From Newsgroup: sci.electronics.design
Don Y <
blockedofcourse@foo.invalid> wrote:
On 8/4/2026 7:19 PM, Waldek Hebisch wrote:
My first product had a 6x10" "CPU board" -- CPU, decoding logic,
ROM, RAM, connectors to I/O. I could patch the UVEPROMs soas not
to have to replace them. Or, replace them with OTPs. Ditto the
RAM, CPU, decoding logic, etc.
Nowadays, all of that (except the I/O connector -- which would now
be the pins on a device!) would be an SoC. Is that more repairable
or less? Clearly, if the FLASH in the SoC craps out, the entire
"CPU board" is scrap.
If SoC dies, then it is scrap. But unless you are a chip manufacturer
your product is likely to contain more than just a chip.
That is increasingly becoming the case. Full customs are now
pretty easy to create without bankrupting a project (and,
for high volume products, a great way to get costs and size
to a particular target)
But, the SoC would be smaller and likely less expensive -- as it has
shed all of the extra "matter" that was present in the "CPU board"
I do not think anybody would protest integrating several chips
into one.
Do you know have to make that *component* available for replacement?
And, document it enough that a user could determine if it was
failed or not?
Available for replacement? Likely. Concerning documentation,
if all device functions are inside a SOC you can probably describe
pinout so that user or repair people can monitor signals on
the pins and infer "no expected signal => faulty SOC".
Clearly "I do not want unauthrised people
mess with my design" is not a valid reason. Similarly
I'm publishing the source code. The goal is for others to
mimic and *improve* the design. But, not by making it *more*
expensive to produce! Or, physically larger than necessary.
[You could make ear buds the size of a pack of cigarettes
(with disposable WIRED "earplugs") to ensure the electronics
and battery were repairable. And, rely entirely on OTS components
so there is no manufacturer lockin. Do you think such an approach
would fly?]
"repair is not viable economically so we will not support it".
What do you think the policy is on repairing a set of $35 earbuds?
Do they even want the "defective" units returned (warranty)??
I do not know what is current policy. But clearly there is some
pressure to allow long usage. Which could be attained by
lowering defect rate and in general by maing it more durable.
But that is hard to codify as a law.
The issue is WHO gets to decide how a product is designed,
marketed and supported. If a manufacturer (likely infinitely
more adept at understanding the issues related to manufacturing
vs. a bureaucrat/legislator) decides that a particular course
of action is best for it, should it be free to pursue that
approach, letting The Market vote on its appropriateness?
Or, should constraints be IMPOSED and the manufacturer (AND
CUSTOMER) be forced to live with those constraints.
You assume working market. But advanced designs tend to have
unique features. For mass market products there are scale
effect. Consequently, provider either have monopoly or
near monopoly position and that leads to highly distorted
market. Also, "good for manufacturer" may be quite different
than "good for economy". You may look at old thing to understand:
"easy" way to get some income is to charge tolls on roads.
In medieval times this was extremaly widespread in Europe.
Various local authorities had incentive to keep tolls close
to pain level: lowering toll in one place would only moderately
increase trafic leading to lower income for this place. Only
when modern nation-states started to regulate (or if you prefer
to say de regulate) the economy internal tolls were abandoned and
that was significant factor to increasing economic activity.
How many people do you know who have had their phones repaired,
over the years (including maintenance things like replacing battery)? Contrast that to the number who have just opted to replace the
phone THAT IS STILL OPERATIONAL?
[Hint, I have a collection of phones, all of which work as originally designed but were *discarded* by their previous owners]
Surely, there is reason to impose constraints on the design of
such ubiquitous products -- if only to keep them out of
landfills and make their precious metals more recoverable.
How many 60-100 inch TVs get scrapped -- because of a hardware
failure on a 25 square inch PCB inside?
How many monitors are scrapped because of a failed inverter board?
I do not know. From what I have heard typical problems are failing
fluorescent lamps and LED-s.
And even if _you_ consider your unit to be unrepairable you
may still be obliged to provide spare parts or other support
to parties capable of repairing your device.
