements in that section one by one. Finally I found resistor that was being like 2-3Meg and other time 200k depending of the moon phase or idk but when it was heated a lot it always came back to spec and oscillations dissapeared. Maybe thermall stress when you were resoldering components fixed it enough to work reliable. Try out upside down compressed spray in the primary section. Thermal shock could once again cause failure.

[@donreid358]

The primary switching may have slowed down because the secondary side was no longer drawing much current. It could still have been a secondary side issue or even a shutdown because of some other circuit decided there was a problem.

[@c1ph3rpunk]

After watching your videos for a year I am now able to understand around 35% of what you’re saying. 10 more years and I might be up to 60%.

[@volodymyrzakolodyazhny7740]

Don't stop. I am Ukrainian, and I was like you, but now I understand almost everything (>80-90%).

[@shazam6274]

Its the Over Voltage / Over Current protection SCR circuit in the "mains" power supply. There is a trimpot to adjust it, which works as a voltage divider to set the trip point. It is likely corroded and increases or opens which triggers the SCR to to latch and shut it all down. Clean that with some contact cleaning spray, maybe also "rock" it back and forth a little bit. It is likely to have silver plated contacts, which are tarnished. If it is a "Piher" brand (used thorughout European electronics) it is junk. Philips also made some which were marginally better. Also the link to the Forum is wrong.

[@ikocheratcr]

I was also thinking that the over voltage protection circuit might be triggering .. .

[@tim_buck_too9126]

I saw that in the main PSU diagram - classic "crowbar" arrangement.

[@shazam6274]

Turns out that the "Trimpot" is actually a 1.5K NTC Thermistor, which can be seen in the service manual with schematics Dave put up. Sam problem though, if it opens or increases (solder joint?) then the thyristor latches until power is cycled. It's about an inch from the one Dave's fiddling with, between the two aluminum (OK,OK: Al you mini yum) heat sinks.

[@juergenschimmer960]

This was also a known Problem with Philips TV-Sets. There was almost always some kind of crowbar circuit in this TV-Sets. There they triggered on to high Voltage in the Power Supply or to high Current in the the CRT.

[@IamSwingline]

I would agree. Monitor MV2 test point to see the switcher control signal. The off-time is controlled by V1014, and what looks like the voltage divider circuit R1021 and R1018. As the current flows through the network, R1018's resistance should decrease (NTC resistor), and eventually turn off the SCS (V1014). This will short MV2 to ground, switching off V1019.

[@mikeselectricstuff]

So annoying when things fix themselves - happens to me all the time!

[@lawrencebillson6224]

If ya didn’t fix it: it’s not fixed

[@ThePoxun]

I work in IT so for me just turning things off and on again often causes things to fix themselves

[@KeritechElectronics]

"Repairman syndrome" like Mr Carlson called it?

[@tomaszwota1465]

Lately I had a couple of those "I swear, mister, it didn't switch on at all, now you're here and it works... I don't understand!" cases. Some I could fix, some remained mysteries.

[@SupremeRuleroftheWorld]

the opposite happens to me, i touch it and it instantly breaks.

[@HaseebElectronics]

wow it is fabulous troubleshooting process. best wishes dear sir

[@flymypg]

Contact corrosion? While in the US Navy I often "fixed" equipment failures with a pencil eraser applied to the card-edge. Multiple insertion cycles can have a similar effect. Oxide layers are NOT your friends!

[@brainiac9579]

I really like these videos! I repaired a pre-amp, about a month ago, that had 2 bad, hidden pads. Which made for two bad solder joints (couldn't see them in visual inspection under the microscope). It would work for any amount of time, could be 2 minutes or 2 hours, then the sound would cut out. But at least, I was able to find them both, as the sound would cut out as I would touch the parts lead with the scope probe. I hope your Combiscope's going to stay fixed, but I have a feeling this is just temporary. In my telecom days, every units that behaved like that would fail in the environmental chamber.

[@mysomervda]

Reminds me of a "Watch Wes Work" saying, "Its only temporary unless it works". Scope looks fixed to me, nice work.

[@simonilett998]

And I believe Wes borrowed the saying from Red Green🤣👍🇦🇺

[@michaelweber8136]

Thanks for the repair video Dave! Hope you and your family are well!

