Pentium III Drama: Removing, Resoldering, and Hoping for the Best!

[@danielktdoranie]

You mad man! Nobody does this much work for a slot 1 processor. I am super impressed!

[@bitsundbolts]

The curiosity :) That is why I do it. And because I learn something by trying it! Training for future Voodoo 5 repairs :D

[@simontay4851]

18:47 WOW, that is absolutely beautiful! The music fits perfectly. Its mesmerising!

[@lerkzor]

18:35 Voodoo magic, surface tension centering every ball perfectly. SO satisfying. You also have the patience of a saint, that was a LOT of balls to place exactly on location by hand without a grid to help you.

[@bitsundbolts]

Thanks! Yeah, it took some time, but it worked perfectly :)

[@ruxandy]

It is satisfying, but you have no idea how easy it is to mess up... If you give it too much airflow and/or apply too much flux, those little bastards will combine into bigger balls (also due to surface tension). :-)

[@1kreature]

A flux tip: Do not put flux completely around on all 4 corners as you create a air-pocket in middle. Make sure you leave one side without flux and when you heat it up the paste will flow in under the chip and come out on the side you did not flux. Capillary forces will make this happen and it's a good indication that flux has passed under the middle of the chip.

[@bitsundbolts]

Great tip! Yes, you are totally right! I missed that one. My flux was bubbling on all four corners! I will remember this for next time! Thanks!

[@lerkzor]

Back when Louis Rossmann was exclusively a board-level repair channel, I watched him do a lot of BGA re-balls. He always puts a drop of Antec flux on the PCB, then spreads it evenly with his finger - just as was done here. CAVEAT: I have never reballed a chip. That being said, I did not see anything wrong with the technique, I would wager that the chip was damaged internally - perhaps by over-voltage.

[@1kreature]

​@@lerkzor I was referring to the removal. You can see as he lifts the old chip that the center pads are gray due to the flux not reaching them. This can lead to issues as the balls develop oxides over time and the flux is there to help break this down and more evenly spread the heat too.

[@lerkzor]

@ Ahh, I see. Thanks for clarifying.

[@tony359]

This is actually a good tip, you can see the air pocket when the CPU was lifted! Thanks!

[@T3hBeowulf]

That was an amazing reball attempt. Despite the result, the technique looked fine to me. Excellent work!

[@bitsundbolts]

Thank you! Let's see what other comments I'll get when the video goes live next Thursday. I think it was a nice result, no popcorn chip, and the solder connections look nice. Too bad the processor didn't work in the end 😞

[@ukaszrutkowski8368]

1. Great content despite non working processor, 2. Huge props for background music choice. Jazzy!

[@bitsundbolts]

Thank you!

[@minombredepila1580]

You never fail. You either win or learn !!

[@Vile-Flesh]

Very, very impressive. I never thought I would see someone go to these lengths to fix a Slot 1 CPU. I commend you for your effort and experience is certainly the best teacher. The best way to learn is to try and retry over and over. That is a lot of work to reball that CPU without a stencil. It's pretty amazing how the surface tension makes the solder balls center themselves to the solder pads. My main daily computer from 2009-2011 was a Pentium 3 667ghz and it was good enough for watching youtube back then with Win2k.

[@Mr76Pontiac]

Placing those solder balls is one hell of a feat on its own. Mesmerizing. Even to the point where when I get home (At work right now) I'm going to replay and screen-cap that to a background on my Laptop. This also makes me VERY excited about getting my digital microscope and be able to do work similar to this. Me being 48 at the end of Jan, the eyeballs are getting fuzzy enough that I'm not able to read details on smaller chips anymore. :( It'll be cool to hook up a 21" LCD and be able to see this kinda detail.

[@Vile-Flesh]

I'm right there with you. I'm 44 and my vision has degraded a lot in just the last few years. I cannot see the pins on a Pentium 4 CPU to straighten them without some kind of magnification and I cannot make out any print on small chips or memory modules. I waited a bit late to get into this hobby but I'm still going to try it as I have a mountain of old hardware to tinker with and never enough time and I still intend to fix the broken traces of our first 486 muthaboard dad bought for us in 1992.

