It depends on a lot of things. If the root cause for the failure was poor factory soldering, a new board will likely fail in the same way after similar use where as a field repair may solve the issue.

There are plenty of other root causes where many components were overstressed and replacing them all with a new board would result in less downtime than replacing components as they fail, too. Of course, sometimes you can do well with replace this group of components anytime any one of them fails.

Also, availability of full boards and availability of individual components makea a difference.

Oh, the sweet innocence of someone who never looked inside electronics around him. Probably 1/10 of the products you ever owned shipped with some kind of factory bodge or went thru rework at the factory. That includes silly things like alarm clocks, VCR/DVD players, but especially expensive ($xx-xxxK) scientific equipment, airplane avionics, industrial control and medical devices.

1960s mainframes were “field serviced” for processor errata by going out to every installation and hand-soldering some pre-planned bodge wires onto the boards, according to very detailed patch schematics.

And their backplanes were breadboards!

For anyone who wants to see what I'm talking about, here (https://www.youtube.com/watch?v=mBssIIRGkOw) is a good video of a 1970s minicomputer (which is not a mainframe, yes, but generally built and maintained by the same vendors with the same repair strategies) being disassembled/reassembled. You can see plenty of bodge wires!

You can also see the 1970s evolution of the "breadboard blackplane" into "pseudo-ball-grid-array backplane, with arbitrary collections of bodge wires loosely mated to it via hand-crimped proto-Molex connectors." (The only difference in practice being that you could now plug/unplug a "group" of wires, rather than one at a time. There was still no screen-printed indication of where on the BGA grid to re-attach anything, since the grid wasn't customized to its purpose, but rather was still fundamentally a "generic routing board" in the same vein as a breadboard.)

Possibly, but the component level repair could be a good triage solution while awaiting that replacement board. Not a soldier, but I imagine in warzones and combat areas a good enough solution that works now is more important than awaiting the "right" solution that takes 3 days.

>When I was in Iraq (08-09), we had a juniper firewall go down, didn't have another on back up, and they had to fly a contractor out from the states to Iraq, to change it out. Took 10 minutes to diagnose the problem, and 3 days of waiting, and 15 minutes to install it. We lost feed to the predator drones in the area, phone, internet, coms, and several other mission critical things for 3 days.... a real s** show.

This is highly unlikely, Firstly military comms equipment and drones use their own proprietary comms equipment The coms equipment is typically a combination of (AM/FM/TacSat) it wouldn't be running through a firewall that handles phone and internet (which is almost always handled by a separate contractor even in 08)

I'm not going to say it didn't happen, because its the military, but that is super uncommon, and against so much info sec, it just doesn't stand to reason.

Much of the communications equipment have complex counter measures built in, and sometimes just really simplistic ones that if you say open the case without unscrewing the screws in the right order, they are designed to fry themselves to prevent tampering.

This isn't something you really want to try to do in 120 degree heat, hanging out the side of a vehicle in a very sandy environment.

Historically, "frying themselves" was also meant in a very literal sense, with thermite or explosive charges.

Every so often, something would make it into the post-WWII surplus market with anti-tampering countermeasures still intact. Hilarity seldom ensued.

Often stuff has to be repaired for training and its not equipment that would ever be used in combat. It would not be an issue if it fails again in fact it would be useful for training if it does. You dont have to simulate failure and people learn if their repair is actually lasting.

I did train soldier on technical equipment (not in the US) and we would have broken or known unreliable parts given out so they learn how to find it. Also we had no authorization to repair most stuff and sometimes it was hard to keep broken parts to use the for training because these parts where scheduled for maintenance witch would mean the manufacturer of the part would replace it and turning the perfectly "good" part for training into e-waste. It's the classic problem in hierarchical structure. The top just never gets what the people at The bottom really want and need. They cant warp their head around why a broken part could be of more value than having 100% of the replacement parts functional all the time.

It depends so much on what it is, a simple single component swap seems unlikely to cause issues.

Sure I understand that a full diagnosis/repair of say a water damaged board I might not trust, but if it's just a port or a capacitor then it seems unlikely to cause massive issues if the person has any EE background whatsoever.

In a combat situation, having a thing down now may be worse than an increased likelihood to fail later. You can always order a replacement, but what are you going to do in the mean time?