Julian Sucks: Failed Arduino Uno Repair - Rescued!

This video was inches away from being deleted. This happens from time to time - the video starts well and contains useful information, but things just go wrong. I should have learned by now that solder sucking through plated holes generally results in total destruction. In my defense, the repair worked first time around - trying to repeat the process for the camera was asking for trouble. But my career as a field service engineer taught me that you can't give up - there's always a solution. A bit of lateral thinking and a generous helping of bodgery can save the day. Enjoy :)
Check out NEW UNO R3 ATmega328P CH340 Mini USB Board for Compatible-Arduino #UnbrandedGeneric www.ebay.com/itm/NEW-UNO-R3-AT... via @eBay
Check out 5v Micro USB Tiny AVR ISP ATtiny44 USBTinyISP Programmer For Arduino Bootloader www.ebay.com/itm/5v-Micro-USB-... via @eBay
Check out USBTiny USBtinyISP AVR ISP programmer for Arduino bootloader Meag2560 uno r3 #100New www.ebay.com/itm/USBTiny-USBti... via @eBay
Check out 10Pcs16MHz 16.000MHz 16M HZ HC-49S Inline Feet Passive Crystal Oscillator 20pF www.ebay.com/itm/10Pcs16MHz-16... via @eBay

[@jusb1066]

Julian, is this your audition for the new hulk replacement? Julian smash!

[@MarkAMalo]

Very entertaining and informative. Thanks for spending WAY too much time salvaging a $3 Arduino!

[@CharlieTechie]

You are the type of guy I want on Mars when there are no backup electronics boards for the environmental system and failure is not an option. You are much like my self... work for hours trying to fix something that can be purchased for a few dollars. Enjoyed watching, on the edge of my seat till the conclusion.

[@michaelcarlberg2204]

I'm glad you decided to upload it. I enjoy following the reasoning for all different steps - successful or not.

[@paulnolastname9422]

Great video and it made me laugh to know I'm not the only one with "micro-rage."

[@SpeccyMan]

+Paul Osterman You certainly aren't the only one, I once folded an android tablet in two as it upset me.

[@johndoyle4723]

You have more patience than me,for £3 I would just get another one,but I do understand the annoyance of it not working. Don't you just love surface mount. Like many others I have built my own arduino and this helps with understanding and diagnosing faults. Thanks for the video,it was very interesting.

[@tHaH4x0r]

Why not upload the video if the board was not fixed? I think failures in repairs and projects are far more important in terms of learning than successes! Also, maybe viewers have suggestions if you run out of ideas...

[@JulianIlett]

+tHaH4x0r I agree that far more is learnt through failure than repeated success, but it doesn't always make good telly. The DSO138 scope build I did recently started as a failure, so it sometimes works, but failures always require some sort of follow up. As for viewer suggestions, I'm all ears.

[@Luke-san]

+Julian Ilett About the DS0138 Did you actually do the noise mod yet and do you use it? Maybe it would be interesting to make an episode of what stuff you still use and what is being left in the dust drawer.

[@youpattube1]

+Julian Ilett I have a suggestion for you. I was watching your video series about the MPPT, and you noted something that has been a stumbling block in electronics to me for 30 years. Incidentally, I have to laugh every time I think about your interpreting your wife's "get rid of that monstrosity" (the solar panel frame) to mean "it can stay for the winter". Now that is what I call interpreting. As to the suggested topic for a video, in one of the MPPT videos you changed the load seen by the solar panel (the number of bulbs connected on your LAMB) to arrive at an optimum load for peak solar panel power. When I saw that, it was like an old, still buried splinter being bumped. Speaker impedence matching. Used to maximize power. A continuing mystery to me. Why is it necessary ? If you make an oscillator that operates at 60 hz, at 100v, you can stick a light bulb on the output and it lights up (I think). But for some reason if you use a speaker, you must match impedences. How come ? I hope you will consider this as a topic for exploration. I bet 95% of the people who have even heard of impedence matching don't really know what it means. Just something to impress your girlfriend with.

[@Mixbag]

+Julian Ilett using a smd resonator and a through-hole crystal is actually very cleaver and shows some thought did go into making this. i actually did this exact thing on one of my very own Arduino boards i designed a couple of years ago. i did it mainly because smd resonators go up and down in price and sometimes become very hard to find. were as through-hole crystals are very common and extremely cheap.

