Oh wow!!! Good eyes. I totally didn't see that until now. Good thing I stopped it when I did!

I didn’t notice that before. Wow! Great catch!

@@damianvila When I first saw it I was like.... "Did that fuse just say Hnnnrrgghhh!!?"

@@preferredimage Ahahahaha! Absolutely. You made me laugh. :D

Absolutely!

I was just about to write a comment to that effect. They have a nice low ESR and make good decoupling capacitors. They also really do not like to be reverse biased!

They're damned dangerous things particularly when becoming old or subjected to overvoltage where they can spontaneously ignite and launch themselves as burning fireballs!

Yes, the low ESR enables them to dump or build charge faster. Combined with higher capacity and better temperature and frequency stability than cheramic, they are a typical go-to for decoupling digital circuits.

Haha, yes, I just saw it!! Probably because I work more on older computers, I tend to look out for the little blue caps, but you're right, they can be any color.

The tantalum didn't explode because PC power supplies can supply more than 10A and 7805 only 1.5A

@@VladoT I've seen a 1A rated 7805 deliver 3A for a few seconds before exploding itself.

Well worth replacing it as its been abused too.

Yes, it should just be replaced straight out. Good point.

@@NoelsRetroLab Tantalums especially. Once you heat them up or break down the oxide layer they are time bombs just waiting to short dead solid. Tantalums had much lower ESR than aluminum electrolytics in those days, and nothing else in the same size range had the same capacitance if you wanted 1uF or more. These days just replace it with a low-ESR electrolytic.

@@lwilton The beloved flame emitting capacitors :-)

@@lwilton or a MLCC, those have excellent performance as well.

If I recall correctly, they're smaller than a ceramic for those same capacities too, because they're technically an electrolytic themselves (just not an aluminum, and either partly or wholly lacking aluminum electrolytic's "self healing" capability).

Thank you! 👍😃

Thank you! Glad you're finding them useful. I'm always honored to be mentioned in the same sentence as Jan and Adrian 😃

Right. That would be exactly the same kind of situation. That's interesting.

Thank you! I'm really glad they're helpful. I also learned a lot initially about repairing computers from some early repair videos, so I totally get it.

Yes, good point. Rob told me about it and I checked, but it was already fixed on that board. I didn't include it in the video to avoid confusion, but you're right that it could have been causing all sorts of weird problems.

Great to hear! 👍

Makes sense. At least yours didn't blow up either! 😃

Polar caps that have been charged with the wrong polarity, cannot be trusted anymore. Always replace.

Thank you! The followup is a doozy! 😃

Glad you like them!

Wow, that was lucky! Depending in which part of the circuit they were, they could have exploded.

Also notice that electrolytic caps mark the negative lead, while on tantalums the positive lead is the one that is marked. Aren't standards wonderful?

Interesting! Those details are definitely beyond my depth of analog electronics, so that's really helpful. I ended up replacing it with an electrolytic one of the same value. Do you see any potential problems with that?

Yeah, I'll replace it next time. Stay tuned for the mystery of the CIAs 😃

@@NoelsRetroLab Such a cliff-hanger! Is the follow up video out yet?

How long did it take you to source the parts for it? (TBH I'd love to try doing either a C128 [what I owned growing up] or an Amiga 500 [always wanted one] )

Good luck with that! The hardest part is going to be getting the components (and trusting they work correctly).

You're welcome. I'm editing part 2 right now, so it's coming soon! 😃

Thanks! Glad you enjoyed it. Good luck with that build. I also have an 8088 kit to put together at some point! 😃

@@NoelsRetroLab ooh would be very interested in your 8088 proj when you get round to it.

Glad to hear it!

True, that's probably a better way, but you need to be intimately familiar with the circuit itself to pick the right order and make sure you got all the components required up to that stage. I suppose a kit could provide those instructions, but most of the ones I've seen just have you solder everything, and at best, they'll provide some checks afterwards. I do like your idea though.

​@@NoelsRetroLab Many thanks for the reply. Most kit instructions follow that lottery approach. I designed and built commercial prototypes from 1970's onwards for many decades where it was necessary to use an incremental approach. In the 70's through mid 80's prototype PCBs did not have solder resist so careful soldering was required to avoid unwanted short circuits. I do really like some of the modern mod kits that use bespoke flexible PCBs to replace bodgy wire connection to neaten their installation.

The 7805 might have helped, but you're right, it's not a bad idea to replace it.