[@michaelturner2806]

12:15 beautiful removal there, happy to see no damaged traces from when the chip got bumped 16:48 oh to see the solder wick going over the whole cpu underside, I can imagine the satisfying feeling like popping every bubble in a big roll of bubble wrap manually placing all those solder balls doesn't look as fun to do though; I imagine it's like patching and re-inflating every bubble in a large roll of bubble wrap

[@blakecasimir]

I think the process you followed was solid and the result, physically, looks excellent. At the least you have improved your technique!

[@janusfoxy]

That BGA rework was masterfully done. Thank you for sharing, especially when things don't go well. Who knows how over-heated or over-volted this CPU has been in the past. I love seeing the hardware I grew up with get a chance for restoration even when the chances are slim.

[@justinkovacik9436]

I was really impressed with your reballing of that BGA chip. It's too bad it didn't work out. The solder joints looked just beautiful

[@ctiborkoza8944]

Nice video, even if you didn't manage to save the CPU, I personally use a slightly lower temperature for the rebal of the old PC hardware, otherwise the work on the rebal was fine. I wish you a lot of success in the next BGA chip resale, it only requires practical experience, nothing more

[@bitsundbolts]

Thanks! Yes, I think it depends on a few variables like which PCB you're trying to solder the BGA chip back on. The more you practice the easier it will get.

[@tiemanowo]

17:10 You've laid down all solder balls one by one? .... respect!

[@r4z4m4t4z]

nice efforts again, heat didnt look too crazy. yes doing it to a working core is the way to go.

[@dolphhandcreme]

over 350°C is way to got, you'll popcorn the chip carrier and/or the pcb. You can use 350°C for regular soldering or changing small components, not for such huge parts. around 250-275°C seems better, depending on your heater and the calibration of your hotair.

[@felixokeefe]

Add to that that you should always bake parts before starting rework. 110°C for 8 hours. It is important to dry the parts thoroughly as any trapped moisture can cause tiny explosions as it turns to steam at these reflow temperatures.

[@rebmcr]

"Couldn't find a BGA stencil" You absolute madman!

[@tadeustad]

I'm in awe of your dedication (and patience of a Buddha) to aligning 495 freaking solder balls by hand without a stencil, that's why your videos are so informative and entertaining at the same time!

[@arthurmann578]

Though I'm no pro at reballing, especially now that I am blind in one eye, it looked like you did an absolutely beautiful job on the process there! I think you may be correct in that the CPU was probably already on its way out due to mishandling at the scrap yard where you found it. Keep up the good work and don't let it dishearten your future efforts as you did a really professional job there to me! 👍👍

[@AttilaSVK]

Hats off for having so much patience for a job with such unclear results. Despite the failure it's still amazing content. I just love how the new solder balls popped into place on the bottom of the CPU.

[@HandFromCoffin]

I'd never get that thing working.. nice job. This is my new fav stuff to watch in the background while working.

[@dycedargselderbrother5353]

I'm very impressed with the manual reball. I was thinking that, surely, a BGA rework station would enter the picture.

[@Mamiya645]

Well it was a good attempt, and it makes good practice. Good vid, subscribing. 17:29 atchoo

[@esc2dos]

"How Do You Get To Carnegie Hall? Practice, practice, practice" Excellent work. Seems like such a bizarre oversight not to clearly mark the Freq on the Chip.

[@woodch]

It's a shame this didn't work, but as others have said-- you learn from failures as well as successes. You have the steady hands of a surgeon! I'd never be able to deal with BGA soldering like that.

[@tony359]

Sorry to see BGA didn't work out - welcome to the club! Thanks for showing the PCB and for mentioning my channel as usual! :) I don't think you made mistakes, everything seemed to be nicely done. One thing I might have spotted: when you went to nudge the CPU, the first time I think it moved but didn't come back, indicating it wasn't molten yet. Maybe that misaligned the balls? But it was a minor nudge... If the fan stops, wouldn't that be a short? Did you try checking for shorts between VCORE and GND? Also, was that my impression or the post card didn't show the clock LED? Next BGA will be good! :) Thanks for the video!