[@phdDotCom]

+youpattube1 +youpattube1 Audio amplifiers are generally limited by their power supplies, and the rating of the output power semiconductors. You get the maximum power out of an amplifier when the load is of the right impedance. If you consider a sine wave, ideally you want the peak output voltage to be a close as possible to the limit of the DC power supply voltage, and the peak current to be as close as possible to the limits of the power supply or the output devices, whichever is the limiting factor. In that way, you get the maximum power the amplifier can deliver. If the impedance is lower than ideal, you will hit the current limit before the voltage limit, lower actual power. If the impedance is higher than ideal, the amplifier will hit the power supply voltage limit early, it will 'clip', and cannot reach a high enough peak voltage to get maximum current flowing, so lower power is delivered. Many car power amplifiers can drive very low impedances, as that means they don't need to boost the nominal 12v DC to very high levels needed for high power into higher impedances. Most audio power amps produce maximum power into 4 ohms as opposed to 8, as means the power supply voltage can be lower. Some professional power amps (stereo) can be run in 'bridged' mono mode, combining both left and right channels into a single mono amp. Because the output voltage can now swing over twice the voltage range, these usually produce maximum power into 8 ohms. The current limit is the same, the value for each amp independently, but by bridging the two amplifiers, one out of phase with the other, one goes negative while the other goes positive. The output is twice the voltage and can drive the same maximum current into a higher impedance, so about twice the power.

[@mnelson10000]

This is actually my favorite of your recent videos.... even if it didn't eventually work, it's still very interesting to watch your troubleshooting style.

[@visualkandi]

Videos like these are gold and take courage to upload. Thank you

[@JamesSleeman]

Even if the xtal repair had ultimately been unsuccessful it wouldn't have been time to throw out the board, just temporarily inject a square wave (google: avr rescue clock) into XTAL1, then use your ISP programmer to set the fuses to use the internal oscillator.

[@bpresolve]

You know the unwritten law with electronics "Never Give Up" well done Julian very entertaining you should be on the stage

[@p1nesap]

Nice job staying the mark and emerging triumphant.

[@hamandwine]

The only initial error that was made here is actually a very common one: Temperature. Lot of people solder at 400°C and above, cause it goes fast. But this is often above the temperature where the glue that keeps together all these PCB layers is giving up too, at least on cheap PCBs. If you keep in mind that many solder mixtures start to float at temperatures below 300°C and you should always adjust the power of your soldering iron to the thermal capacity needed... The poor crystal has only two very thin traces attached, so next time try again with the iron set to 280°C and only increase if needed. If the quality of the PCB is "unknown" but probably cheap, then prefer de-soldering wire over suction tools. And yes, de-soldering wire adds some thermal buffer, so you could use a soldering iron at a higher temperature.

[@norm1124]

Great video, fixing is such an art. Didn't know more than replacing Caps and resoldering connectors yet. - Please more of this kind!

[@mrtom64]

True British grit and determination....nice one!

[@amtpdb1]

Great to see the process you went through to fix it. Thanks for the video.

[@lindsay5985]

Julian, you are a legend - you don't give up! Brilliant.

[@HoZyVN]

Great video, I couldn't help but laugh at the "hulk smash" moment as you tore the components off the board.

[@slucas601]

Thank you Julian for teaching us a little bit about perseverance. By being of firm resolve to see this through, others are inspired to 'give it another go'. Thanks.

[@mtslyh]

You have way more patience than I do. Nicely done!

[@JerryEricsson]

Now how did I EVER miss this video before? With a name like that, I would think it were posted by a Jullian Troll! Thanks for this, it reassures me that I am not the only one who lifts pads when I am working on boards, here I thought it was my cheap soldering iron! Great learning video!

[@BogusNotions]

If you place a short piece of rubber tubing on the end of the solder sucker with about 2-5mm showing past the end of the original plastic tube then you might find it works much more effectively. The tubing makes a better seal and stops the solder sucker sliding about so much.

[@JulianIlett]

+BogusNotions Good idea. It's the hard end of the sucker nozzle that does all the damage when recoil from the spring smacks the nozzle into the hot fragile PCB pad. Solder suckers really do suck.