That is the nature of the question. What LEGALLY are the
REALISTIC obligations? If the factory policy is to discard
defective product ("no replaceable parts inside"), then can
legislation REALISTICALLY require the manufacturer to support
repair by end users?
Yes. The start point was manufacturers policy that people
deemed to be unreasonable.
Manufacturers, like all creatures, aim to protect themselves.
They've spent money creating a product; what incentive to
make it easy for others to copy or service THEIR product?
The *customer* should dictate what they consider to be essential
(or desirable) in a product offering. If you are NOT likely to
have your device repaired, how happy would you be to discover you
paid more for it, initially, so that it *could* be repaired?
The phone that I presently use has a removable battery -- I can
swap it out in seconds. Yet, this is NOT the norm. Obviously,
The Market decided that this capability was not worth its cost
(connector, removable door, physical constraints on the design, etc.).
That is nice example: at point of making purchase typical customer
does not know if battery is removable. That becomes an issue only
few years later at which point there may be _no_ phones which
are acceptable to the customer which have removable battery.
The user/customer can always find a product from another party
that is "closer" to his desires.
Maybe if you have a lot of money. But if you want "nonstandard"
feature expect to pay 2-10 times as much (and possibly much more)
compared to typical item.
There likely will be consequences and
tradeoffs involved -- but, the customer is best qualified to make those decisions.
Nowadays, it is relatively easy to share experiences you[ve had
with products, manufacturers, etc. Whereas manufacturers have to
PAY for advertising, word-of-mouth can be an effective counteragent,
often having an outsized place in a new purchase decision.
There is a lot of products and tiny number of people that can
competently judge those products. And many things get known only
after delay. Various experiences in many cases only shows users
lack of knowledge about products. And there are paid influeners,
who pretend to be neutral, but have monetary interest in promoting
specific opinions.
E.g., when I see folks looking at the refrigerator that we purchased,
I *offer* the observation that the "skin" dents easily. And, can
almost always find such a dent on whatever floor model is in front of
them. This, my way of putting pressure (indirectly) on the
manufacturer's offerings, going forward.
[Note that I make no claim as to the RELATIVE ability to dent it vs. other manufacturer's offerings. But, likely cause the potential customers
to "notice" dents in other floor models!]
Should the manufacturer have to provide
proprietary replacement parts "at cost" (?) to folks wanting
to do this?
IIUC that is in proposals.
Should manufacturers be required to sell PRODUCTS at cost?
Well, I take "at cost" to be just a shothand. Clearly, without
price limit offer of spare parts is meaningless. Consider manufacturer
who do not want to keep any spares. Technical people may find out
that 6 months is enough to get production line for given part
running, and accountats could calculetate that cost of setting up
production line and getting single run of parts is say 5.0e6 US$.
So, manufacturer could say: parts are available, just pay as
5.0e6 US$ per part. Of course, for product that costs say 1000 US$
this offer not different than providing no spare parts.
And, shipped by the fastest/cheapest means (lest
the user be inconvenienced)?
Shipping is in a sense a trivial matter, there is well developed
delivery system and I do not think law would require you to
better than available means. OTOH if you try obstruction there
it would be visible and easy to catch.
Who sets the standard. I ordered an RTC replacement battery from
a place in an adjoining state. It took almost two weeks to arrive.
Is it OK for me to deliver the replacement parts for your device
on a similar schedule? Should I be 8required* to offer expedited
shipping? Aren't you then legislating winners and losers in
the markets on which *I* rely?
Two weeks looks long. In Europe I think that norm is delivery
within 24 hours from shipping. If you order on Friday evening,
then shipping can happen on next Monday. More delay usually
indicates that vendor do not have the product (or is hopelessly
badly organised). OTOH, depending on the product two week
wait for spare part may be reasonable. FYI, last time when I
had my car repaired repair shop did not have spare parts
on the spot. They ordered them from their vendor and had them
delivered in few hours (actually, I think that this is a norm
but fortunatly I have very little contact with car repair shops).
There is no need to have very detailed regulation. Lawyers
use phrases like "vendor should ensure timely delivery using
customary delivery mechanizm", with meaning that as long as
vendor do normal things it is OK, but negligance or willful
obstruction is punishable.