[@beaseac]

The repair videos are by far my favorite!

[@paulbeijer341]

These are diode stacks. Multiple PN junctions in series to increase reverse blocking voltage.

[@pufero1]

the old ones on BW TV was like a buch of PN discs inside long ceramic fuse .

[@jabbalechat7910]

Thanks for the video Dave. I've got one of their analog PM3094 scopes which had exactly the same image shrunk to dot in center of screen and then stopped working failures. Watching your video gives me the confidence to explore the psu, which is exactly the same, following your methodology... 👍

[@Mi-Di]

Dave, you can call it fixed after bending pcb and wiggle all primary side components while powered on 😉

[@thomasunsworth425]

One of the black art mysteries of fixing intermittent electronics. I had many years fixing simialr issues in high power transmitters. Dry joints caused a lot but not all issues. Yours sound a real mind bender.

[@TheCod3r]

I just watched the score video to this. Good to see another upload Dave

[@simonilett998]

My favourite theory is that Murphy is always watching...pissing himself laughing at situations like this🤣👍🇦🇺

[@arenaengineering8070]

I had a situation where a 1 MΩ resistor gave a flickering break under voltage, and when measured with a multimeter, it showed that it was working. This led to the fact that the microphone amplifier generated at a frequency of several tens of kilohertz. Possibly high ohmic resistors in the high voltage feedback divider may have problems from time to time.

[@dktr2]

I love this delight with the unusual voltage of the high-voltage diode. Nobody normal will understand it;)

[@thesolderman861]

The rectifier Diode in a Microwave oven at the big capacitor for the voltage doubling is similar, has several Volts (e.g. 6V) in forward voltage. As always: Be careful, the -4200Volts there could kill you or your Multimeter in a blink of an eye.

[@dherrendoerfer]

Time for some percussive maintenance perhaps, plastic rod and a tappedy-tap-tap on the switchers while on ?

[@thehobe2111]

I suspect a fractured solder connection that is thermally stressed due to expansion/contraction. Large physical parts like transformers are mechanically connected but their electrical pins going to the PC board see this tug of war from hot to cold during warm up and cool down. These bad joints normally show a circular break around the soldered pin on the PC board. I have run into this time and time again with consumer electronic hardware, i.e. auto ABS solenoids, projection TV's high voltage transformer etc.

[@EEVblog]

That's why I resoldered them all.

[@mrpappa4105]

Thanks, one of the best fix it videos ever even without the final answer, so interesting & informative to follow your thought process and of course i feel a little better when mr. Murphy not only visiting my bench.

[@iamjadedhobo]

Early on in the video, when pointing out all the voltages going to the bus connector, the +58HT is clearly visible as a signal coming back in from the bus just below the +58V line going out. That whole section can be switch externally !

[@Frank-bg6xz]

The issue is not likely the power supply, contradicting I know. Check the board on the side of the tube and the one on the back of the tube. Those drivers tend to heat up and cause shut down. I've had this happen to my pm3384 a year ago. Unplugged everything and plugged it all back in, it's been working reliably ever since.

[@helifynoe9930]

I had a case that drove me up the wall. A transformer was soldered into the PCB. It had the usual wire wrap construction, and the same posts were thus plugged into the solder holes in the PCB. Anyhow, it turned out to be one of the transformer wires that was just barely making contact that was the cause of the eventual shutdown. Even though this one wire was wrapped around the post several times, it was not soldered properly. When the unit warmed up, it became an open circuit.

[@emmettturner9452]

Reminds me of the Atari 5200 2-port I had that was black and white only. All I had to do was leave it on half an hour then power-cycle. Since then it’s been in color the moment you fire it up. That was in 2017.

[@M0XFXUK]

Yep intermittent faults have to be the worst. If it starts playing up again it may pay to monitor the current levels on the voltage rails. The current may be creeping up to the point where it fails. Just a thought as may help narrow down the fault. Really enjoyed the video.

[@ovalwingnut]

Good Show. Quite iNtErEsTiNg & my "pea sized" gray matter appreciated it, immensely. Thank you.

[@1CM69]

For what it’s worth, we had a family Philips TV back in the ‘80s that would switch off randomly. From vague memory it was the anode cap on the tube breaking down and shorting on cold days when condensation would form on the glass, joys of living in a damp house with no central heating.