[@d18c7db]

At 10:55 the datasheet for those chips is freely available, from the datasheet: The SN74TVC16222A provides 23 parallel NMOS pass transistors with a common gate. The low on-state resistance of the switch allows connections to be made with minimal propagation delay

[@Stratotank3r]

Respekt, dass du alle Kügelchen von Hand aufgesetzt hast! Schade, dass die CPU jetzt keinen Mucks mehr macht. Ich würde folgendes testen und machen: 1. Fordert die CPU immer noch 1,8V an? Das Board startet aber der Lüfter hält an 2. Möglich, dass einige Balls keinen Kontakt herstellen. Einfach Flux verteilen und das Ding wie bei Tony aufheizen und leicht andrücken wenn das Lot flüssig wird. 3. Diesen Core auf ein anderes PCB setzen und den anderen Core (funktionsfähig) auf dieses PCB. Allerdings dann nur mit BGA Schablone.

[@Geomanb]

12:48 this is so kind and sweet

[@DefenderOfBoston-yo2tl]

Ladies and gentlemen, we've finally found him: the most patient man ...in the world (to be vocalized Jeremy Clarkson' style 😉). Placing 495 solder balls manually is quite the task - you've certainly got balls! 😄 The process looked flawless to me, although it also looks like the chip and surrounding area may be a bit discolored afterwards, which might be indicative of too much heat? May also just have been the lighting though. You just can't save them all. Obligatory Picard quote: "It is possible to commit no mistakes and still lose." As always, A for effort though! You certainly did get some personal improvement out of it and it's still satisfying to watch, so I consider it a win nonetheless. 👏

[@bitsundbolts]

Haha, yeah :) Aligning almost 500 solder balls seems to be a lot, but it also is relaxing :) at least for me because I know what comes after. Did you SEE how those balls nicely flow into place? When I did my first Voodoo 3 BGA reflow, the balls were floating away because I used too much flux on the chip. So, technically, I had to place more solder balls during that project because I had to do it twice 🙃

[@DefenderOfBoston-yo2tl]

​@@bitsundbolts Yes! That's one of the things that make your video so satisfying to watch. The springiness when the intermolecular forces hold the chip in place that can be used as confirmation that the solder balls are successfully re-liquified are a highlight for me too. True that you placed more balls before, but then you decided to do it AGAIN. 👏😆

[@envoycdx]

Rough with the smooth Alex. Hopefully the next is more successful, damn good try though!

[@bitsundbolts]

Thank you! Next time, I'll try on a known working CPU. There will be a day when I'll succeed! 😅

[@aspinx]

Hands up for manual reballing of this thing... That's dedication! Also, I think you used too much flux when soldering it back.... not sure you can clean it all under the chip without using ultrasonic cleaner.

[@bitsundbolts]

Thank you!

[@ileox4170]

Exactly the BGA process I was talking about last year ❤️ you did a very good job doing it! Aspecialy without a BGA Station. I also reballed them from hand, Hope you did it your first try, i‘ve messed up the CPU solderballs so many times before it was done properly 😂 The Problem with hotair is the not so well heat coverige of the CPU PCB itself😮 that‘s why the Color is darkening. Burned my first two as well 🫣 But my offering for SL4BS is still there! They work very good in Xbox Classics as a 1Ghz Upgrade 😍

[@masterofdisaster1904]

Hello. For reballing try to not exceed 340 degree for leaded balls or 360 degree for non leaded balls. Also the temperature for preheater do not exceed 165 degree (for this PCB). Anyway i admire you for the patience in reballing without stencil - great job.

[@bitsundbolts]

I think the PCB never got hotter than 120 degrees over the preheater. there is a substantial gap between the PCB and the heated plate. But thanks for sharing - so, in general, I was a bit over with the heat.

[@Pauel3312]

Wow awesome! I was wondering what is the music you chose at 14:50 ? I think it sounds really awesome

[@drgusman]

Really amazing even if it does not work, that manual reballing is a first class work.

[@pavelfara9333]

You did a good job. I assume the chip was damaged by the scrapper who pryed the cartridge. I have seen this many times and chances are it damaged the die. Keep it for parts - sometimes they crack resistors or other things and you would be happy to have a donnor! Sometimes they also ceack the PCB during this process and you can find a good die on a bad PCB. Another chance to pratice and get a working unit.

