Hi,
We all know about it....a small power supply inside a huge product fails...and the entire product gets de-installed and scrapped...totally replaced with a new one.
Just because some small component failed. No attempt made to replace the small component.
Dave Jones had a good one on here of a large product which had failed , and it was just the mains PSU that had failed, ill see if i can fish it out.
Dave also shows some monitors having been dumped in his "dumpster dive", and it was just the electrolytic caps that had failed...Dave replaced them and got them back working.
Dumped Monitors with failed caps..
http://www.eevblog.org/video/EEVblog347-BadCapMonitorRepair-480x270.m4vAnyway, ill start it off if i may, with a....
1.... Large 3 kiloWatt electro-plating power supply, these often failed , and the failure was simply a cooling fan power supply that used a under-rated linear reg. Entire product de-installed and replaced when it failed.
Got any others?...anything bigger?
well sometimes its because they want reliability and if that broke no one can tell them if its good or not after a repair (or they just don't believe it).
if its considered mission critical for someone with money they might think its liable to do something again due to cascade damage.
and less delay time to buy new. The money firehose principle.
I see it happen the most with TV/Computer because someone will definitely want to kill someone if the TV goes out during a international presentation. Or test equipment that is used on a tight standardized test schedule that has a big impact. Any delay or deviation (its often measuring something 'easy') will be like a scandal. The lab budget shoot down broken equipment at the same rate of a CWIS system in this case. They even want it in a ejection system into the dumpster because god forbid someone uses it after it gets a 'service needed' sticker on it. This happens with shared labs. They just know someone is gonna come and measure something their not supposed to measure to expedite the guarantee of something or another unofficially and then basically drive company into a ditch and its gonna go too far up the chain of command, implicate too many people to even take action against it internally. Kind of like if Mr Weyland or Dick Jones told you its OK to do something and you use broken stuff but you know their kinda cool with it but it fucks up anyway and you got protection but its wrong so people end up doing some crazy 1400's British admiralty shit to try to make it look good and it just ends up costing alot of money and no one can really do anything about it and some how its way too authorized.
well sometimes its because they want reliability and if that broke no one can tell them if its good or not after a repair (or they just don't believe it).
if its considered mission critical for someone with money they might think its liable to do something again due to cascade damage.
and less delay time to buy new. The money firehose principle.
I see it happen the most with TV/Computer because someone will definitely want to kill someone if the TV goes out during a international presentation. Or test equipment that is used on a tight standardized test schedule that has a big impact. Any delay or deviation (its often measuring something 'easy') will be like a scandal. The lab budget shoot down broken equipment at the same rate of a CWIS system in this case. They even want it in a ejection system into the dumpster because god forbid someone uses it after it gets a 'service needed' sticker on it. This happens with shared labs. They just know someone is gonna come and measure something their not supposed to measure to expedite the guarantee of something or another unofficially and then basically drive company into a ditch and its gonna go too far up the chain of command, implicate too many people to even take action against it internally. Kind of like if Mr Weyland or Dick Jones told you its OK to do something and you use broken stuff but you know their kinda cool with it but it fucks up anyway and you got protection but its wrong so people end up doing some crazy 1400's British admiralty shit to try to make it look good and it just ends up costing alot of money and no one can really do anything about it and some how its way too authorized.
Makes a lot of good points here. Prevenative repair practices (think right to repair) that started 20-30 years ago play a huge role in this. Because they help to foster this 'throw away, buy new' practice. It can be detrimental to company budgets, especially when they are a small company trying to grow. An example is the company I work for now. I am an in-house Lab Instrument Engineer for a biotech company - I maintain all their lab equipment. Before I started there, the cost to maintain their equipment was easily six figures due to either outsourcing the maintenance or just 'throw away, buy new'. I've brought that cost down 90% some years. If manufacturers of test equipment actually provided the service manuals like they use to, customers wouldn't have to worry so much about whether or not the equipment would fail again because it wouldn't be a black box - they would actually understand how it works. You've got lab equipment manufacturers selling this complex lab equipment to companies for hundreds of thousands of $$$ and won't even provide a service manual with a block diagram of the internal circuitry. It's ridiculous!
its good in theory but the fact is in some company setups that basically make YOU the voucher. There is no one offering confidence in the process besides you. So if it goes wrong it fall on you. Some environments that is good enough but other are all intrinsically based on the blame game for problems... i.e. legal. Even legal is reasonable ,to a extent, but IMO as soon as it starts to look 'complicated'... good luck. They are at a disadvantage if they don't have some major manufacturer 100% behind them, at least for some views.
BUT IMO that also makes yourself look incompetent if you need the big test equipment company to verify everything for you.
at some point its kind of like a nanny situation. can't pour milk without the nanny. starts to feel like being some posh sheltered royalty that is not allowed to use a butter knife... like 1000bc emperors court.
It is often a small thing that actually goes wrong, but it can take quite some time to find the actual defect. It takes experiance and some luck to find the defect and there is a chance spending 1 hour on looking for a simple fix and still not find it or only find out that there is something really bad broken. With bigger units there is also the problem that one can hardly take all the tools and little spares on site. So the more usual way is to have a technician comming out with only a few modules or a complete replacement. Depending on the parts they may actually repair the broken units in a specialized shop.
I remember at the university an HP workstation (AFAIR some 7000 series and no longer cutting edge) went bad still in waranty. An HP technician came out an replaced the motherboard.