I can imagine provision that repair centers could order replacement
parts in advance and possibly in quantity bigger than 1.
But, those would be NON-factory repair centers.
Sure.
Because the
factory doesn't support repair of that product.
(One can always choose not to sell into markets that impose
criteria that conflict with your approach!)
Concerning future, I can imagine system simiar to current
carbon credits, where you pay for any unrepairable part
of your design. That may be coupled with strict repairability
mandate for elements considerd part of infrastructure.
Having first-hand experience with recycling "in the wild",
a huge chunk of every (electronic) product is recycled by
grinding it up and chemically extracting the precious metals.
The low hanging fruit are things like copper and aluminum heatsinks,
the aluminum frames of disk drives, lead from VRLA batteries, etc.
Gold from edge connectors or CPU pins. Steel/tin are worthless (1p/pound). >>>
So, if you *add* materials to a design to make it serviceable,
you are just increasing the amount of non-recyclable material
that taints the product when it *is*, eventually, "retired".
Such recycling clearly discards most of utility that thing had
before. If you look at publicised cases you will see electronics
that is part of expensive products that customers expected
to function for long time. In particular as long as electronics
works there is no need to replace it,
That's not true. Witness cell phones, laptops, desktops, etc.
One publicised case is tractors. New they seem to have six digit
prices in US$ (maybe some in high 5 digit range). A farmer could
expect them to work for say 50 years. Newer models may be more
comfortable, easier to use. They may offer better fuel economy
and some extra functions. Still, 50 year old tractor can perform
its work. But newer ones have computer control, which when it does
not work effectively disables the tractor. Manufactur tried to use
copyright on the software to block independent repairs. And that led
to legal battles.
In other thead you write about ovens. Did you consider that your
oven works "as intended" by the manufacturer? That is manufacturer
wants to either charge large service fee or possibly encourage
you to replace the oven by a new one.
People replace things because they can. My neighbor just bought a new PC ("What's wrong with the old one?") Each of the big screen PCs that I've rescued was likely replaced because they didn't have room for *it* AND
an even bigger one in their home!
[I think few people realize just how much "stuff" gets discarded
that is still operational. I should start taking photos...]
The thirty three 4T SAS drives that I recently rescued were discarded
after just *5* power on hours! No idea how many hours the 48 drive
DAS that I pulled them from had on it! That;s a "business decision"
that obviously made sense to SOME business owner!
while failure creates need
to repair or replace it. If you add material to make it more
repairable, then likely effect is longer useful life, so less
Only if the additions:
- don't adversely affect reliability
- are exploited by someone WILLING to incur the cost of a repair
E.g., I solder down the memory in my designs instead of adding
a socket. This means the user can't repair, replace or upgrade
the memory (obviously, not a very user-friendly design decision, eh?)
I probably can not repair that (I succesfully handled 20 pin SMD
parts, but to the moment did not succeed with 48 pin one). But
repair shops succesfully replace chips with many pins, so that
alone does not contradict possibility of repair.
I would prefer device with user upgradeable memory, but if price
is adequate and installed memory is adequate, then I have no
trouble with such device. Trouble begins when vendor takes
advantage of technical difficulties of users and third parties
to play monopoly tricks.
Yet, now don't have to rely on the cost of the socket, labor
to install it (and populate it) *or* the ignorance of the user
who thinks sockets can be used repeatedly without consequences
(and anything that physically FITS "should work!")
waste overall. There are now voices which demand longer useful
life from ordinary "consumer" products. It is not clear how
That doesn't mean that "repairability" is the solution.
If you want to encourage longer life (discourage replacement),
then legislate with THAT goal in mind and let the manufacturer
sort out how he is going to meet that goal.
I designed a product many years ago that had a stated 30 year
expected, "service free" lifespan. The costs of servicing
(or replacement) were probably 100X the cost of the device.
This drove HOW the product was designed as you had to consider
the impact of every design decision on that aspect of the design.