[@lexluthermiester]

@BitsUndBolts You did everything right. I've done reflows, you didn't miss even one thing. Nor did you overheat that chip. If anything you might not have given it enough heat to melt and bond the solder balls in the center area of the chip. Those chips & PCBs are designed to withstand up to 440C for extended times without damage. That die was likely already damaged in a way that is not easily seen. I say well done!

[@monad_tcp]

10:18 The chip seems to be just a voltage translator from the CPU voltage to the motherboard TTL voltage.

[@felixokeefe]

Probably a bus buffer IC. Probably custom order ICs as i can't find any info online either. SSOP-48. Same type of package often used for bus switches and buffers of the same time period.

[@SidneyCritic]

I doubt chips on the edge would matter because it would be laid out far from the edge in case the silicon wafer separation cut nicks something. Check if any nets are shorted, and even reflow it again, because you've got nothing to lose, and what happened to that lifted pad, ie, was it still connected. ERS presses down on the chip.

[@bitsundbolts]

The pad was fine - it just looked like it lifted, but it was just how the pads are made - they are on a small plateau that exposes the pad to the top. The surrounding gap is to prevent the pad and the solder ball to make connection to the sides. That gap looked like the pad lifted, but that wasn't the case.

[@zabykrinich3392]

Hells job to set million balls to million pads. You are very patient 😂. Great job and great video

[@bitsundbolts]

Reballing takes a bit longer when it's done by hand, but still very possible with those chips. It's a nice exercise to keep your hands steady 😊

[@michaeldibrino6320]

Great video! Maybe try just hot plate with no hot air next time to avoid overheating the chip?

[@Geomanb]

excellent technique, the fact that the CPU doesn't work has nothing to do with your flawless work

[@TheTallGirl]

What a crazy maniac who put balls on the chip by hand. That's the craziest thing I ever saw.

[@stoptheirlies]

I know how you feel, but you can't win them all mate. I see nothing wrong with your re-ball, nicely done. Bob. UK

[@bitsundbolts]

Thanks! Glad to hear that! I am also very happy with my second attempt to solder BGA chips!

[@mstrVLT]

10:50 SN74 TVC16222A provides 23 parallel NMOS pass transistors with a common gate

[@AnnaVannieuwenhuyse]

I agree! They're likely all a similar type of voltage limiting transistor array. The purpose is the same and the board seems the same.

[@thebayandurpoghosyanshow]

If I remember it right, all those chips are 22-bit voltage clamps, acting as bus controllers for the CPU.

[@AnnaVannieuwenhuyse]

Rumour has it it might be 23 bits 😮 48 pin chip, 23 pairs plus one set of gate/gnd

[@thetechgenie7374]

Good job manually reballing the CPU itself one by one. The issue I saw was with the flux putting it on all corners and wat to much. just put a very thin layer on pads instead. Center CPU once and use a square nozzle which you can get for the Quick stations. For larger BGA you need a larger nozzle or fully cover the whole chip, on smaller ones you can get away with the regular nozzle, I done large ones with it but you have to have more control and a preheater set at about 160c and bring up temperature slowly as you can popcorn the PCB. Hold it steady above and you known it done when you see CPU move slightly, or you tap it and bounces back in alignment, like you did at the very end, or you can use the moving of the handle to dance it as well. When you did the first nudge, it didn't bounce back, so might have thrown it out or bridged a ball. The trick to better chance of success is to use less very thin layer of flux. I learn that on early myself when started with PS3 many years ago. Then who knows the CPU many have not worked to begin with, as at the end did see the CPU bounce back correctly?

[@tigheklory]

You did a good job regardless! Your wife is awesome!

[@alexanderg2211]

Have you checked the resistance on power pins for any abnormal resistance?

[@ruxandy]

Me, while watching your video: "We interrupt this program for the RTX 5090 reviews, we'll be back shortly. 😀"

[@bitsundbolts]

Hehe - I heard they are slow :)

[@ruxandy]

@@bitsundbolts, yep, boring reviews, continuing your video, because it's certainly more interesting. 😀

[@M0UAW_IO83]

That TI chip is a voltage clamp, SN74TVC16222ADL, I would be 99.9% certain the Phillips one is the same thing

[@bitsundbolts]

Yes, it is probably the same chip. I think I could have just moved any of those chips around.

