I remember watching this video a few months ago. As I write this I am debugging open source electronics I designed for our open source farming robot.
It seems like people see the growth in opaque proprietary electronics and they conclude that we should move back to old mechanical systems. But mechanical systems are not always more reliable.
What I really want more people to understand is that electronics are not the issue - proprietary designs are. Open source designs inherently have enough documentation that even if they don't come with some nice service manual, the nerds on the user forum can dig through the schematics and code to sort out what's what.
Unfortunately open source hardware just isn't getting enough coverage for people to really see the value in it, despite the entire internet being built on open source software. We need to find ways to communicate the power of open source beyond software. We could have open source modules in cars with open standards. Sure the manufacturers need a rugged PCB that can control lights, read sensors, interact with CAN BUS, etc. But there is no need for that system to be proprietary. We need to build and demonstrate the value of open source hardware to help people understand what is really possible. Because the call to move back to mechanical systems is really just a call to move back to systems that are intuitively understandable, and today's shadetree mechanics can read a forum post as well as anyone else. What we need are open, documented systems. Electronics are not the enemy, proprietary systems are.
You are right on openness. This is key to understand and fix broken machines. Mechanical designs used to be more open too. We bought tractors in the 1980's and these would come with a workshop manual with parts list and disassembly / assembly instructions, with tools list and torque requirement.
The key is that mechanical designs cannot be copyright locked and you cannot build a closed ecosystem and demand insane prices for access to that ecosystem.
I feel like open source isn’t the point, it’s the solution. The point is that proprietary lock in creates artificial scarcity and allows protectionist rackets, and open source AND all mechanical systems counter this ugly consumer deleterious product choice that is incentivized by the market effects of protectionist proprietary ethos.
Reproducibility of mechanical objects is only the case in the modern day! It used to be that reproducing mechanical parts was difficult enough that a particularly sophisticated part could be just as locked-down as a particularly sophisticated bit of software is today.
For an illustrative example, imagine that a part on your tractor suddenly becomes a bit of twisted, wrecked metal.
These days? You ask someone else with the same tractor to send you some pictures, pull their part out and measure all the holes and their positions and tell you what the bearing surfaces are made of, you scrape off some slivers of your wreck and assay them to figure out what alloy it's made out of and how it was heat-treated, you go and machine a test piece or two on your mill and try it, you buy a bit of bronze bushing off mcmaster and it works, etc. Maybe someone else has already done this and published a CAD model and instructions! We have solutions for this.
Back then? You have no designs. You have no way to communicate with other people who have a complete working part. You might be lucky enough to remember what bits the part connected to if you'd looked at it before. You may have some idea what the intended geometry was, but you're never going to figure out that the geometry is actually just a few degrees off right angles because that keeps you out of a complicated kinematic singularity that blows up the part every time you turn left. You can guess at some of the materials, but systematic classifications of iron-carbon alloys don't exist so all you can say is "it's some kind of steel". Even if you figure a lot of this out, there's no real way to write it down or share it with people because a lot of the terminology for representing this stuff simply doesn't exist. Even if you make a functional part, its expected lifetime is months rather than decades.
The original creator, by comparison, can just buy a new block of their supplier's "Alloy Number 5", throw the forging dies back on the presses and stamp out a blank, slap the blank into the jig that presents every hole to the drill press at exactly the right angle, put the part through their proprietary heat-treat process, and then ship you the result. Those dies and jigs and heat-treat protocols are exactly the same kind of proprietary protection that a git repo full of source code is today.
If you look at hydraulic and pneumatic fittings, you can see elements of this lock-in still extant in the modern world. Everyone had their own thread sizes and thread geometries for their own tools, getting them to interoperate was impossible unless you had the right adapters, and you could not make the right adapters unless you were a dedicated specialist in hydraulic fittings. You were locked in to your tools provider's pneumatic toolchain in exactly the same way that people are currently locked in to their current power tool brand because they own $50k in tools and battery packs that agree that they're made out of a specific manufacturer's 18650s with specific voltage drop and internal resistance and etc etc etc. (go watch the torque test channel's segment on battery adapters a bit, it's enlightening! - https://www.youtube.com/watch?v=WgJI8Ikrd6Y).
(if you watch enough of Forgotten Weapons, you'll also hear things like "X bought the design for Y gun but it took them 3 years to transfer the tooling and get it up and running" - that's the stamps and jigs and heat-treats and everything that I was talking about earlier, and if those tools were destroyed a gun generally became totally unmanufacturable and had to be redesigned from scratch to continue production, and those redesigns frequently failed to perfectly reproduce the original!)
You've never heard of tracing attachments for lathes, or Pantograph mills. Copying things and making parts to fit goes back well before the industrial revolution, it used to be the way all things mechanical were made.
If there's enough time and budget, anything, even crashed things from other worlds, can be repaired. I've seen replacements made for the bull gears that hold up Bascule bridges in Chicago, and I've repaired atomic clocks... people improvise, adapt and overcome issues like the odd broken part.
The only limits are artificially imposed, by governments and market capture. I'm currently helping a friend repair a National HRO-500 receiver, likely made before I was born. It's amazing that they managed to build a synthesized general coverage receiver with Germanium transistors (the old ones are NOT reliable, nor anywhere near interchangeable, even with the same part #), but it's getting repaired after 60 years of service.
At that level of effort entire electronics systems might be replaceable too.
If a washing machine circuit board breaks (It probably won't, so e mechanical thing will), I could probably swap everything to run on a PLC if I really had to.
I wouldn't have a clue what to do if the mechanical timing stuff broke, other than cleaning it and swapping it for a PLC if that failed.
I keep hinting, to the various TLAs* watching the internet, that I'm willing to sign disclosures, and help fix things / reverse engineer them, if they need me to. But nobody shows up at my door, telling me I need to help my country. ;-)
I'm hoping Barry-1[1] works, and pushes into a higher orbit (increasing the Semi-Major Axis on the graph), without propellant. If it works, it shows our "laws" of physics are wrong, and forces us to refine our understanding.