(Of course, we were designing for quantities of 30 million units so
you have extra opportunities available to your design process)
far this will go, opponents of regulations in this direction
are quite strong. One possiblity is to have long warranty
period possibly coupled with mandatory insurance to cover
cost of repairs in case of manufactur default.
That still doesn't address the user who willingly "discards"
the device in favor of a new version. Do you punish him by
applying a "deposit" -- that is not redeemable for N years?
Or, ban him from making new purchases for that period?
Electronic devices sold in EU include cost of disposal in their
price. I could imagine that consumer who unexpectedly early
discards some device could still be requested to pay disposal fee.
OTOH EU prefer to regulate manufacturing and supply side and
let consumers to do silly things if they really want to.
In slightly different spirit, a lot of people in EU feel "squized",
they pay more for neccessery things and their income does not
increase. In many cases they would welcome opportunity for
longer use of various devices. And they are more willing to
use devices that were previously used by somebody elese.
Are you acting as his advocate -- protecting his "investment"
(despite limiting his ability to take advantage of newer offerings)?
Or, are you trying to advocate for society by reducing waste?
Electronic devices tend to change quickly. E.g., in 13 years,
there have been effectively *8* different versions of the
Nest thermostat. You've probably replaced your cell phone
two or three times. A laptop, once or twice. Car, similarly.
Roof, once. Major appliances, less than once.
I have my cell phone more than 20 years. I have a new desktop
now, but for previous 16 years I used the same desktop. I have
third laptop now. The first one I used about 10 years, but
the touchpad failed and later power jack had problem (work in some
position, does not work in different position). I looked inside
and power jack is surrounded by (IIRC) video curcuitry with its
screen and heat pipes. It looks that without special equipement
it is impossible to access power jack without demaging critical
components. Seems that designers spent some effort to make it
unrepariable. OTOH it was not good job as it worked way past
warranty period. The second one was a brilliant design: it failed
mechanicaly close to end of warranty period. And manufacturer
warranty does not cover mechanical failures. The third one
is probably at end of its useful life: it developed mechanical
problem which does not prevent use, but will likely lead to
complete failure later and seem to have problem with power (I
did not investigate deeper, so I do not know what is really
happening).
If not the failures I would probably still use the first laptop.
This is a desirable trait of electronic products -- there is
continual improvement. People seem to WANT that in those offerings
(even when they complain about it!)
I think that a lot of peoples buy new devices because old "does
not work". It may be just failure of electronics. It may be
sofware that needs upgrade to newer standards. It may be dependence
on discontinued service from the vendor.
There are devices for which progress is in nonessential features.
For example audio in late 70-s reached level that ordinary people
wound not hear defect in the equipement. One can say that newer
devices are better, but it makes almost no difference for
odinary users. Modern LCD screens offer higher resolution, less
flicker, lower power use than old CRT-s. They are lighter and
almost flat. But for purpose of viewing TV or recoded films old
CRT-s were quite adequate. Basically the main rational reason
to replace them is power efficiency. And, AFAICS 10 year old
LCD TV set should be essentially as good as a new one, except for
possibly planned obsolescence, like discontinuation of services
offered with old TV, or changes to video encoding.
I see improvement in computers, but it is not as fast as it used
to be: in 16 years from my previous desktop to the current one
single core performance grow about 3 times and the new one has
3 times as many cores as the previous one. In the past one would
expect similar improvement in say 6 years. And to say the truth:
for ordinary use performance of this 16 year old machine should
be excellent. Recent AI mania may change this, LLM-s (and some
earlier methods) need insane compute power
How much has to be "discarded" to facilitate that is debatable.
But, anything that requires labor will have a distorted sense
of price/value.
E.g., the true "savings" in m current products comes in the act
of installation as that's on-site labor. Do you have someone
come to change the filters in your HVAC system each month?
replace a $10 filter for the cost of a $100 service visit?
[I'm sure there is some way of CLEANING filters and reusing
them -- but, it is impractical when the cost of the filter
itself is minimal. Likewise, if you're going to "spend" a
considerable effort to access/remove/replace a device, how
much more do you think you are going to save by *repairing*
it?]
--
Waldek Hebisch
--- Synchronet 3.22a-Linux NewsLink 1.2