[@cybcarr]

23:38 top right corner seems to b damaged. rip.

[@kokodin5895]

are qr-like codes on the cores all the same or different?

[@bitsundbolts]

I am not sure about those QR codes. However, there is a tool from Intel that can read QR codes on the heat spreader. Unfortunately, that only works from Core 2 and newer. The QR code on those CPUs is a mystery to me.

[@kokodin5895]

@@bitsundbolts yes but you only have few models and visual comparacy could shine some light on it, is it a core revision or entire cpu if they all match that just a core revision if they all diffrent check if the one on the plastic edge match the one on the core

[@turbinegraphics16]

you could still take it off to see how well you did

[@bitsundbolts]

Maybe I will try it once more, to reflow, then to remove it. I will do that in case I find another PCB with a Pentium III at the scrapyard without markings.

[@Daveyk021]

Damn, manually re-balling that CPU; you have much more patience than I do - lol.

[@bitsundbolts]

It does take time, but once you get the hang of aligning those balls, the sky is the limit ;) It took about 20 minutes to reball that CPU.

[@andreabc1469]

nice work!

[@Raptor3388]

That reballing is a bit crazy ! I have several really damaged socket 370 and socket A CPU dies, and it's absolutely not a problem. Even a Duron that's missing a chunk and works just fine. I don't think those slightly damaged edges are an issue.

[@Andr0-Zero]

Dig around the Original Xbox modding scene. We have stencils for this CPU as we sometimes have to reflow or upgrade our 733Mhz to 1000Mhz. Think I have a stencil file somewhere...

[@bitsundbolts]

Yes, I've heard that those 1GHz CPUs are sought after to mod old Xboxes.

[@mikehensley78]

It was probably already dead and you just made all the connections is why it acts different. Don't give up. You'll get a good result if you keep going.

[@samio3907]

I'm pretty sure you could 3D print a stencil for those solder balls.

[@dosdoktor]

That looks like a very baked good chip. STILL good chip. It did not exfoliate, no bulges, had a few cooked boards and chips, they did still work, you may have not failed to reball it, but the IC on the back is most likely not compatible.

[@1kreature]

BTW: The fan powering down could be from overcurrent as the CPU activates something. Check the carrier board for a shorted rail? Possibly a cap?

[@OVERKILL_PINBALL]

A+ effort

[@kstricl]

From the looks of the work, that cpu was just Balled up to begin with... (Side note: I burnt up a 600mhz Celeron of the era based on the same platform. I believe it was when Intel went back to socketed MB's. I had brief glances of the overclock potential of that chip just before it failed.)

[@menotyou8369]

GTL=Gunning transceiver logic, 11c=Emitter-coupled logic, probably buss transceivers, but incompatible logic levels.

[@awilliams1701]

I thought you did great with the reballing. I'll bet the chip was already on its last legs.

[@danielhn93]

I've got an old Compaq Deskpro EP/SB computer and I was trying to learn how the components from back then worked. I'm more experienced with modern hardware. I bought a Pentium 3 500mhz cpu thinking it would work on the motherboard, but nope. So I tried a 300mhz Pentium 2 to replace the old Celeron 244 and it did work. But its a 440bx motherboard and I can't seem to find much info on cpu support for it. The 500mhz just didn't post at all. But overall, I'm astounded by how these Slot 1 CPUs work.

[@WinrichNaujoks]

Och Mönsch, die Enttäuschung in deiner Stimme!

[@webfischi]

Next episode this guy is buying a cnc to produce his own stencils

[@braschlosan]

Make a new video where you reball one of those other good cpu. I enjoyed this video even if it wasn't fully successful

[@kenabi]

at this point, (if its fixable) there's either a problem with the translator chip, the pcb, or the passives. the translator can be swapped, the passives can be tested, the pcb.. ehhh. its also entirely possible the chip itself is just toast, or that it won't boot with the 1.8v mod. though that last ones less likely to get zero result, with just a freeze being the normal scenario. good attempt though.