For me, that marks a turning point, and a start down the road to being an Interstellar species.
How far back are we talking about? Because even in the 60s and 70s enthusiasts could create all kinds of parts for, say, cars, on their garage workshop.
You're right. Televisions used to be more repairable too. I found a microfiche reader by a dumpster and for fun ordered a lot of 100+ microfiche cards from an old TV repair shop on ebay. Zenith for example published detailed repair manuals including detailed steps for recalibration, parts lists, schematics, annotated PCB layouts, and more. These were for repair shops not the average person, but modern TV companies don't even produce those materials anymore.
Honestly I only think it’s cheaper to replace because we’ve made repair more expensive than it needs to be. I’ve fixed two TVs that stopped working with a few dollars of capacitors and my own time. That’s much cheaper than buying a new TV! But I had to go digging through forums when manufacturers could have provided this information. Instead they refuse ti share info and charge extreme prices for repairs where they will just throw out an entire board with one bad component which would cost 20 cents on its own.
I agree it’s a terrible system I just wanted to clarify that the “it costs less to throw it out and replace” narrative is only true so long as we make it true.
>they conclude that we should move back to old mechanical systems. But mechanical systems are not always more reliable.
In general I agree, but I worry this glosses over the subtle cases in the middle.
In a non-zero number of cases, the only reason electronics are "superior" to mechanical solutions is precisely because it's easy to enforce proprietary restrictions. In this subset of cases, it is proper to replace the electronic solution with a (here, superior) mechanical one.
.
My point is, we should acknowledge that there are some cases "right on the edge" that would naturally favor mechanical instead of electronic solutions, but the ability to cheaply enforce proprietary restrictions pushed them "over the edge" to using an electronic solution instead.
In these cases we should indeed consider replacing electronics with mechanical solutions — if not for retrofit, at least for later product iterations.
The cases right on the edge are real, but not common.
All the examples I can think of are hundreds of years old and have no serious competition from digital stuff, like eating utensils (for abled people, stabilized spoons seem to be a great invention for those who need them).
What do you consider an edge case that should be mechanical?
Electronics have wear and tear too because at the end of the day it needs to run on some hardware. Solder joints are never built to last until the end of time, just to not lead to a lot of repair claims within the warranty period and any longer is an opportunity to cut costs for a future manager.
I refurbish a whole lot of ewaste as part of my job, and with a few exceptions any time a device fails it's either the electrolytics in the power supply or something mechanical.
Solid state electronics technically _can_ fail but their expected service life is an order of magnitude more than the mechanical stuff. Same with properly done solder joints. If they crack, it's either due to mechanical stresses (e.g. constant reinsertion on battery jacks) or it's a design fault (Xbox 360, PS3 cold solder).
That order of magnitude longevity gap is a profit opportunity for the component manufacturers, and people are absolutely working to figure out how to make them cheaper.
I dont mean this cynically, only that no one wants to spend extra on overengineered parts.
DRM is a very modern concept. We originally made electronic solutions to mechanical problems because electronics allows for much more complex and nuanced response to input conditions, as well as the physical advantages of size and complexity.
Some systems are better as simple mechanical devices, but I'd put that number at a very, very low percentage.
Take your washer, dryer, dishwasher, or oven. These things can be purely mechanical, and are generally incredibly reliable when they are mechanical. In modern times, we've computerized these things for very dubious reasons mainly down to user tracking and enforced obsolescence. But adding network connectivity can be genuinely useful. Get a notification when your laundry or dishes are done. Automatically turn off the oven if you forget to when you leave the house. Track the runtime of your dishwasher and get notified when you need to clean the filter.
People would use features like that if it were possible to do so without a proprietary app that barely works, but demands your email and phone number and access to your microphone and camera. Or mandatory firmware updates that slowly degrade performance until you buy a new one.
The number of problems that can be solved mechanically and for which electronics cannot provide any extra value for the user is almost certainly very close to zero.
Just to add to your first point, many washers feature digital control of motors at the very low level, especially for BLDC motor types. This can reduce noise, increase efficiency and longevity of a motor. This requires electronics and software for precise timing control.
Eh, I think that's a different thing. The line between machine and electronics is pretty blurry when you get down to it. Is an LED attached to a mechanical switch electronics? What if the switch controls a transistor that switches the light? A relay? A CPU? What if it's an incandescent bulb on a mechanical timer? The distinction isn't super clear, and a lot of electronic circuits are much less complex and more reliable than a mechanical replacement. Is that good or bad?
I think the discussion here is less about "uses any electricity at all" and more "uses internet connected DRM to be actively shitty"
Indeed, I get the gist of the message, however I’m just highlighting how much advanced electronics and software drive something as ‘mundane’ as a washing machine motor to improve its efficiency [0].
This often requires real-time software calculation of precision timing for FETs that control the motors. Very loosely related to what an ECU does in cars.
I agree this is different to the DRM/internet connectivity stuff, but does disservice to the rest of the electronics and software housed within.
[0] Trapezoidal Control of BLDC Motors Using Hall Effect Sensors
> Hm, I disagree. ...
> *Some* systems are better as simple mechanical devices, but I'd put that number at a very, very low percentage.
We actually agree. Yes the number of cases is small (hence my wording "non-zero"). ;-)
I warned everyone I was making a minor point about "subtle cases," right? Both you and eternityforest thought I needed "correction" and pointed out it was a small number of cases, but I already agree. My reasoning process was the same as both of yours, and I came the same conclusion.
But while it's not many cases, we still shouldn't forget it. Imprecise thinking is the mind killer.