Thank you! I think you're going to like the root cause of the CIA problems. That's why I decided to save a whole episode for it, so I can take the time it deserves.

@@NoelsRetroLab Looking forward to it!

That would have been crazy! (unless you heard something rattling, that would have been an easy hint 😃).

@@NoelsRetroLab it was crazy, going downstairs to the parking many times. I discovered it by cleaning the circuit. #be-noel-for-a-day

Oh I didn't realize the C128 was already available. Maybe that's how I'll finally get my hands on one! 😃 Good luck with the build!

It's already in the pipe. Coming up next Thursday!

Glad you liked it!

Agreed. I'll swap it out as soon as I pull the board out again. I honestly didn't realize they would be permanently damaged, but everybody seems to be in agreement about that. Thanks!

@@NoelsRetroLab : The issue with tantalums is that like aluminum caps, they're electrolytics, so internal misconfigurations of the chemistry are caused by reversing their polarity (which is why electrolytics require a DC bias to properly handle AC uses). However, while aluminum electrolytics will _slowly_ repair themselves if subjected to a low current of the correct polarity, tantalums don't, and supposedly go through some sort of collapse cascade.

If you're going to swap it with a ceramic, just leave it out. Tants have different frequency response than ceramics do, so replacing it with another ceramic won't help. There's already a ceramic on that rail.

Good point. I didn't go there because I knew that this board design was already tested and working with other people. But you're right that there was always the possibility of this one being defective somehow.

@@NoelsRetroLab Sure, but what i was talking about wasnt the design but rather the possibility of a manufacturing fault of that perticular PCB that you got in your hands, like i mentioned it has happened to me in some cases that one board out of many had faulty traces.

Probably!! 🤣 It was right there in plain view.

Nah, I trust my oscilloscope to see that we had a black screen 😃 (actually, I think I did plug it in and it just confirmed there was signal but black screen).

They way they made them in the 80's was to buy defective memory chips for next to nothing, that could only address half their capacity, use the half that worked in each chip and mount double the number of chips.. Sinclair did this with one of their ZX Spectrum revisions. I think it was with 4164 chips, not sure Or was your comment a sarcastic nod to this?

@@paulmichaelfreedman8334 I understand perfectly what you say and no, I wasn' t sarcastic at all. I just quoted the obvious: Somebody bought something in 2021 and it just didn' t work...

@@paulmichaelfreedman8334 You are correct, Sir; the first few revisions* of the ZX Spectrum 48K used those "half" 4164 RAM chips. The boards were designed so that by soldering in some jumpers, it could accommodate chips with either the lower half or the upper half working. Another set of jumpers allowed the use of RAMs from different manufacturers. *Except the very first version; that one used a small daughter board that sat above the main PCB.

Seriously! It must have been a second away from a nasty surprise 😃

Editing it right now!

Thank you! Cheers!

Thank you! I overlay everything directly on Premiere Pro (main view, oscilloscope view, plus any schematics and graphics).

@@NoelsRetroLab OMG! Thank you again for a so hard work! It’s not only a matter of filming, also an incredible amount of post editing! Thank you

I think you're right. I didn't know about that but other people mentioned the same thing. Maybe the 7805 helped it a bit by shutting down, but you're right that it would be safer to replace it. Probably an electrolytic one would be fine too.

Good luck!

Wrapping up the edit right now, so hopefully in a week or two.

I noticed that, too. I have red, green, yellow, blue at hand. This is usually a color code for the maximum voltage of the tantalum capacitor and it is possibly manufacturer dependent. But there is also a norm specified. Search for "Capacitor Voltage Colour Code Table". In that yellow is 20V and blue is 35V for Type J (Dipped Tantalum Capacitors).

Also: The reverse driven of the capacitor possibly has the electrolyte and foils damaged and it has a very shortened lifetime. It is better to exchange that capacitor with a new one.

That's what I'm hearing, especially PCs. The ones I encounter on the computers I repair are usually blue (like the original one in that C64).

Yes, good idea. Planning to replace it.

That's great! I'm still unclear why the original board is full of ferrite beads. I get what they do electronically (kinda) but I don't understand why they felt the need to use them everywhere. There are other computers of the time without a single one.

A little bit 😃 But at least all the major problems were solved before that.

@@NoelsRetroLab It sounds like the most interesting one is still to come though - looking forward to it! :)

Thank you! No, there's no concern about accidental contact. I actually removed the metal pin completely, so you're left with the plastic from the pin header separating the two solder points.