[@joaoc_PT]

Well, it still is a great experience oportunity, that's the thing with learning curves, the last 10% are exponencial. Maybe first test cpu on another good working cpu-board first? I don't know if those are multilayer boards also. Also maybe its possible, like in other sockets, to test the voltage and front side bus before soldering the CPU.

[@B24Fox]

I'm sorry things didn't work out... 😕 But as I commented on your previous video: you REALLY need to start insulating things that don't need to get hot, using some multiple layers of paper tape. (like the CPU die, in this case 🙄). P.S. upon some testing with my finger on a hot surface, I concluded that paper alone is better than paper with Kapton tape on top. Kapton tape spreads the heat and and holds on to it for longer. So if you have good paper tape that doesn't singe easily, use multiple layers of it, without kapton tape on top. P.P.S. chips and/or ICs usually have a heat tolerance time specified in their datasheets. Something like "350C for 40seconds", or whatever. Keep this in mind; because depending on the situation, high temps for minutes on end, usually can lead to disaster. It sometimes can be better after you preheated the board, to kick your hot-air station into high gear, and finish the job as fast as possible.

[@DeveloperAccount-m5i]

Didn't even considered manual reballing an option ;) Very well done

[@timschonherr7674]

The number looks like a compaq sparepart number, they are usually using a similar pattern i.e. 1234567-001. maybe those were build for compaq?

[@VladoT]

Your technique was OK, I guess the cpu was already damaged and gave up the ghost when heated again. Those chips were already very fragile from the factory.

[@bitsundbolts]

I think so too. That chip was scraped over the floor, had so many scratches and dents. I still wonder where the C0 came from. It might have been some secondary system that reported that code right before the CPU takes over. But that is just a guess.

[@Neksus-M06]

I have 2 cpus in similar condition, C0. I know they used to work, making a huge difference. I'll try to reflow if all else fails. How much time did it take to melt the solder balls when you started using hot air?

[@bitsundbolts]

If you are heating the PCB where the CPU is still attached, then it will take 60-90 seconds. It is not quick. Depending on what technique you used. First, I heated up the PCB that was already at around 120 degrees from the Preheater. Then you slowly bring the board to higher temps. I just get closer and closer with the hot air station instead of changing temperature.

[@_____pyro]

Pay attention to corners of the crystal. If one of them or more are chipped there is a very big chance this can be the fault

[@Hadisabetghadam]

0:82 That CPU Remind Me Of Zhaoxin CPU Or Chipset Zhaoxin Is Part Of Join Venture Between VIA(YES THAT VIA) And Shanghai Government

[@lemagreengreen]

Did you put the original smaller chip back on the board after determining it wasn't the fault? I'm just wondering if that chip is programmed somehow with something specific about the CPU. edit: apparently according to others it is just some voltage clamp chip, probably doesn't matter.

[@K10driver]

To say it in german: Zumindest hat Du 495 Balls of Zinn gehabt. Unfortnatly it doesnt work. But to have the patience to reball the whole CPU without mask is at least worth a video :)

[@harukaapl]

Maybe there is no connection for one or few soldering balls after reballing. You can try again rebail CPU, now it could be only better :) Good luck!

[@bitsundbolts]

Hehe, maybe. But I think one reball is enough. I might try to reflow it again, but if you get that bounce-back the CPU settled in place and all solder balls should make proper contact with the pads.

[@vasiliakulich1608]

I wanted p3 so much at the time..

[@bitsundbolts]

Ah, I skipped the P3. I used to have the Athlon Slot A and stayed with AMD until the Core 2 Duo.

[@vasiliakulich1608]

@bitsundbolts yeah, a year later, around 02 parents bought me the first PC with Duron :) which wasn't bad at all! I read so much books about windows, dos, programming, pc journals etc even before I've got the first computer.. then came Athlon XP and from there I had a job myself and could afford whatewer I like.. but still have the Athlon completely setup with Win XP and want those pentiums so bad.. :) tech was mystical back then, now it's 5090 and whatever I don't care.. maybe it's the age but probably not.. thanks for the response - your channel and you personally are great!