If a designer categorically decides (to save time) that we humans should never again revert anything (presumably, that has ever been done electronically, by anyone) back to a mechanical solution, they'll produce a worse solution in certain cases. Probably a lot of cases, actually!
One may counter, "I don't mean anything that's ever been done electronically by anyone, just the ones where electronics are better", but how do we distinguish the two, beyond in obvious cases? Worse yet, if we need to do that analysis anyway, what brainpower is being saved by having a rule to never switch away from electronics? :-\
TL;DR we should resist the temptation to shortcut with a simplifying rule "thou shalt not switch electronics to mechanics", because it neither rules nor simplifies!
Electronics separate form and function. The behavior can be anything we want with almost no extra cost. Plus, we can make module and reuse it in thousands of products with only code changes.
Electronics lets us pack in more complexity, giving us a place to compensate for imperfections elsewhere, such as users leaving babies behind in hot cars.
Plus without proprietary restrictions, nearly every vehicle and appliances in the world could probably use the same 5 modules or so for everything, making repairs rather easy.
But if the electronic design is open then what's would be the point of forcing an electronic solution when its inferior? The hypothetical is that electronic designs are open instead of closed.
It’s easier to fix with off-the-shelf parts (blown capacitor, damaged MOSFET, etc) than a part in which is manufactured for and by the automaker (e.g. variable valve timing component).
Totally nailed it. Without the information on proprietary systems to repair them or effect the cars re-use (if say doing an EV conversion), the car or anything becomes either a massive reverse engineering problem or landfill once the manufacturer stops supporting it.
Thank you, yes. Here is a blog post with more information about what we are doing [1] and a direct link to our funding page [2]. We are working on moving to a fully donation supported engineering model so our work can be freely available for everyone the world over, and your monthly contributions make a big difference. Also here is our github with our electronics, software, and hardware designs all permissively licensed. [3] Notably our open source dual brushless motor controller design is coming along nicely. [4]
How weather resistant is it? Is it aware of its surroundings yet (for instance: irrigation installations)? Are any of these in use outside of the people that develop them?
(sorry for the questions barrage but if I didn't live on another continent I'd hop over to help). Nice Al welding btw.
It is very weather resistant, our prototypes all live outdoors 100% of the time so it is developed in the rain. We don't get snow where we are so we will need to do more testing in snow as things develop.
It is not yet aware of its surroundings.
We are still finalizing the first version to ship, so currently no other machines are in use outside of our farm. We are working on getting there!
> Nice Al welding btw.
Thank you, though I don't understand. I do all the welding myself by hand. We also have a CNC plasma cutter for cutting parts. It doesn't use AI, just Gcode. :)
Be careful about that. Just last week he lied about open sourcing the original roadster. Musk very clearly said "all design and engineering for the original roadster is open source" but that statement is false. There's a few PCB designs available and one diagnostic program, but there are zero mechanical designs and I suspect there are missing firmwares, among other things. Given these glaring omissions (and Musk is certainly qualified to understand what "all design and engineering" means), I suspect there is a trove of other "design and engineering" documentation that is missing as well.
I believe none of the cars shipping today contain any open source electronics. Teslas are notoriously full of very expensive proprietary electronics. He has indeed released some "goodies" but is nonetheless as bad as every other shipping car manufacturer today.
It is honestly hard to say he is making an effort when he is so blatantly lying about the roadster and his other vehicles are completely proprietary. Not saying he is worse than the other companies, but I can't give him credit for something he is not doing.
> Musk is certainly qualified to understand what "all design and engineering" means
I'd say there is a mountain of evidence from SpaceX and Tesla that's not actually true and that he has little to no understanding of anything related to engineering, nor the ability to manage engineering teams effectively; I think he's surrounded by people that are actually talented and do a lot to scrambling to make stuff happen and fix his messes.
The huge 'box' his demanded his team put together, to rescue people from a cave with passages so narrow diver could not wear tanks on their backs? To me that demonstrates he has no ability to analyze a problem at high level and correctly apply the engineering resources available to him.
Examples:
* Dishy, which had basic design flaws, such as a permanently attached power/data cable that if damaged meant you had to purchase an entire new unit
* Hyperloop.
* SpaceX's repeated failures, which they spin as "it's okay, we meant to fuck up, we got data!", including the disaster where his insistence on launching on 4/20 resulted in substantial vehicle damage (5-8 rockets failed), heavy damage to the facility, a 380 acre wide debris field (so severe you could see debris landing in the ocean from an aerial shot well outside the launch danger zone), a forest fire, and heavy damage to a car (parked within the "hazard area.")
* Musk's repeated production-related issues at Tesla, and his insistence on automating as much as possible even when experts in their field who work for him tell him it's not possible
* Repeated demonstrations of engineering incompetence at Telsa. The Model S drivetrain units used to be disposable and would be quietly replaced during routine service at anywhere from 10-30k miles. They still can't make a Tesla with proper body panel gaps. People have found bits of wood/wire/duct tape holding their Tesla's guts together. For years Model S's couldn't be driven in heavy rain or through large puddles without water ending up in the drive units. Windows randomly shattering while cars are parked in driveways and garages. Etc
Stop, just stop. I’m not going to reply to the points I’m unfamiliar with but this crap about SpaceX has to be answered (“someone is wrong on the Internet!”)
SpaceX operates what, by reasonable measures, is probably the most reliable launcher in history. That’s the Falcon 9. They are absolutely capable of reliable space operations.
Their two test flights of the Starship stack are test flights. Perhaps a better term would be R&D flights. They’re building something way outside the envelope of what’s ever flown before. If they followed the approach taken by big government contractors and/or classic NASA, we might expect them to fly around 2042 after spending many billions of taxpayer dollars. (Exhibit A, SLS). What you see is a company taking risks to advance the state of the art.
Be thankful or be ignorant but please, please don’t be loudly and proudly wrong!