[@JJayzX]

+Julian Ilett I use a desoldering iron with the suction bulb - rayshobby.net/wordpress/wp-content/uploads/2011/08/IMG_0577.jpg - I love it.

[@DanieleGiorgino]

+Julian Ilett >Solder suckers suck. Definitely. My life changed when I bought some solder wick, it's seriously great stuff.

[@jpopelish]

The solder sucker jams itself forward in reaction to its piston jumping backward. This causes it to mash down on your hot soldering iron tip causing it to push the pad away from the sucker. The trick is to back the iron away from contact with the pad and sucker, just as you fire the sucker. This will save most pads.

[@jaaasgoed]

For easy desoldering with a solder sucker: 1: Use a chisel tip, even on small vias/pads. The heat transfer of a pointed tip is to small. 2: Put fresh tin on the vias. You where working on "dry" joints, that is always a no-no 3: Keep the tip on for at least 5 seconds. It's a double sided board, so especially with small vias, it will take a long time for both sides of the board to get hot enough. When you use a large tip and fresh solder, the surface tension of the tin will make it so you can keep your tip 0.5 to 1 mm away from the hole or via and still get good heat transfer. At the same time, your solder tip is not covering the hole, so that the solder sucker has all the room it needs to do it's work properly. Loss of a pad or hole is (almost) always an indication of not enough heat to the joint. People try to "make up" for this by applying more mechanical pressure to the iron or the sucker. That in turn leads to a damaged board.

[@TheEPROM9]

I would recommend getting a desoldering station. One of those combined with solder wick makes desoldering a breeze.

[@RaspberryStraw]

Good job, Julian!

[@joblessalex]

When solder sucking, always add a tiny bit of new solder. It makes it heat up faster and sucks it through the hole.

[@YensR]

Fun and educational, thank you for pulling through!

[@ultimateaquatics9114]

I had huge smile At the end when it worked Well done 👍 Thanks 😊

[@fantuznet]

great guess on impedance around pins 13 !! got the same issue on a Mega and had already changed the clock with mo results. will retry tomorrow to jump solder on the chips correct pin (arduino mega clone)

[@simonzodiac90]

there is plenty to learn from this video ... thanks for uploading it :)

[@NickStallman]

Worst case, you can use an AVRISP mkII to manually send a clock to the AVR just to reprogram the fuse bits, and tell it to use the internal oscillator (~8Mhz) rather than the crystal. It would be half the speed but it would work just fine. Whether different programmers can send the clock would vary depending on the programmer.

[@shaputer]

Thanks for uploading ! Perserverence DOES pay off. I learned a lot as always. :=}

[@webcompanion]

That is awesome! Nice fix, I would just leave the blink sketch on it and mount it somewhere in your shop... it has a story and history behind it now.

[@iforce2d]

I've had a strange spate of failed components lately, two arduino nanos, two nRF24s, a brushless ESC and the accelerometer of an IMU breakout board.... I think all the good luck I've had with this stuff until now is swinging back into balance. Unfortunately my diagnostic skills are not as good as yours and for the price they cost, I can handle buying a replacement :)

[@JulianIlett]

+iforce2d Yes, but do you violently attack the offending modules just to show them who's boss? Loving your videos by the way - it wouldn't surprise me if Amazon started their last mile deliveries using cheap ass nRF controlled drones.

[@muh1h1]

+iforce2d I am breaking one ESC after another these days, just had a 25A ESC fail on me, aparantly it couldn't handle a 2212 1000KV motor, that was run for month before with a 18A ESC...

[@JohnBailey39]

+Julian Ilett Don't worry.. You are far from the only one to extract a measure of revenge from uncooperative objects.. Best not ask what happened to the two chilli plants in my kitchen though..

[@iforce2d]

+Julian Ilett I can neither confirm nor deny any reports you may have heard about violent attacks on electronic equipment under my care. Occasionally I may find myself fantasizing about driving my car over my useless wifi router and using the power steering to grind it into the concrete, but this has not happened as yet.

[@Slider2732]

+iforce2d If it does, can you put it on KZbin please :) Thanks. Oh, multiple camera angles too, if you would be so kind.

[@DavidWatts]

Well done, I am not sure I would have thought it was the crystal.