About his tantalum explosion? Adrian's is one of my favorite channels! 😃

As a matter of fact, I do! 😃 www.amazon.co.uk/Spiratronics-TH2-013-IC-Straightener-Tool/dp/B0093Z7JH0

@@NoelsRetroLab Thank you... It's $52cdn on Amazon.CA 😱

Most of the ones I see in computers that go through my workbench seem to be blue (like the original one on this C64 board). Maybe it's a European thing, or maybe it's just the computers I usually work with. I've been told that most of the ones in PCs are yellow, which I usually don't work with, so maybe that's the main source of the difference.

ah right, the ones on the older machines I get are yellow, and usually failed, are the blue ones more reliable at all?

I agree. I think a lot of it is also due to the fact that you can wrap your head around one of those computers intimately and figure things out yourself. Once you move into the 90s, things start getting a LOT more complicated.

@@NoelsRetroLab If I were a professor I would personally teach computer-applied electronics basics on old high level computers (C64, AMIGA, ETC.) rather than new low level computers (Raspberry, SOCs, etc). And yeah, I'm a 90s kid (not a kid anymore actually haha) and I remember the HUGE spike between pentium generations and how Windows XP was "impossible" to run properly. Such thing hasn't happened nowadays. And with the semiconductor shortage, the hardware spikes will get even slower. It sounds "bad", but is it?

@@tspawn35 Agreed. And also with this "circular economy" tendency today and the insane amount of electronic waste we do, there is a lot of good stuff that could happen. In my country, repair shops now only repair blenders, refrigerators and stoves. I have some broken LCD displays, fully functional CRTs, motherboards and Pentium 4 era hardware (PCI tuner cards, IDE drives, DSL Modems, etc) that don't have a "use" anymore but I just don't feel like throwing them away. As you say, doing board level repair makes sense as it is cheaper but I'm no sure if it is, you know, "marketable" for companies to let you repair your stuff :(

​@@SUP3RP3DR0L1V3Was also thinking about the whole idea of how you might teach programming, and the vintage systems that have the OS baked onto the hardware (like the Commodore machines) also have another huge advantage--it's VERY hard to screw up the machine itself---e.g. a hard power cycle will put you back to a known good state. Have you heard of a project called the Commander X16? This is a "modern 8 bit computer" based off the vic-20 architecture. This also got me thinking (and yes most of this is a pollyanna view :D )....once the CX16 it is released in kit form, would be some sort of "summer program" that would teach both the lower level programming and practical soldering/electronics. The idea is the students would either start on the emulator, or physical CX16s to learn the coding, but along side of this, they would also learn practical electronics/soldering/debugging as they assemble their own CX16 from kit. At the end of the program, assuming things went well, they'd end up with a fully functional "modern vintage computer" that they get to keep

​@@artstrutzenberg7197 I'm an IT Engineer personally and only took a small subject on electronics at school (my university had a totally software-oriented curriculum, maybe some networking over here and there). I learnt to understand high level programming languages just by practicing and not reading theory or the principles behind it (it was cool at the beginning). But just when I thought I was a prodigy, I got hit in the face by some OS scripting problem on Linux (due to PTSD, I rather not remember what it was haha). Programming logic at OS level (C modules on linux) is a WHOLE WORLD different to what I do at work as a Web Developer. From there, I started to try to understand the "low level" under the iceberg but that led me to needing to understand (and I mean understanding, not what I used to do back at school :P) math (discrete math, algebra, calculus, numerical analysis) and electromagnetism and optics (physics). At the end I ended up here, talking about electronics which is the basic form of modern computing. So, if you look at it, I did everything that was needed but in the opposite way (Frameworks -> high level programming languages -> operating systems -> electronics -> basic sciences). The way I did is a very good way of keeping interest on the students, but a terrible way to actually understanding stuff. So, you should go step by step of technology (Basic Sciences -> Electronics -> Computers -> Low level Programming languages -> Operating Systems -> High level programming languages -> Modern Technologies). Commander X16 as a project or even as a whole subject to study on the first courses (if not the first one). Keeping your first computer is a literal milestone hahaha

True. I don't know why I had it in my mind that it was 1uF, which would have been in the upper end for a ceramic one. So I guess the real question is why not use an electrolytic capacitor then. Were they significantly more expensive?