> The huge 'box' his demanded his team put together, to rescue people from a cave with passages so narrow diver could not wear tanks on their backs? To me that demonstrates he has no ability to analyze a problem at high level and correctly apply the engineering resources available to him.
The solution they came up with for those kids was extremely risky, involving drugging them and hoping they didn't drown in their face masks. It had never been done before in cave rescue history, and no-one really knew if it would actually work.
Cave passages can be widened. It's actually quite common for cave rescues to involve widening passages. It's also relatively common for cavers to do that in general as part of ordinary exploration - I personally have helped widen an impassibly narrow passage with explosives myself. And yes, this is done underwater too.
If the drugging solution had resulted in a kid drowning, there is a very good chance that the submarine solution would have been used.
> Dishy, which had basic design flaws, such as a permanently attached power/data cable that if damaged meant you had to purchase an entire new unit
That sounds like a typical engineering/marketing decision: rather than coming up with a tricky waterproof connector, just seal it permanently and sell the user a new one if they break it. Starlink as a whole has been very successful.
> SpaceX's repeated failures, which they spin as "it's okay, we meant to fuck up, we got data!", including the disaster where his insistence on launching on 4/20 resulted in substantial vehicle damage (5-8 rockets failed), heavy damage to the facility, a 380 acre wide debris field (so severe you could see debris landing in the ocean from an aerial shot well outside the launch danger zone), a forest fire, and heavy damage to a car (parked within the "hazard area.")
SpaceX is probably the most successful aerospace company in history., and they have some of the most reliable orbital rockets in history. Obviously their focus on meeting schedules and iterating is working for them. They just launched Starship again with even more success, and will probably do so yet again in a few more months. I really don't think SpaceX themselves care about occasionally spreading some debris around, setting some trees on fire, and damaging a car in the designated hazard area. There's lots of rocket engineers that have managed to destroy their own cars because their rockets failed and debris hit the employee parking lot.
Open source schematics and a closed Supercharger network until this year, at least in the US. It's hard to separate Tesla's intentions in opening up their charging network from the legislation in both the US and EU incentivizing it.
No they did actually follow through on the charging stuff to the point where NACS is now the new north American standard, which is kind of wild to see given how they've acted before.
The conclusion in the title isn't justified by the content. Bad designs are bad designs, regardless of whether they're electronic systems or mechanical systems. Ford and Nissan have shipped plenty of mechanical systems that leave something to be desired.
Funny enough, I had a Ford truck in the 80s that spent a week in the shop because a shorted taillight bulb caused the entire electrical system to misbehave. No CAN bus required.
And, dealer only parts? That's hardly an 'electronics' issue. Try to replace anything other than body panels or drivetrain parts, and you'll find a lot of this.
> And, dealer only parts? That's hardly an 'electronics' issue.
If resistance differs among any of six coils, halt.
That sounds like a software issue to me. It prevented him from replacing a single faulty coil, and required him to buy and replace all six at the same time. So that's dinging him with dealer part markup plus 5x the dealer part price!
I am saying that "dealer parts" aren't a frustration that is caused by electronics. They're caused by the aftermarket not making them.
The design of Nissan's fault detection doesn't have anything to do with it being a dealer part. Something dumb like a cupholder is 99% of the time going to be a dealer only part.
And on top of all of that, it's not like 'electronics' is enabling Nissan to force you replace all the coils. Automakers have done this mechanically for decades too. Just make it a single part: https://www.amazon.com/ENA-Compatible-Econoline-Thunderbird-...
Drivetrain parts are the easy things to find in the aftermarket, because they're wear parts, and people have to replace them to make the car go, so they're profitable to make.
The kind of stuff that's hard to find in the aftermarket are things like oddball interior parts that don't need regular replacing.
Yes, true. But what I meant (and failed to explain, apparently) that for instance a gearbox can be paired to a particular ECU and that the manufacturer does that to stop the gearbox from being sold separately from the engine, which makes it harder for a chopshop to do its thing because and engine without a serial number is kind of suspect but a gearbox (which, especially for a larger car can be quite expensive) would often be overlooked during an inspection. By tying the two together it now becomes mandatory to not only swap out the gearbox but also the engine (or the ECU, but that's not all that easy either).
I've had a VW van stolen right from in front of my house and to this day it's a mystery to me how they did it, there was no glass so they must have had a key or a way to open the vehicle and those are fairly solid. You can't start the car unless you have a key with the right transponder in it and that transponder is known to the ECU. So you'd have to bring a key that works on the door, open the door, teach that key to the ECU (or swap out the whole thing) and then drive off with the car. But it was all done in a couple of minutes at most.
It's 8 years ago now or so and I'm still looking for it, every time I see a long wheelbase silvergray VW with a camper roof my head turns all by itself to check the roof raises at the back (which is a very rare combination).
Yes, there are also anti-theft features built into electronics on cars too. (There's also frustrating anti-theft features built into mechanical parts on cars too, e.g. locking lug nuts, fiddly key mechanisms, parts with hard to access or nonexistent fasteners, etc)
But also, most vehicles don't really technically have ECUs anymore, despite that name sticking around in common use. Most stuff in the past 30 years has a PCM, because there's a lot of features that makes sense to coordinate between the engine and transmission in a modern car.
Even back in the 90's heyday of EFI engine tuning, one often had to be conscious of which transmission their "ECU" was expecting, because some transmission features were staring to get integrated into the engine control logic.
Your explanation may well be the right one, never figured on that but it makes good sense.
The transmission issue bit me a while ago, I had a transmission rebuilt and as a part of the service they replaced an electronics board that had some exposure to lubricant on account of a broken seal. But it was keyed to the engine of the car and we had a really hard time getting the car to recognize the rebuilt transmission, in the end the old board was put back in and everything worked flawlessly. But that was a ton of extra work. ('97 MB E-class kombi). Lesson learned there, 'if it works don't fix it'.