[@NivagSwerdna]

@6:58 True engineering skills in action! Nice video. I have ordered 8 of these NHdunio boards to use as donors since the parts are so cheap. I will look out for the XTAL shorting problem. Thx

[@linksmith1057]

This is why I don't use a solder sucker, I use fluxed, copper wick for removing solder.

[@BobMellor1954]

Sounds like Julian turning into the Hulk! Thanks for uploading.

[@nafarjadeed7378]

Great video, Jullian. Thank and best regards.

[@agonymobile]

There supposed to be gasket under the crystal in the manufacturing process! And because those pads are for SMD crystal and got solder on it.. You need an oscilloscope for checking boards... Its a must for almost any high tech device. Btw, some ic's like those replace ATMEL ftdi fail to program or communicate with the pc, one of the boards like that didn't work with 6 laptops but fortunately works with my laptop.

[@Edmorbus]

Nice save Great video

[@hk84]

Nice work! Hate when everything looks well to then discover the pads have started to lift.

[@SadamFlu]

Yay! Nice save.

[@reggiebacci]

I like the old DIP version of the uno much better than the surface mount ones. I had the actual 328P itself pack up half-way through an upload. Got a new one off ebay with bootloader for just over a quid, shoved it in the socket and good to go again.

[@hippynurd]

Tip, always pull the iron away before using the sucker to prevent this kind of damage. Ive actually repaired this kind of thing so that you never knew the bad thing happened, but it takes a lot of effort. The easy way is to just remove the iron before hitting the sucker, and you will never damage the board.

[@arcadeuk]

Hi Julian, the old manual soldering pump was a very harsh method of desoldering (as you well saw yourself) If you don't have access to a dedicated desoldering iron, might I suggest that you consider using a high quality desoldering brain instead? If used properly they are actually very effective and less damaging to the PCB. Keep in mind that solder braid varies significantly in quality, and thus ease of use, so if like many people, you have had poor experience with braid in the past, I would try a decent brand (3M and Chemtronics spring to mind) Thankyou for sharing your videos

[@zaprodk]

+arcadeuk Desoldering brain :D I approve.

[@arcadeuk]

+zaprodk Damnit :)

[@boldford]

I've found it is sometimes useful to give the desoldering BRAID a liberal does of flux.

[@vidasvv]

Nice save !!!

[@viscomjim]

Julian, I am not sure if you covered this somewhere else, but could you divulge what your camera and microphone set up your use when you make these wonderful videos? I was trying to do something like this and can't figure out how to do the lighting and camera setup to look as good as yours and how you get such good sound. I really enjoy your videos and have learned quite a bit from you. Thanks for your efforts!!!!

[@BowersElectronics]

I think once that pad was messed up you should have placed the Crystal in from the bottom up. or even when placing the Crystal in from the top keep solder on the lead, and make sure the whole lead is tinned, and push it in. then you wouldn't need the 22pF cap. correct me if I am wrong.

[@RupinChheda]

I think nail polish or any nail paint on the crystal body would have also worked.

[@Coolkeys2009]

Wow perseverance does really pay off. Can those processors run from an internal RC oscillator?

[@dh1ao]

don't be afraid measuring SMD, see you soldered through hole stuff to the smd pads, so measuring is even easier. Yes these whang han lo boards are miserable in quality but for a few cents you can't go wrong, at least you may salvage parts from them. Thanks for the vid Regards Peter

[@ozgemmo3445]

Hi Julian - what an excellent video; I totally agree with tHaH4x0r's comment! We all greatly learn much more about how the Uno works from your dogged persistance! keep up the excellent work. Cheers from Down Under, Ian

[@protuningaustralia6795]

Well done Julian, you made me laugh so hard when you were growling and saying you were cross. I would have smashed that UNO into a thousand pieces with a large hammer if i was in that situation. So once again very well done on keeping your composure and rectifying the issue, Regards Dave

[@1pcfred]

Good job! Sometimes the solder sucker is a bit too aggressive for board rework. Then solder wick can be a safer bet.

[@deslomeslager]

I see a lot people commenting about desoldering. After I managed to take the component out, I have my own trick. I look very carefully where to heat up the solder, and then I turn the pcb around and I blow through the hole. And yes, there is a paper to catch all that solder flying around. And be sure not to spit .. .. But this is a cheap trick and has worked for me at least 4 out of 5 times. I hope other people can use this.