A few months later we had a similar defect (no network connection) - but out of warranty. It turn out to be a blown fuse for the AUI network interface. With a new fuse and different connection on the AUI output the computer worked for at least the next 2 years.
JayCar Electronics had a warehouse clearout sale here in New Zealand last year, and if you've seen Dave's videos about JayCar warehouse sales, you'll know what it was like: huge amounts of unsold/faulty/returned goods on sale for cheap. I was in early and got hold of a pallet of mostly 1kW sine wave inverters for only a couple of hundred bucks - all marked as faulty of course, presumably customer returns.
Looking through them yielded a few quick insights. On one unit, the return note suggested that someone had tried to use it to start a big pump and returned it because it hadn't done the job. There was nothing wrong with the inverter at all.
However the most remarkable thing was that probably 70-80% of the units had one particular fault: the mains output socket was broken so you couldn't plug anything in. It was just a bit of broken plastic, so the part that holds the contacts had fallen away from the front faceplate. Obviously a weak part, some sort of design/manufacturing error. So tens of inverters got binned, each worth hundreds of $$ - and the only fault was a $5 output socket being broken, the inverters themselves were still working! Many of the faulty units looked brand new, still in the boxes with all their accessories, so they must have been dead on arrival or just a bad batch. What surprises me about this case is not just the cheapness of the failed part, but the number of failed units, and the simplicity of the fault, and thus how easy it is to diagnose and repair. You literally don't even have to open the units to work out what is going on. One diagnosis and one box of replacement parts would've enabled the straightforward repair of thousands of dollars worth of stock. It's possible someone did realise this but liability or company policy got in the way of fixing and re-selling them. Either way it's kind of sad.
And it makes me feel like such a scalper to make a profit from this!
But the alternative is them going to landfill, and nobody wins in that case.
I remember at the university an HP workstation (AFAIR some 7000 series and no longer cutting edge) went bad still in waranty. An HP technician came out an replaced the motherboard.
A few months later we had a similar defect (no network connection) - but out of warranty. It turn out to be a blown fuse for the AUI network interface. With a new fuse and different connection on the AUI output the computer worked for at least the next 2 years.
Speaking of university, here in the US, the higher education institutions are nortorius for throwing away or auctioning off very expensive lab equipment for the simpliest of issues. Which is surprising because you'd think someone at these places would know how to fix them! A colleague of mine that I source lab equipment from has dozens of stories of equipment that universities and colleges spent tens, hundreds of thousands of dollars on that they've gotten rid of for issues as dumb as what RLP describes above!! And he'll buy it for pennies on the dollar and make a profit. There's your tax dollars at work!!
However the most remarkable thing was that probably 70-80% of the units had one particular fault: the mains output socket was broken so you couldn't plug anything in. It was just a bit of broken plastic, so the part that holds the contacts had fallen away from the front faceplate. Obviously a weak part, some sort of design/manufacturing error. So tens of inverters got binned, each worth hundreds of $$ - and the only fault was a $5 output socket being broken, the inverters themselves were still working! Many of the faulty units looked brand new, still in the boxes with all their accessories, so they must have been dead on arrival or just a bad batch. What surprises me about this case is not just the cheapness of the failed part, but the number of failed units, and the simplicity of the fault, and thus how easy it is to diagnose and repair. You literally don't even have to open the units to work out what is going on. One diagnosis and one box of replacement parts would've enabled the straightforward repair of thousands of dollars worth of stock. It's possible someone did realise this but liability or company policy got in the way of fixing and re-selling them. Either way it's kind of sad.
^^This just makes me sad and annoyed. These would have been so easy for their previous owners to fix. This is one of the reasons why our planet is doomed.
It's not quite the same, but a medical friend of mine told me that he had just had to scrap a 3 year old, £150k antenna coil for an MRI scanner. The reason for this accessory to be beyond econoomical repair: the plastic case had cracked, and seals had failed resulting in contamination of the internal electronics.
The price was explicable by this being one of these new fangled "digital multichannel" coils - so, not just one receiver element and LNA, but 64 active elements, with a full 64 channel coherent 150 MSPS/16 bit SDR and high speed optical interconnect.
The reason for the failure was rather more surprising - the materials used for the casing and seals are not compatible with hospital disinfectants and detergants!
Sign of the times. Skilled electronics engineers are almost a thing of the past. Diagnostics down to component level is fairly rare now. Used to be commonplace in the 80’s and 90’s
The reason for the failure was rather more surprising - the materials used for the casing and seals are not compatible with hospital disinfectants and detergents!
Surely that would fall under Fit For Purpose laws, if any?
"Big products that get de-installed due to a tiny internal component going wrong?"
Sometimes that happens with people, too.
If a very large device of the order of several million dollars replacement cost fails, a replacement will not only cost that amount, but will entail "sitting on your hands" for however long the replacement takes to travel from the other side of the world, plus being installed, which will also entail large costs, & the new one may not always work on arrival.
In that case, a repair is obviously cheaper & faster.
If it is, say a TV transmitter, try explaining to the station manager that being off air for a month is ok, because full replacement is "the only way to go".
The reason for the failure was rather more surprising - the materials used for the casing and seals are not compatible with hospital disinfectants and detergents!
Surely that would fall under Fit For Purpose laws, if any?
It wasn't incompatible with
all detergents and disinfectants, just the common and preferred ones. This was all documented in the small print in the manual, which someone either didn't read or didn't realise the significance of.