This problem isn't just cars. I've had 2 washers fail in the last couple of years, each time the main logic board. The proplem is (mainly) the electronics, or more precisely, highly integrated, highly proprietary parts.
When they break, you can hardly diagnose them, let alone fix them, or replace them with 3rd party stuff. Buying replacements from the manufacturer (if you even can) is often uneconomic, and also risky, since maybe it broke because of some other failure, etc.
The solution isn't going back to simple wires and mechanics - his example of 'wiring in a tail light' would double the weight of the car with wires if applied to everything.
I couldn't agree more. I have a 65" TV that's about 3 years old. It was a midrange model with a nice picture. It's now sitting here waiting to go to ewaste because the mainboard failed. Unfortunately, it turns out that was a common failure in that particular model, and replacement mainboards are no longer available. And because of the proprietary nature and integration you mentioned, replacing it with something else just isn't feasible. It almost makes me sick to think about trashing this giant piece of equipment (which is part of the reason it's still sitting here) but I just can't see any hope for it.
What kind of washers if you don't mind sharing. I want to avoid buying that brand. My guess is Samsung because I have had really bad experience with Samsung stuff over the years.
Purchased a washer and dryer set when our old unit was shot and had multiple issues within two years. Also purchased an oven with a similar outcome.
Fortunately my tv is going strong ish. Had to have the mainboard replaced around 4-5 years and patiently waiting for the day it won't power on. The TV situation isn't as horrible with my only lasting gripe being hdmi-cec getting out of sync and only full power loss for 15-30+ minutes seems to fix it.
I'm avoiding their products like the plague after all this....
Their designs look great but... lipstick on a pig.
You could take look at Speed Queen. I went for their TC5 machine recently. No lid lock, real deep fill option, etc. It's a $500+ premium over what you can get from Samsung (or anything at Big Box for that matter), but I think it's worth it to have a proper machine that can actually be serviced.
Can't get seem to get SQ where I am, but currently looking at older GE/whirlpool models with minimal or at least modular electronics - anything from before the Great Enshittening.
I can't figure out if any new models are modular/repairable and built to last, or just trading on old reputation, but actually the same disposable garbage as everyone else.
Electronic systems can quite easily be far more reliable than mechanical systems.
High voltage ignition systems are a really good example. I used to own a 1973 Buick LeSabre with a stupid low-voltage ignition system for "emissions control" purposes. Any water inside the engine compartment, and your engine would stall. Put a modern high-voltage ignition system on that and you could damn near submerge the engine and it would keep going.
The problem here is a lack of interface standardization (for example: old school taillights all connected to bulbs the same way). If auto manufacturers were required to document interfaces and accept third party parts, they would have to design everything defensively and these kinds of silly issues would go away.
The problem isn't electronics--it's engineering. And the problem isn't engineers--it's management cutting every corner they can.
Another issue is the physical layer of the CAN bus system.
For something meant for automotive deployment in massive numbers I'm not particularly impressed with how the typical physical layer is implemented. There is relatively little margin for error and the topology is just designed for cost efficiency, not for reliability or redundancy.
It's so strange all the comments in here about how modern electronic stuff is completely fine because "it just needs to be open sourced so it's maintainable". Yet not a single modern car does this, nor does any manufacturer have any motivation to do this, nor does any manufacturer plan to do this, nor will any car ever actually have this. It's a completely dead argument.
Governments can shift the balance in favour of consumers with right to repair laws. The argument is only dead if we accept this situation. If it's dead in the US, that's because democracy is finished.
As engineer and entrepreneur in the digital space I am biased, but I believe what this man is saying does not make lots of sense. It remembers me the analog photography guys buying analog gear because this "crazy digital thing could not last". People will always love the grain and so on.
The key of the issue are proprietary designs. Cars' tech is established and hence a commodity, easier than ever to replace things thanks to technology like computers, 3d scanners and 3d printers. That made possible lots of third parties selling you repair pieces without having to pay an arm and a leg for the official part.
The future are electronic cars, battery or hydrogen or ammonia, because it just way simpler and better.
With electronics you can make entire designs open source, so you are not dependent of the seller of your car, but you can also make the interface specifications open which is the really important thing.
The IBM PC was an example of open interface specifications. Anybody could build new cards and insert it on the slots and start using it.
You do not need going back to mechanical systems. You need open specification of interfaces so different companies could compete on equal terms.
Really not sure about this dude’s, like… thing, but I absolutely agree that cars have reached a point of almost complete unmaintainability. It wouldn’t be such a bad thing if they were engineered in a fault tolerant way, but they aren’t. Their electrical systems are brittle as dry spaghetti, completely finicky and relying on very small margins to prevent basically frying everything.
I have a 1987 Toyota pickup, a 1983 Toyota Land Cruiser, a 1989 BMW 325i and a 1974 BMW 2002, a 2022 Toyota Tacoma and I used to have a 2021 Ford Transit.
Of all of those cars, the only one that had electrical issues was my Transit. Twice in fact. Once right after I bought it. The cluster went belly up. Then again with the driving lights on the left side of the van which would intermittently not work. Dealer couldn’t figure it out. I couldn’t figure it out.
They wanted to start playing the game of just start replacing everything until the lights work again. Knowing the quality of modern Ford mechanics, I opted out knowing they’d probably break something else in the process of playing whackamole with my car. It wasn’t a serious issue, just obnoxious.
Modern cars that have electrical issues cannot be fixed. Whoever is responsible for this change of things should be dragged out and shot. It’s criminally wasteful.
> Modern cars that have electrical issues cannot be fixed.
Some of it is just "volume" too. When you scale up the amount of wires and connections so much that formerly easy problems become impossible. Things like a somewhat buried wire getting some exhaust heat because a plastic retainer failed, or a squirrel getting into the engine compartment and chewing. You can sort that out in a car with dozens of circuits, but it's harder when there's hundreds.