[@Slider2732]

+deslomeslager Neat trick, will try it ! Am forever leaving thin bridges or covered holes and never thought to simply blow the rest away.

[@kamil004]

roarrrrr wrrrrrr xD masive destruction. good job ;)

[@daab889]

Nice Job!

[@flymario8046]

LOL! Epic Save!

[@CSO3DOnline]

haha,Julian rage

[@andljoy]

Dont you just hate it when something should work to repair it but it does not . That's why i keep a broken macbook pro logic board around for random beatings when something goes wrong with the more expensive macs :P

[@y0rkiebar]

Bodgetastic !

[@dumle29]

I've noticed a lot of people talking about the desoldering, some saying that a pump is a harsh method. It doesn't have to be, as long as the pump isn't moved or jerked on the hot pads, it's fine. Something like this helps a ton with that: www.ebay.com/itm/30W-220V-2-in1-Soldering-Iron-PCB-Solder-Sucker-Desoldering-Pump-Welding-Tool-/361351959330?hash=item5422415322 It helps a ton, because you can heat up the join with the pump itself, and aren't under time preassure to apply it and trigger it. This means you can jus place it directly over the hole, hold it steady, and trigger it.

[@donaldfilbert4832]

I'm a little fuzzy on how you arrived at the crystal as the culprit in the 1st place (before de-soldering and observing the short). Can you expand on your logic there ?? Did the programming attempt with the USB AVR programmer NOT work? You didn't mention that. And - did you say that the USB AVR programmer WOULD require a functioning crystal? So - should I assume that the USB AVR programming didn't work and did require a functioning crystal only (i.e. none of the other board parts are involved in AVR programming) - so that's how you concluded that it was a crystal failure?? Why didn't you think it was a board trace failure? Very interesting trouble shoot Julian !! Thanks for your response !!

[@maurospagnoletti4787]

Hello, please don't damage the board. To put out a component do you have to touch with solder both the terminals togheter, than it comes out from the other side. Bye

[@Disillusioned_one]

Grrrr the hulk, don't make me angry ardy.

[@_MisterG]

IT'S ALIVE!!!! - Frankenduino

[@MithatKonar]

I use round wooden toothpicks to clear out through hole, er, holes if it seems there might be a danger of pads lifting.

[@BeginnersTech314]

i have a nano clone that i think i burnt out somehow. i was messing around with it one day and i think i crossed some wires while it was plugged in and it just stoped working. but those nanos are so small i dont think it worth the time to fix.. by the way love your videos. keep up the great work.

[@norm1124]

Had one failed too, check the diode ;-)

[@BeginnersTech314]

norm1124 ok thanks for the suggestion!!! i may make a video and if i get it to work i may post it.

[@pauldusa]

22 puff puff mate. I respect you 99% jolly o

[@jameslamb4573]

Whoo-hoo! If you can't build it, bodge it (following hours of painstaking, not to mention eye-straining, iron-work and research)!

[@zaprodk]

Julian, i thought you got yourself a better soldering iron other than that old Antex?

[@websitesthatneedanem]

Nothing like a bit of constructive destruction!!

[@leobav2425]

My arduino uno had a bad input diode that rendered the 3.3 and 5 volt rails bad. The fix was putting a rectifier diode in place of the surface mount diode.

[@louigi6001]

Sucking the solder like that is time consuming ... get a round wooden tooth pic, heat up the solder and push the pointed end in with a decent force (squeeze fit the wood in the hole a litte to get a good clean) .. then it cool off and take out the tooth pic ... much much faster ;)

[@nonoaidnono]

As far from genuino lol loved that bit

[@maicod]

its a chinduino :D

[@douro20]

The only problem I've found with my NHduino is a handful of bad solder joints on the thru-hole components...

[@yuppiehi]

You can't leave the soldering iron on the soldering pad that long. That's the the problem using that sucking device. You should get a good desoldering gun. I have the Hakko FR-300, and although it is expensive, it is well worth the money because you really should be doing things properly, including rework.

[@1pcfred]

+yuppiehi GTFO You can buy like 100 NHduino UNO boards for what a Hakko FR-300 costs.

[@JerryEricsson]

Ok, it's been a couple years now since this video, are you still using that board? If so, did you find a project where it could be mounted and made a permanent say, ghost finder or some such trinket?