No idea about price differences back at that time. In that application I wouldn't have thought performance differences would be a factor. Tants will tend to be physically smaller, but there doesn't appear to be a space problem - could always have used an axial elec. if there were height concerns. Really easy to overlook it sat alongside all those resin dipped ceramics though (and the schematic not actually putting any units next to the value doesn't really draw attention to it either - it *could* have been 4.7nF - although that wouldn't make much sense in parallel with a 100nF decoupling cap).

As long as that's a good kind of hooked, it's all good 😃

Yes, I hear on PCs are mostly yellow. Not sure why, but in most of the computers I work with they tend to be blue and my eyes are trained to look for them (especially because they short with time).

Ugh. That's bad luck. And the worst thing is that resistance checks wouldn't work because all decoupling capacitors would show up as shorted. Did you have to desolder them all, or did you track it down with sensitive voltage measurements?

Thank you and welcome on board! That was one of the first retro t-shirts I got many years ago. I think I got it randomly searching one of those t-shirt sites like Spreadshirt or something like that.

That's the next step: Enhanced replica boards. This particular one is an exact copy, so I don't think it can be easily changed.

@@NoelsRetroLab I don't know if you read my previous replies, but I was talking BS. The 6510's 16 bit address bus can only address 64kB at a time. So if you want to use more memory than 64kB at a time, you will need to add a separate bank of 64kB and a bank switching circuit which will need a dedicated CPU line. It will probably also introduce latency on the memory and require a lot more memory read and write cycles.

@@paulmichaelfreedman8334 And completely customized software to use it. You can't just add bank switching in some transparent fashion.

@@stargazer7644 indeed. Probably the reason why nobody bothered in the first place.

My pleasure!

That makes sense. Is there a reason to use tantalum over electrolytic in that case, or is it a perfectly fine substitution in that case? And yes, I believe the boards I have are designed by you now that I think about it! 😃 I've had them for a while, but I'm hoping to do the build sometime soon. Thanks for making that project. It looked amazing!

@@NoelsRetroLab : Lower internal resistance & inductance, which are usually both good.

@@NoelsRetroLab tantalum have better filtering characteristics at high KHz to low MHz frequencies - ie exaclty the frequency area that the buses would be operating at. Ie better at filtering any noise induced on the supply by normal chip operation.

Not in this case. The problem was subtle and malicious.

@@NoelsRetroLab I watched the rest of your video and man, that was interesting to say the least. You've got me excited for commodores and im hoping to build one just to play around with it. Good content man!

You're welcome. The CIA problem runs much deeper than that. So much so that it deserved a whole episode to itself. Stay tuned.

Right!! I think it was the 7805 that saved me because it shut down as soon as it started getting hot.

Yeah, I suspect it was on a countdown to detonation. Tants do not like to be abused.

Right! Working on a SID-replacement comparison episode at the moment. Coming soon! 😃

@@NoelsRetroLab Yeah, I saw a recent video with an FPGA SID playing Lightforce tune and it sounded OK at the bit near the end where most emulations fail. Not sure on the state of VIC-II direct replacements. CIAs are less important for now and most of the other chips are no problem as far as I know. Of course you have to recreate the exact smell of a brand new breadbin C64 from 1983....maybe one day I will do a video on what readily available item will do that :)

You may be right about that. I didn't know that but other people pointed that out. The only thing I can think of is that the 7805 shut down faster than the cap was damaged. Alternatively maybe the cap was damaged but since it's just a decoupling cap, we can't really tell anything is wrong. Probably safer to replace it though.

That's another problem with tantalum caps: They seem to short with time. Not a fan of them.

@@NoelsRetroLab I read here and there that tantalum caps should stay stable longer than ceramic caps, so I guess it's down to quality. Commodore (among many) wasn't above buying inferior components to cut cost.

@@paulmichaelfreedman8334 : Perhaps they stay more stable in regards to e.g. temperature, but past a decade or two I'd expect the tantalum to be worse, because they're a type of electrolytic cap that lacks the self-heal capability of aluminum electrolytics.

The CIAs are both reported as faulty, independently of which ones I put there (it just takes another pass for the second one to be reported). The cassette part... I did show up as faulty and that just went away. I thought it was because I cleaned up the edge connector, but maybe you're right and it just wasn't pushed in all the way.

@@NoelsRetroLab I think it is a remote possibility, :) and maybe you will find the right issue in the next episode!!!! :)