> It wouldn’t be such a bad thing if they were engineered in a fault tolerant way, but they aren’t.
There are multiple CAN specs, but fault tolerance is a built-in part of some of them... enough that you can cut a wire entirely and it'll still work. That beats any analog vehicle electrical systems. Although, if it costs a penny more, I'd bet Ford cuts that corner.
The video he refers to about the $5000 repair for a taillight problem.[1]
This is a parts cost problem. Each tailight assembly has some LEDs, a control module with a CANbus interface, and a rear-obstacle radar sensor. Two of those are over $5000.
The other parts don't need replaced though, they were just showing as failures because the taillights were jamming the bus. That isn't ideal, but Tony made it sound like parts had to be replaced all over, not that it was just difficult to diagnose failure in the taillights along with the taillights being ludicrously expensive.
Not sure why he’d go into the topic of a BRICS alternative to a global reserve currency. That just ain’t happening. Especially a commodity based currency. That’s just not going to scale.
There’s a force aspect to the dollars role as global reserve currency, yes. But the main factor is that America built the world’s best currency and the best international banking system. It’s the only currency that could scale to meet global demand, and so of course it became the global reserve currency. None of the proposed alternatives have the properties needed to replace the dollar. So there’s no reason to think it’s being replaced anytime soon.
He might be a brilliant mechanic, but he sounds just like my paranoid anti-vax coworker when he starts talking about the BRICS currency. The solutions are open standards and documentation, not going back to the stone age.
I'm not a mechanic, but I'm an amateur mechanic as far as my YouTube viewing goes, and I've never heard of this. Plenty of people I watch, get junkie, old cars, and all kinds of things to fix but this one hasn't come up.
Is it possible? I'm sure, does it happen all the time I'm somewhat suspicious.
Yes, it's absolutely possible. You have a whole slew of stuff connected to that bus and if it shorts out then it can take other parties connected to the same bus with it depending on what is shorted out and how.
What you think of as a taillight is actually a couple of micro computers, a DC-DC converter and a whole bunch of LEDs, connectors and so on on it's own circuit board with potentially a bunch of other IO for the safety features mentioned in TFA which could basically be anything. So all of that lives on the secondary can bus (which can have 10's of clients). Short out that bus in a creative way and you may well end up creating a nasty cascade of issues that run from one end of the car to the other. Normally this shouldn't happen. But the potential is there. In my car - fortunately - I just have regular lightbulbs. They're a pain to change out because of how the front and the back of the car are constructed but the worst they can do is blow a fuse. I keep a very small socket and driver in the car just in case I have to do that on the road because without the right tools it is impossible. And even with the tools it is a 20 minute job.
The can bus can tolerate +-12VDC. Actually, something like the TJA1051T can take +-58V. The problem is that when there is a short on the can bus, you can't communicate over it. It isn't magic. I don't know what the exact problem in this case was, but I've seen it where the ABS module self destructed, and because it was on the same bus as the ECU and BCM, the BCM wasn't able to tell the ECU to start.
Those taillights have their own LED drivers and not rarely they pulse the LEDs with far more than twelve volts to get them to be brighter. The pulses are super short so they don't damage the LEDs. I don't have any idea if the Ford in the article uses that trick but I've seen it on a couple of other cars as well. Usually they are simple DC/DC converters using buck-boost drivers to make the higher voltage. Moisture (tail light leak mentioned in TFA) could lead to that voltage being present on the bus (at low current, but those bus drivers are puny).
Datasheet of a typical driver. So 60+ V out, nice little series inductor, 6A peak current on pulses a few ns wide. That'll do it, though the best CAN bus drivers on the market could easily sustain that without damage I'm pretty sure that not all of them would and I've seen them give up the ghost in e-bikes from transients (CAN is used in e-bikes extensively, both for the main components (battery, motor, UI) as well as for auxiliaries (front and rear lights, charger).
On another note: the price for the repair is ridiculous, but that's mostly because anything involving a transistor from a car brand is marked up 1,000%. Boards that don't cost more than $20 to make sell for hundreds of $, and replacing them is going to be a ton of work because of how it's all put together. A (sealed) taillight assembly could easily be $800. Headlights are usually even more expensive.
In general, about Ford: great when it works. But fragile and bad engineering on plenty of parts, notably: light assemblies, various braces (including the one that holds up your transmission, don't ask me how I know about that), trim pieces and most of the interior. Engines are great, it's almost like they are designed and built by a different company. Warranty is a ridiculous back-and-forth on things that are clear manufacturing or even engineering faults and you, the customer end up in the middle (and usually pay for the privilege). I had an F150 in Canada, and I liked it from the utility perspective but the reliability wasn't there and Ford as a company utterly sucks and will never see another dime from me.
25 years ago Ford was one of the biggest brands here, today they have a negligible market share in the personal vehicle segment.
> (Ford) Engines are great, it's almost like they are designed and built by a different company.
They were. Ford had not only an Engine Division, but a separate V-8 Engine Division. Both were their own fiefdoms with their own management. To this day, Ford has separate self-contained engine plants. GM is divided up differently, with foundries that do hot metal work for multiple parts of the vehicle. No idea how Stellantis does things.
Ah, that explains quite a bit. Thank you for that bit of insight, I knew about Ford SP for the racing world and crate engines but never realized that the division ran much deeper than that.
A friend of mine spooned a Ford V8 crate engine into a Midget and the result is as much fun as it is dangerous. Also the shortest clutch throw on any car I've ever seen, from fully engaged to fully free in about 1" of travel on the pedal. The firewall had to be cut out and the engine lives for a considerable portion in what used to be the interior :)
Just, no. A taillight short doesn’t cause the radio not to work because of canbus. I have no idea what this nonsense is doing here but a guy advocating for carbureted ICE engines because “computers and electronics bad” in 2023 shouldn’t be take seriously on this site of all places.