[@NoLandMandi]

of course, it's left on the healing bench overnight, and now it's working!

[@seculi7757]

That PTC moved way too easily.

[@DIY-valvular]

A Philips made apparatus suffering from an invisible dry solder on power supply? hmmm, let me think... Yeah! It's like almost every Philips tv set and vcr I fixed in the 80's and 90's! You can get all service manuals, do all troubleshuttings you can imagine, you can measusre every component and find nothing. And then the only thing to do is re-solder all the power supply like crazy and pray a rosary and three padres nuestros and hope the client is satisfied... That was my experience with Philips stuff in my youth.

[@dedamarsovac]

The components at 2:37, I'm pretty convinced they weren't reflow soldered but wave soldered. That's why there would be large pads on the edge of ICs, to "wick off" any possible excess solder during the wave process, to avoid short circuiting

[@TVV-04]

So the fix was a Fluke? I'll see myself out

[@andycarter5617]

When the switching frequency suddenly dropped from 50kHz to 2kHz …. I was thinking you would go straight to the switcher IC datasheet and see what controls the frequency. I had a similar experience and it was the timing capacitor gone wonky.

[@stephenbell9257]

No switcher IC in this power supply. Apart from a couple of op-amps in the feedback circuit It is made up entirely of discrete components.

[@richardbrobeck2384]

Yes, I had a NEC monitor that had issues after it ran for a bit, and it ended up being a Zener diode .

[@oraz.]

I remember seeing a heterodyne frequency analyzer being teared down, on this channel. I hope you find another one in the dumpser.

[@DLSDKING]

Should have given a few whacks to the scope before attempting to fix it as it would have zeroed the fault to solder fracture type failure. I do this test for bulky audio amps and CRT type systems.

[@qzorn4440]

wonderful troubleshooting video 🥰 however to get back to the nifty cheap O-scope purchase... most hobby ebay folks will not have keysight stuff, so testing with the multimeter and circuit switching frequencies sounds can tell a lot 🤔trouble-shooting switching power supplies can be a real challenge and this problem is a head scratcher. thanks a lot

[@poprawa]

I have DIYed passive HV probe for 1Mohm voltmeter, witch consist of many resistors wrapped in PTFE plumbing tape and heat shrink tubing. With 99Mohm probe and 1 Mohm meter it just divides range from 1kV to 100kV. As shielding is poor meter still can die, but it is cheap, risky and dirty way to measure HV

[@poprawa]

It works for any voltmeter with known internal resistance (witch can be measured in some cleaver ways if spec is unknown), just probe must be 99 times the resistance

[@primateinterfacetechnologi6220]

I don't gamble so I don't know about the "risky" part, but cheap and dirty? Count me in... um, what were we talking about again? Oh wow: The stupid grammarly algorithm thinks my comment is "disapproving"... Which is of course, bullshit- because I totally approve. May peace, or whatever you're into, be upon you, sir.

[@radarmusen]

GW121 is very useful for checking hv diodes. I guess hv-diodes are stacked inside, some of them breaking very easy.

[@grahamebenson112]

Having spent years mending Philips kit, looks like a typical Philips dry joint

[@NICK-uy3nl]

I would hit the board with a heat gun, start raising the temp on different sections of the PS board and see if that triggers a shut down. Hopefully that will isolate it to s specific component or parts.

[@Badspot]

The evil spirit inhabiting the machine knew you were getting close, so it jumped into one of the other devices in your shop

[@IvyMike.]

All those caps look like a city from the side view.

[@jonathanrees3765]

Check the joint at 2:54 - short protruding wire, sort of looks OK - but have seen ones like this that are not OK. You have disconnected and reconnected cables - so maybe a simple connector problem. Frequency dropping on input probably rules out crowbar or protection circuits - as still running but in degraded mode.

[@jeffm2787]

Tin whisker that finally blew itself out...who knows.

[@ovalteen4404]

Reminds me of a CRT monitor that a friend had that would randomly shut off for a while. It turned out to be where the VGA plug went to another plug inside of the monitor. One of the pins wasn't making good contact.

[@Ozymandiuus]

I know you've decided this can't be an electrolytic, and I tend to agree, BUT, electrolytics do have a sneaky tendency to reform themselves after a period of time under power. Just sayin'. Could try cooling the entire unit down--throw it out into the snow for a night--and power it back up. Another free suggestion worth its sale price...

