Those look like standard 2.54mm pin headers. Super cheap and just cut them to length.
Soldering the male dupont wires will also work, but I'd install headers for flexibility.

I was mostly curious about methods that don't involve soldering, particularly if it could be as simple as just sticking male end of jumper into a hole. Considering what was written about doing the same with the pin headers that are meant to be soldered, I suspected those wires by themselves might have the same problem. But there was a hint posted it could work:

Though it was followed by the definitely surer method - press fit headers that don't require soldering:

Source: https://electronics.stackexchange.com/questio … older-for-proto

I got around asking at my workplace about it today, coworker from production department said one could do that and it may work, but it's not the surest method. For this experiment, it has to be sure so if it doesn't work out, it will be because hardware's done and not because of flimsy connection.

He also told me it's not a problem for him to solder pin headers on it so I don't have to hunt around for press-fit headers. Soldering is not one of my skills, plus doing it at home and needing equipment, just not worth the bother for one lone experiment with high probability of failure. There was a power outage recently and it doesn't boot since, all LEDs except Internet and DMZ just blink forever. That's with a good power supply that came with another device, the original presumably doesn't work properly anymore, with it, LEDs light up for cca. 5 seconds, then darkness. No life on LAN ports. Doesn't respond to usual tricks with power cycling and reset button.

Anyway, I'll see if it says anything over serial port, USB to RS232 TTL converter should do the job. I'm not so sure about JTAG method anymore. The only way that I'd bother would be with Raspberry Pi. The old method is crafting special JTAG cable (way over my head) or maybe getting it from across the ocean, again not worth the bother for me, and plugging to computer's LPT port (my main PC is old enough to have the header for it on the motherboard).

Getting Raspberry Pi just for this would not be worth it neither, but I see myself putting it to other uses. And here's the catch, for the latter, the latest and greatest Pi 5 would be preferable, but the one program and its fork for performing magic over JTAG only surely runs on up-to Pi 3 (https://github.com/oxplot/tjtag-pi). I've read about GPIO stuff and it's way over my head. I've done some patching of old games in the past, writing bits in C and x86 ASM, but this is totally alien to me. I suspect getting that program to run on latest Raspberry Pi is not as simple as changing hardcoded peripheral base address in the source code and compiling it.

So the only surest way would be getting otherwise less interesting Raspberry Pi 3. Sigh.

Yeah you need some kind of spring or wedge to make full contact, the fishing line is a cool idea, but you could wedge each pin in with a wooden toothpick also.

This one time at band camp... okay college... I had to connect a male serial cable to male port and managed a shaky connection by chopping the ends off a bunch of bic pen refills, rolling aluminum foil up and posting it inside, and using those as sleeves to make basically a double sided female socket. If you're trying to do male pin header to male hookup cable pins though, that are individually positionable, then bits of stripped wire insulation, slid over the pins with the other male forced in alongside it are often enough temporarily.

It's all clearer now, I found fishing line (wire) method described here, the additional wire involved that ensures good contact has totally slipped by me.

I also asked about it at work and the coworker in production agreed just putting the jumper wire in by itself is not the sure way. I didn't know we have a bunch of plain pin headers available, he said it was not a problem for him to solder them on my router's board. So both JTAG and serial ports are now ready to be connected to with the female end of the jumper wire.

Thanks for pointing out the woodpick method, will keep it in mind. Interesting experience regarding putting both male ends together.

I've been recently staring at different source code examples written in C dealing with those GPIO pins on Raspberry Pi. If I understood it right, it may not be as complicated as initially thought. At least I think tjtag-pi could be easily changed in a way that it will drive those pins through the WiringPi library, which recently got support for Raspberry Pi 5, rather than being hardcoded to work with specific hardware revision.

Changed title a bit to be more descriptive.

After trying another good power supply that has a LED that is normally lit when it's powered, it was blinking along with router's LEDs, I believe this means short circuit protection kicking in. I guess this means faulty capacitors in best case scenario and something else fried in worst case scenario.

So I suppose if one wanted to try anything in this case, soldering would be unavoidable. I wonder though, could multimeter be used to pinpoint anything, eg. this capacitor definitely doesn't work, without taking it off? I remember someone on YouTube showing measuring with both ESR and multimeter, when capacitor was still on-board, ESR is more suitable for in-circuit testing from what I've read. I don't have an ESR meter, but I have a multimeter that does both capacitance and resistance testing.

Do you have access to a thermal camera? With all those caps on the board the quickest way will be to inject 1V into the board and see what heats up.