It can and does. I had a 2021 Ford Transit with a non working front passenger turn signal. I figured at best maybe a bad assembly, at worst, maybe something was up with lighting module.
Nope. Turns out the light shorted and killed my cluster while it was at it. It was under warranty so no cost there, but it would’ve easily set me back almost $2k for parts and labor.
I work on cars a lot, and if I wasn’t a software engineer, I’d be a mechanic (I’m actually going to be opening a restoration shop in the next few years). Basically every system on the CAN bus is integrated with every other system and even if they aren’t, there is a ton of cross pollination of systems you’d never expect.
That same van had a feature that would turn down the a/c blower if you received a phone call so as to not be too loud. Neat feature but i was just counting the days before that somehow cause my radio, a/c or car control module to go belly up.
I can easily imagine how that particular fault could occur. So I wouldn't rule out this particular complaint without having a thorough look at the schematic of the vehicle involved to see through what pathways the CAN bus or the radio itself (usually on a separate bus, but not always) could be drawn high or low. That's the problem with a bus: if you force it low or high that's effectively a denial-of-service.
I've had a case recently where a radio locked up hard because of a CAN bus issue that had nothing to do with the radio at all... (2005 MB CLS). That was quite the headscratcher and it took multiple people with a lot of gear to debug the problem to the point that we figured out what the culprit was (a little box on the main CAN bus embedded quite far away from the radio that apparently controlled the power to the console portion).
CANbus is on the way out. The future of networking in cars is Ethernet. A proper switched network with TSN [1] can dramatically cut down on the amount of copper in the car while increasing bandwidth and reliability.
The problem with Ethernet is that it's (usually) not a bus. CAN is still plenty used in newer projects and they're getting even a newer higher bandwidth version.
Ethernet in cars is based on the idea of putting a switch right on the die of all your microcontrollers. Then every device has multiple Ethernet links and helps to route traffic on the network. This way there’s no bulky wiring harness (bus) and everything connects with a mesh topology with single links between devices.
From what I’ve heard, it can save about 100lbs of wiring per car!
Sounds like a routing nightmare. Do you have a more detailed reference about this? Full disclosure, I work in automotive SW, but maybe I'm not working in the right ecu to see this. Timelines for these projects are quite long.
I worked on the networking side at NXP Semiconductors who are building the chips to do all of this stuff. Here's an example of one of their recent devices built with these sorts of applications (automotive and industrial) in mind [1]. They've partnered with Bosch and TSMC on this [2]. Here's Bosch's description of the tech [3].
Ooh, that's related to the whole "domain ecu" buzz. I see a lot of push in this direction, but I'm skeptical of the direction they say they're taking. It does not make sense to me, but maybe it's just my failure to wrap my head around it all.
Networking as in the SW stack or some silicon magic?
Would love to pick your brain further about this if you can talk about it because so far, no one has managed to clear to me how they intend to do this magic transition to "SDV", all I saw were buzzwords and hand wavy explanations.
I worked on a pretty narrow feature so I’m not sure how it all fits together. I do know they’re really excited about TSN [1] and its potential to let you mix bulk traffic (in-vehicle entertainment stuff) with safety-critical signals and never drop the latter.
I think it’s fair to say that a lot of this stuff is being pushed from the hardware side and it’ll be up to software devs to figure out how to use it all and make everything work together properly. A lot of this networking tech is being pushed down into silicon though, so on the software side you’ll just be tagging Ethernet frames and letting the hardware do the rest.
That's the problem with a bus: if you force it low or high that's effectively a denial-of-service.
This obviously doesn't always happen in reality but CAN is supposed detect/deal with that - pegged bus should be detectable, bad nodes are supposed to float, etc.
...but that is what happened. Ford designed a car that bricked itself if that failure mode occurred. Whether it was born out of malice or incompetence does not matter.
I've been customer of yours for decades. For all that time, one of my main concerns was ease of maintenance, and for that I regard a pillar to be SIMPLICITY. This article is an extreme example of why I don't like your electronics in your cars -- an exception for the fuel injection and ignition system and, sure, a radio.
Otherwise, for ANY electronics, I don't want it. Get rid of it. To me a car with more electronics is MUCH less desirable and valuable. I'll work HARD to find used cars that have no such electronics. None.
E.g., my favorite car was a 1986 Chevy S-10 Blazer with a V-6 engine, electronic fuel and ignition, and NO more electronics. I'd still have it except winter road salt caused too much rust, e.g.,actually rusted through the oil pan.
Another car I liked was a 1971 Chevy Nova -- Chevy small block, 350 cubic inch displacement, V-8, automatic transmission. Sure, electronic fuel and ignition would have helped some.
Bluntly, in fact, no doubt, I'd take that 1986 Blazer over any and all your current cars with electronics. No joke or exaggeration.
To improve on the Blazer? Sure, do a LOT to resist corrosion.
Yes, I know, your electronics engineers have a long list of various aids, assists, reminders, monitors, etc. that can be done with electronics and are eager to include in the new cars. Get them jobs doing something else, maybe using electronics to automate parts of your manufacturing, maybe making toy cars for kids.
Just keep those engineers far away from any cars you want me to buy.
E.g., WWII movies show a Jeep. Okay, I'd prefer it to any of your electronic cars.
Are you getting the message -- I don't want electronics in my car? And why? Because just as in the article, it is a reliability and maintenance problem. The electronics cost a lot more and makes the car much less desirable, useful, and valuable. MUCH less.
For more, follow the famous, KISS principle -- "Keep it simple, ..." E.g., for headlights, just use the old, common, standard, cheap, really good, readily available "sealed beam" headlights. Use 4 of them, 2 low beam and 2 high beam -- get LOTS of good light, cheap, reliable, rugged, easy to maintain. For your stylists eager to create some stylish, high-fashion, dashing, sweeping, space age, new ..." artistic, emotional expression, send them off with the electronics engineers. Just use those sealed beam headlights.