[@JulianIlett]

Pretty sure that Uno is now on my penny organ project - and it still works fine :)

[@jenniferbonete4852]

The microcrontroler is a smd component not like the atmega328p

[@raymondheath7668]

I'm the same way, I'll spend $60 in labor on a $3 board!

[@GeorgeGeorge-xj2bc]

I bought a clone named M.J duino and from the factory had installed LM1117-3v3 instead of 5v regulator and the result was only working if powered from usb cable and not from 12v external.Malfunctions to work around and about but not exactly and the repairings of them are usually expected.

[@TheUnnamedGent]

I've wrecked a number of pads on PCBs with a solder sucker. Do you think solder wick/braid would be less likely to damage stuff?

[@JulianIlett]

+TheUnnamedGent Yes, I reckon solder braid would have worked here. I don't know why I continue to use that sucker - it's killed more through plated holes than I care to remember.

[@MrOpenGL]

+Julian Ilett If you put leaded solder on it before sucking, it will suck it better since it has a lower melting point than the lead-free rubbish.

[@HillOrStream]

+John Ridley Strongly agree on the hollow tip heated sucker bulb thing, especially for bigger jobs. Just keep extra tips around for it.

[@tillytony]

I would have enjoyed this video even if the repair wasn't successful so would have been worth uploading anyway. Quality content is quality content.

[@AndyWJP]

Maybe that area of copper on the underside was not connected to anything? You could have done a quick continuity test. And I thought that you had a DSO so why not probe the signals to look for oscillations? Ha ha, I would probably spend far too long trying to rescue some cheap kit as well.

[@SabretoothBarnacle]

By not soldering that one leg first time around didn't that mean that the 1M resistor wasn't connected?

[@Madmax23419]

Atmega dont need those caps to startup, only the crystal must be enough, atmega328 works without those caps, maybe not 100% on right frequently but he will startup. The caps are probably not the problem, there must be something else wrong.

[@petti78]

No scoping of the clock signal anywhere during the work? Funny that. Especially since I remember fighting with a bunch of these same cheapo e-bay crystals a while back. Turned out that many of them have a really weak output and are not capable of driving the chip, with or without the caps. I just chucked the bad ones away and kept the good ones after measuring them all in-circuit with a simple micro. Often you can even just lose the caps and it will still work just fine.

[@superdau]

+petti78 Scoping won't work. The input capacitance of the probe will in many cases stop the crystal from oscillating. It's like adding another of these load caps.

[@UkSkin]

Ha-Ha - You win! Take that Technology!

[@andymouse]

Do you still have this board?

[@ElGatoLoco698]

Is that what Dave Jones means when he says "bodge"? hehe

[@reeseyme9613]

if i would to have this issue i will most likely remove the critical & useful components; say atmega328, crystal, voltage regulator & headers then transfer to a perf board and program the chip with an ISP instead. the minimum configurations....

[@channel_B5]

6:56 I burst out laughing :D

[@dainiakasble]

I'm I the only one who thinks, that there is no ground connection where you soldered those capacitors at the back? i don't have this board, so cant check, but in the video doesn't look, that it's connected anywhere. :)

[@maicod]

just saw this video for the first time. It was so funny hearing your rrrrrrrraaaaaaaaaawwwwwwwwwwwing :)

[@leedevries5660]

Flooding both pads from the bottom with a little more solder and then quickly alternately heating the pads with a quick downward tap and the crystal would have fallen out. I actually use two TS100 solder tools to do this which is even better as gravity will do the job. IMHO solder suckers are quite brutal on tracks. Desoldering braid is only a little better. I enjoyed the vid. Rexlee

[@goopy0078]

Der teuerste UNO aller zeiten bei dem Aufwand 😂 sorry not in Englisch

[@wasanthawimaladharm]

Cheers jullian

[@zaprodk]

I would have removed the xtal alltogether and programmed the chip to use internal oscillator and using it like that - bearing in mind that the oscillator isn't perfect.

[@JulianIlett]

I was all set to go down that route. But programming a new bootloader needs the MPU to run and that needs the crystal until the fuse settings have been changed. Kind of a catch22.

[@bergarteric5713]

Very good job realy ...fine and cleaver cooooolllll........