[@TigTex]

Before calling it dead, I would reflow it one more time. Because you've used a heat gun in circular motion there's a chance that the balls right under the core are not soldered. Everything else looks fine and even if you've exceeded the temperature I highly doubt the chip has suffered any damage. Usually if the temperature is too high, the pcb will start to "popcorn". If you are brave... You can also put that core in another pcb as the one you've got has suffered a lot of damage and might be the reason why it wasn't working in first place.

[@notneb82]

18:13 Please, take your seats.

[@bitsundbolts]

And fasten your seatbelts!

[@StephanHampel]

Maybe the CPU isn't dead at all. Maybe there's just a defective capacitor on the back of the board.

[@impulse_guy_22]

I used to work for an OEM. What you may have there is an Intel Conf cpu (Engineering Sample...ES). It'll be a one-off in the sense that e.g. two trays got made and sent out for eval. As such, it may not match up with anything you'll find again. It may well be that the retail chip was socket370. If it does boot, you may well find the bios doesnt id it correctly. These sample chips were super weird and mostly unstable depending on the stepping. And other times they were reasonably representative of what you would purchase two months down the line. Good luck!

[@bitsundbolts]

Interesting. Well, I did find all kinds of weird stuff at the scrapyard and wouldn't be surprised if there would be something like an ES in between all those CPUs! I definitely found a few fake Pentium CPUs which will appear soon in a video :)

[@goclunker]

370 era chips were suuuuper fragile, especially socket a amd durons. Try swapping it to another slot 1 card

[@alanarmstrong2323]

No failure in trying

[@bitsundbolts]

Very true!

[@guillaumedelaroque8519]

350 degrees is too high for me i suggest 280 degrees max for soldering and desoldeing if you want re-use a chip desoldering

[@dolphhandcreme]

Your hotair is way to high! The solder should melt around 220-230°C (assuming you're using PB balls). I'd go for a maximum of 250-280°C as you are also pushing heat from the bottom via your hotplate. With over 300°C you'll popcorn the bga parts and maybe the pcb. That's the end. Learned it the hard way while changing chipsets. ;-)

[@bitsundbolts]

Yes, I agree that too much heat will end up pop with popcorn. I will try to dial in my temperatures to a better lever in the future. I did follow the protocol to dry the chip before removing it by leaving it for several hours at around 100 degrees over the preheater. That should have driven out most - if not all moisture if there was any.

[@dolphhandcreme]

@@bitsundbolts while drying components isn't the worst idea, there's another form of popcorning (or maybe, we shall say "bubbling"). It's when parts of the pcb or chip carrier reach temperatures in the range of 300-400°C and the epoxy/bond/glue starts building gas. The catastrophic delamination of the layers starts which most likely kills the part. The interposer under the cpu chip itself, which connect the chip to the bga-balls is very allergic to this failure. Sometimes you only hear a small "click", sometimes you see bubbles forming. In anycase the part is lost. I did this failure several times until i learned to stay roughly 15-30°C above the melting temperature of the solder, use a good preheater and take my time. The chip can withstand 200-250°C for a relatively long period of time, but 350-400°C only for some seconds.

[@AndrewFremantle]

hello 😃❤

[@peteregan9750]

try a known 2 chips that are good swap over, then you will know its not your technique

[@huberthans4312]

Shit happens.

[@jerry1333]

He's dead, Jim!

[@arubberroomwithrats]

100th like 🤤

[@bitsundbolts]

Legend! Thank you!

[@arubberroomwithrats]

@@bitsundbolts np man

[@tlrptg]

reballing these cpu-s is a waste of time and resurses.

[@RivaDynamite]

some say it is also a good practice - me aswell ;-)

[@pufero1]

On the resoldering never use to much flux theses air bubles made balls jump .

[@bitsundbolts]

Yes, I saw that. I create a seal by applying flux all around. Next time: less flux and leave one side open for it to breath.

[@MrHimer12]

20:44 you have warped chip, look at those "balls" I don't know older chhips but newer intel mobile chips are prone to warping/getting out of shape. Not entirely your fault.

[@MrHimer12]

and from persfective of canera later also banana PCB so could be both.