I'm not joking -- for me, that WWII Jeep beats all your present work. It was simple, likely cheap, rugged, reliable, easy to maintain -- to me, that's a dream car, better than anything you are trying to sell. Yes, for that Jeep, rear wheel drive only would make it still better. The Nova and Blazer were much better still. The Blazer would have been still better with rear wheel drive only.
No joke, to summarize, get rid of the electronics and other junk that is expensive to buy and maintain but delicate and not rugged or reliable.
This is unintentionally a good example of someone who's aged out of his profession because he's not curious about how things really work anymore, despite the lengthy intro claiming the exact opposite.
He was repairing a Nissan Maxima, and either got an alert about a bad coil, or one coil was just entirely dead. He then wanted to replace just that bad coil.
This is after he was told they should all be replaced, note how he never thought to even ask "why?" at that point, just assuming "what do they know? I've done this before!".
He then finds that the car's computer checks all the coils, and just proceeds to alert on the next coil shortly thereafter.
This isn't because each coil has some DRM module to make you put another dollar in, but the engine is checking the performance of the coils, presumably it's picking up on something as simple as resistance mismatch.
So he blames the car, computers, the modern economy etc, and finally ends with some conspiracy theory about how BRICS are going to replace the USD as the world's reserve currency.
There's lots of little annoyances related to the modern CAN-bus in cars, but not this sort of thing. If he'd gone with his initial plan the customer would have unknowingly ended up with a car where the performance of each cylinder differed, which can lead to subtle and more major mechanical issues down the road.
It's also perfectly fine to just replace the one coil, get check "check engine" light, and just explain that problem to the customer. If they only want to spend $100 and not $100 times the number of cylinders that's their decision.
But whatever the issue is, it's not that the computer notices that the engine's performance is abnormal, and alerts you about it.
So you have no problem with junking cars that are most likely completely fine mechanically? Just because electronics is poorly made, prone to chain failure or because software does not support replacing some parts that otherwise could be replaced? Maybe having slightly different performance on one cylinder is completely fine. But with all the proprietary stuff, we aren't allowed to even ask!
How are you making the leap from the check engine light being on to the car being junk? You can just ignore it and keep driving if you think you know better.
In this case the customer was implicitly asking the mechanic to repair the car to the point where that diagnostics wouldn't fire.
Now, you can argue that those diagnostics are overly anal etc., but let's not blow this out of proportion.
This is analogous to replacing a single blown tire with a new one of a slightly different size. Most modern cars will also detect and warn about that (e.g. noticing the difference in spin at the differential), and the manual most likely says "don't do that", but the car will probably be "fine" in practice.
The mechanic made the leap not me. He replaced thousands of coils and everything was fine, till the manufacturer decided to be anal, thank you for the right term. And then there are scrapyards full of fine, but artificially unfixable cars.
I sometimes try repairing things myself and it's like this not only cars but with everything. So much unnecessary waste. But for you I'm just some weirdo that wants to drive with one slightly different wheel right? :(
I'm not saying the manufacturer is being anal, I'm saying that's really the extent of what you can argue here. The car isn't "broken", it's just reporting an engine problem.
The same car from the 70s would have had the same coil issue, you just wouldn't know it unless you opened up the engine and checked them with a multimeter.
I think you misunderstood what he was saying, that's understandable because he's really talking about at least four different things: This Nissan story, another story about an F-150, observations that scrapyards are full of "OK" cars, and how BRICS is on the rise.
I'm only commenting on the Nissan story, as it's the only one where he's offering enough specifics to see that he's simply wrong about it.
A sibling comment pointed out that you can get these coils for around $20 a piece, he went to the dealer, so he likely decided to get the OEM certified ones for $100 a piece.
That a mechanic doesn't know that is baffling. Every manufacturer does this, it's because the OEM parts are really intended for their own dealer and repair networks (which get a hefty discount) while the car is under warranty.
Once it's past warranty you can get the exact same parts from any number of generic manufacturers. Often from literally the same factory as the OEM, the only difference being the lack of an OEM stamp/certification.
I also repair things myself, including my car. I've got lots of what I think are legitimate beefs with the computerized aspect of doing that.
But that doesn't mean that everyone who runs into any sort of "computer says no" issue is right about their particular complaint.
I don't see how that particular issue is baffling to you. These coils might not have been so cheaply and readily available back then. That happens all the time with proprietary parts.
It's baffling that an experienced mechanic wouldn't know that there's often a 2x-5x price difference in "genuine", "OEM" and "third-party" auto-parts. It's something you usually learn the first time you repair your car as an amateur.
So he knows about that, but "I couldn't wait for overnight shipping, or otherwise I could have replaced all the coils for the cost of just one OEM one at the dealership" doesn't build the same amount of YouTube drama.
It seems like people see the growth in opaque proprietary electronics and they conclude that we should move back to old mechanical systems. But mechanical systems are not always more reliable.
What I really want more people to understand is that electronics are not the issue - proprietary designs are. Open source designs inherently have enough documentation that even if they don't come with some nice service manual, the nerds on the user forum can dig through the schematics and code to sort out what's what.
Unfortunately open source hardware just isn't getting enough coverage for people to really see the value in it, despite the entire internet being built on open source software. We need to find ways to communicate the power of open source beyond software. We could have open source modules in cars with open standards. Sure the manufacturers need a rugged PCB that can control lights, read sensors, interact with CAN BUS, etc. But there is no need for that system to be proprietary. We need to build and demonstrate the value of open source hardware to help people understand what is really possible. Because the call to move back to mechanical systems is really just a call to move back to systems that are intuitively understandable, and today's shadetree mechanics can read a forum post as well as anyone else. What we need are open, documented systems. Electronics are not the enemy, proprietary systems are.