[@greeceuranusputin]

"throw it out into the snow for a night" It's summer, and it don't snow there, ever.

[@VincentGroenewold]

Yes, true, perfectly fine looking diodes can be dry or simply not having the proper characteristics anymore.

[@vampiremuffinman7183]

Guess: Surface mount component being affected by a failing solder joint near it. You fixed the solder joint, but when you first had it back together, there was still some heat there, or other stress. When you turned it back on after that initial failure, that stress had eased and now it will work forever...

[@piotr86]

There is a thyristor on the primary side. Usually it is a memory element in a protection circuit. It blocks the operation of the system until the power is turned off. Investigate what can trigger this thyristor.

[@MatthewSuffidy]

When the problem was in the primary side at least you did not probably need some weird flyback thing no one had. It sounds like something moved during the soldering. The fact that it messed up after the soldering suggests it was not reliably fixed. I would not call it fixed yet. Not a lot of experience but I was thinking bad switching control IC. Maybe it is a cracked lead on a component?

[@6581punk]

Never rule out something simple like the power switch or socket being iffy. I had that on a Vectrex game console, the contacts on the switch were arced, a quick clean and it's worked ever since. Of course, unless the switch is doing more than one voltage here it's unlikely to be responsible.

[@davesdigitaldomain]

David it looks like you have a dry joint at 2:55, the pin that joins up to the transformer on the HT side of the power supply also is possible the CRT cathode over-current protection circuit is shutting down the power supply.

[@rael5469]

We see things like this on aircraft all the time. They are very complex systems only on a larger scale than your Combiscope here, and we see this all the time. Go through the system, everything checks good. Put it all back together and what do you know??? the fault disappears. The favorite theory of what could happen is .....micro-corrosion on contacts. That sort of thing. What else?

[@sadteeto]

Classic “my code works but I don’t know why” moments

[@Tekjive]

Figures right, lol. Can’t wait to see part 2 :)

[@KeritechElectronics]

"Nasty don't-touchy" is such a cute way to say this :) I really don't like heisenbugs like this. If it's repeatable, it's something that can be worked on. Will need a lot of extended testing, I see. Nice episode :)

[@donepearce]

Time to wield the long plastic poking stick. That will usually find whatever is being a bugger.

[@Nik930714]

It's a bit ironic that Dave is using a Keysight Educational scope for the repair, that is basically with the same specs, in terms of bandwidth (not channels, or usability or UI, etc, etc), as the professional level scope he is fixing. Ok, ok the Fluke is 10Mhz faster and its historic and the repair was fun.

[@Capt.Marco-Hawk-L.L.A.P]

Dave and your magic fingers, wins again.

[@wernerfritsch6436]

The primary switcher contains a thyristor and a photocoupler. These are the parts I would have an eye on. I have seen thyristors fail in weird ways. Photocouplers tend to lose gain over the years.

[@originaltonywilk]

I'll second that, opto gone 'soggy' is what my money's on :)

[@FireballXL55]

I was going to say I don't think it is the EHT failing, as when EHT drops the picture dims and gets larger with lower EHT.

[@stephenbell9257]

Yes, the fact that the trace collapsed indicates that the deflection voltages had failed not the EHT. If only the EHT had failed but all other voltages were OK then the trace would have expanded due to the CRT deflection sensitivity increasing with lower EHT. I guess that nobody under the age of 60 understands how electrostatic deflection CRTs work anymore so they make incorrect assumptions about the circuit operation and they go on wild goose chases trying to find non-existent faults.

[@Stoneman06660]

The lack of a few firm taps seems to suggest your debugging isn't too thorough, Dave!

[@danmenes3143]

Maybe some electrolytics became leaky from sitting unused for a while. The unit was therefore drawing excessive current, and tripped the OCP. If the current was right on the edge of the trip point, it might take a few minutes, as voltages and currents drift when the unit heats up. But after you repeat the cycle a few times, the capacitors have spent enough time under bias to reform and the problem goes away.