I have a feeling the old Linksys power supply could be responsible for causing more damage than it would have occurred otherwise. That thing was rather hot during normal operation, probably not normal for 1A power supply, at least by modern standards, maybe that was "normal" back then. I thought about it few times, but, as it just worked until now, it didn't come into my awareness that maybe it would be a good idea to replace it.

So if I plug the original supply in, the lights stay lit for about 4 seconds, then they go off. Back then when it broke I must have tried power cycling about 2 - 4 times, last time I heard whining noise coming from the router that lasted about 2 seconds... Subsequently, I was only trying with the other known good power supplies, that apparently cut the power right away when they detected short circuit.

The old supply was cold when I got home, the power outage occurred about 2 hours earlier, lasted over half-hour.

No, I don't have a thermal camera at hand. These things are pricey!

Soldering equipment on the other hand can be acquired at much lower prices. Do you need anything expensive for simpler things like (de)soldering capacitors? It doesn't look too complicated on YouTube and who knows, the skill might come in handy in the future for another computing device. So if I got this right, you need the solderer and the wire, correct? And be careful to not burn the board (or yourself).

At least that's the only idea I have for ruling out the capacitors. I've got a bad feeling it could be something else, but you have to start somewhere, right? If it's something else, then well, at least I'll know, but if it's really just capacitors by any chance, maybe I can have a router working again to play with. There are also coils, these can apparently be tested, though no idea about replacement, and even beyond capacitors alone, there are chips on the board that are ancient and it's probably more complicated to get them on, even if I could get a replacement somehow...

Thoughts?

I recognize that board.
https://www.badcaps.net/forum/troubleshooting … bad-capxon-caps
From that thread, there's a chance that bad caps can damage the main ICs, but a good de-CapXon job might put it in a state where it will accept firmware. If those purple caps are bad enough, you could measure a drop in capacitance, but they tend to fail with good capacitance and high ESR.

I'm usually just using bent pieces of jumper wire for temporary through-hole connections. Simple and works reasonably well.

Sorry about potato quality 😜

Didn't see the replies, this forum has way too much traffic 🤣. I've put this off for a while since since I didn't feel confident about doing any of this myself, though I did rush to buy some equipment back then, including multimeter and (probably) way too cheap 40W soldering iron.

Today, I finally got around to trying something, the initial idea was to just get one capacitor off, the one near where the power goes in and measure it with multimeter. Well that turned out to be a failure. I got the capacitor off, but the legs are still stuck inside, sooo I probably can't do anything with that. Desoldering looked so much easier on YouTube, but in my case, just nothing melted or very slightly and all I did in the end was yanking the capacitor off with brute force with the result as written above. Well, it seems the heat still helped slightly loosening it.

Ugh, despite the initial enthusiasm and investments, it seems in these cases, I'm better off simply with the good old "throw it away when it dies" approach...

I probably shouldn't try to power the router now that one capacitor is off, huh? Just to see if there's still short circuit, but there could be risk of frying something further? I don't know, I've seen people write about their motherboards working with a missing capacitor. Was lti's suggestion in that direction, try to run the router without a single capacitor?

You don't want to run it without the cap. That will just make things worse. Since you mentioned the router making a whining noise, I was suggesting a recap. I must have missed the part where the power supply was shutting down.

Removing the individual leads is easier than keeping the cap intact. Cleaning the holes and soldering in a new cap won't be easy if your soldering iron is that weak, though. If you haven't soldered before, it takes some practice. Adding fresh solder will actually help with desoldering, and you can then use wick or a solder sucker to clear the holes. It needs a lot of heat, and with a cheap iron, there's not much time between successfully removing a component and damaging the board.

Newbie mistakes, for all I know, I might have already fried it further by attempting to get that cap off. Back then I didn't know already having solder at hand could help with desoldering, so I didn't buy it as I was going step by step.

I did a test with that capacitor I yanked out, used sewing needles to give it legs temporarily, then tried measuring capacitance with multimeter and I guess when I got (hopefully) good contact, it showed 0L, manual says that's an overload. I wonder if that was even accurate or there could be possibility of the needle to be touching another separate area the metal legs didn't which could give false reading, assuming that scenario is even possible.

So one thing is clear, I'm probably not going anywhere with that 40W Puctec. You'd think such a small device wouldn't be as tough. Perhaps it has to do with 21st century boards using lead-free solder, which is less toxic but more difficult to melt, or so I've read. Today I just found another video of the guy doing desoldering of capacitors with just 18W soldering iron, the brand was Weller (apparently a decent one), but still.

I only did a little bit of soldering back in school, which was half my life ago, cca. 15 years, I don't remember if we ever used it on electronics, what little electronics stuff was done on a prototype board, but at some point, we had to make a metal cube and I had trouble with making joints, they came apart too easily.