[@BretBowlby]

I need to get something like this. I want to see if I can read the CH2 out on the daisy chain ports used on the Meanwell drivers for Growlights. If I can get it to work, I was thinking that I should be able to put an ESP32 on CH1 that will house a wifi server (for connectivity purposes) and allow it to be programmed to send mimic light cycles that match Day/Night cycle along with seasons (going from a 6000k to a 3500k and some deep reds). Basically make them a bit smarter. I'm guessing this would be a good tool. Sorry, I'm new to the whole engineering stuff. :D Love the CHANNEL btw!

[@ALIKHAYRI]

It maybe a mechanical stressed connection which happens to be cracked, and a thermal effects which make it to be open at non fixed time. And by manipulating the board it just to be in the correct fixed connection. You may manipulate the board later and break that connection and the condition will reappears again.

[@flos251]

You should check the PTC is movet too much when you touched it slightly 0:51

[@gordonwelcher9598]

Sometimes when you disturb an intermittent too much it goes away for a long time. I have seen this with microcontrollers in NEC and ADI computer monitors.

[@andymouse]

Thats a bugger, Murphy dictates it will do it again in 1 year during a moment of maximum usefulness...cheers.

[@eh42]

At some point you stop asking an engineer and start looking for a priest :-) Tech gadgets seem to fear me - if it doesn't smarten up in my presence, then once I pull out the soldering iron it often mysteriously falls back in line and behaves. This scope thought you might be a user (they're cursed in the other direction) and was taunting you until you pulled out the soldering iron. The mischievous angry pixies have been properly disciplined. for now....

[@gordonwelcher9598]

I had the power switch fail on one of these types of scope.

[@fieldlab4]

Freeze spray and heat/soldering gun on low can be useful.

[@krnlg]

Stick a hairdryer in there - helped me find a thermal fault in my scope's trigger circuitry :D Edit: or resolder that PTC!

[@lilstubthumb]

I can't be the only one that was rooting for it to die again in the middle of the outro.

[@adipuiu5603]

That's why I really need a digital oscilloscope, not cutting it with just DMM's.

[@Kc12v140]

I’d be looking in at v1001-v1004 diodes in the mains input to the fly back transformer. Something in the mains input could be dropping the input voltage to the transformer causing the fly back frequency to drop. Book says it should be 20-50khz depending on transformer load and voltage input to the transformer but it would pull down to 2khz and the outputs on the transformer collapsed

[@Kc12v140]

If one of those diodes was disconnecting from the board you’d be getting half rectified ac into the transformer on one side which would screw up the flyback circuit and collapse the secondary voltage immediately. You can prove this by probing for the line trigger voltage as that’s generated prior to the transformer input rectification.

[@AndyHullMcPenguin]

Did you remember to indulge in some percussive testing? A few well placed dunts here and there might re-introduce the fault. If it has gone completely, then it might be (or more likely have been, past tense) a tin whisker somewhere. The re-heating may have weakened it, and the power on failure after heating it may have been its last stand. Is this thing not from round the time when everyone was clamouring to remove leaded solder from their products?

[@mcconkeyb]

Nice Video! My journey is still continuing. My symptoms are similar, but quite different. On my scope, the display never dies. But after 5-10 minutes the scope resets itself, and that means the input voltage range and the time base all go to start-up values, also the triggering, cursors, and all time and voltage measurements also go to their default states (usually off). So I'm thinking that in my scope the processor rail (+5V) must be dragging down slowly and then it hits a reset point. But I don't know why this reset would unload the +5V rail and reset this slow lowering of the rail. An alternative problem, might be that the processor or the flash chips that hold the firmware, are bad and that after a certain period of time they just loose the ability to work, and reset. If this turn out to be true, then I could replace the processor, but I don't have access to new firmware chips and that would mean the end of life for my scope. Dave, keep us updated on this item, I'll be interested to see what happens.

[@primateinterfacetechnologi6220]

The problem with your scope is as follows: damned if I know- I am no expert as it happens, so if I were you, I wouldn't waste my time reading my comment. But if you already are basically committed... which seems likely, well no major harm done. My point being- I don't have a point... Sorry man: guess you're the one today. On a different note- I've been doing this stuff for years and I've never owned an oscilloscope... which may have something to do with or be directly related to, the fact that I generally exist in a confused state, with no idea what the hell is going on... and restricted to all kinds of roundabout methods for determining even the simplest thing... The fact that I am ever able to repair or build anything is somewhat of a miracle under the circumstances. Perhaps due to magic or a similar phenomenon, an oscilloscope will appear in front of me sometime soon... but that seems unlikely... bummer. I've considered trying to build an oscilloscope but that might be difficult without an oscilloscope. money, of course, would solve that but I haven't seen much of that lately. Instead, I have its equal equivalent= time. I believe it was Einstein that proved that- "Time is actually money"- ...he said, clearing his chalkboard before anyone could see what was actually written there; probably obscene stick figure drawings... That's why brilliant theorists often have "people" for the basic arithmetic. or nowadays I guess you just use a calculator. However, I haven't seen a calculator yet that had buttons for the obscene stick figures that are often necessary. Perhaps there is a marketing niche still left open. Peace be upon you, sir. journey on.

[@EEVblog]

Same model?

[@mcconkeyb]

@@EEVblog Almost, my version is the 4 channel 100Mhz with 32K data points. The full featured version (PM3382A). Everything inside looks the same, except the input board at the very bottom (4 vs 2 channels).

[@IanScottJohnston]

Dave, is one end of the SMD cap @2:35 on the underside unsoldered?........kinda looks like it on video. Upper right quadrant of section B4 I think it is.

[@edwinschuddemat8280]

With CRT's nine out of ten times it's the line transistor that shows this kind of behaviour when failing. It's the part that also comes the most under stress. In the past when dealing with CRT's with this kind of behaviour this was the part I always checked first, useally with some cold spray. In this case it's transistor VL1109. BUV28 was notorious, but don't get fooled when you take it out and measure it. You probably won't find anything wrong with it then.

[@Dreamlgider]

That diode at 4min was blown in my 3380 scope. I hope i can get it up and running again.

Did you knock the powered up board with the back of a screwdriver? This might induce the trouble back...

[@hicksbrad]

Nothing is more frustrating to fix than an intermittent issue.

[@petermikus2363]

Thinking of buying one of these for 220€ mind you everything electronics related is expensive as heck in my country unfortunatly

[@whatevernamegoeshere3644]

I worked at a repairman's workshop for a brief while and he was really puzzled when he saw my psu-diode-multimeter ring on the table for checking the diode of a microwave oven. That rotten little thing needs 15 volts across it

[@jerrydaugherty3781]

Fixed while fumbling, That was some intensive fumbling.

[@VincentGroenewold]

Maybe you cleaned something off by pure chance that was causing a short or something when heated up? Super weird stuff sometimes happens after soldering and leaving some fragment of something. Or the other way around and it simply was a bad joint that you reflowed.

[@EngineeringVignettes]

Ruh Rro. Reforming caps pulling the PSU out of Tol? That's skinny but the only thing I can think of...

[@mghumphrey]

Send it to The Signal Path KZbin guy; he is very good at troubleshooting.

[@nathantron]

So would you say in most electronics, the tool itself usually is always the surviving part and the power supply is usually always the fault point? Couldn't that make a HUGE opportunity for a company to make reliable or replaceable power supplies for every shape and size so companies could just offload the work to a standard "size" or replacement part?

[@EEVblog]

Power supplies are usual culprits because they dissipate the power and have high current switching.

[@user-rq6pi6cs2h]

Think about what can cause the reduced frequency of invertor

[@monchiabbad]

I guess the first time after touching up all the soldering points the heath detection point was pre-warmed-up due to the soldering and failed after 4 minutes. Beforehand the part was warming up due to a dodgy soldering point ?

[@barakandl]

When you where talking about upgrading the sample memory it Looked like there is a (bypass?) cap in the way for the larger chip. I suppose it is not necessary and redundant to the larger 1206 bypass looking cap above it.

[@EEVblog]

Yeah, that was a bit weird.

[@matthewsylvester9103]

Please please please upgrade the memory, that would be so cool!

[@gergelysoki1705]

what i do in this cases i take my heat gun with a small nozzle and start to heat up components, not very hot just a little bit. and when it fails, that is the component is bad.

[@simonsmith2802]

I had a mosfet do that once what's the power supply had come up to temperature then the power supply which it down with no warning it turns out that the MOSFET had change thermal characteristics once I changed the component the power supply was fine after that

[@bertblankenstein3738]

I really like the repair vids. Thank you!