11 ms·
> Until the late 1980s, a cable on the engine throttle connected directly to the accelerator pedal, giving drivers total control of their engine speed and power
by kube-system 5y ago
> Until the late 1980s, a cable on the engine throttle connected directly to the accelerator pedal, giving drivers total control of their engine speed and power. Throttle problems were quick, easy to diagnose, and, more importantly, fixed at home without paying exorbitant mechanic prices.
> Cars now use an ETC (electronic throttle control) managed by a computer, as is just about everything else on engines these days. Naturally, this makes vehicles more difficult to repair, not to mention the glaring “right to repair” issue growing by the day when everything runs on a chip.
It isn't more difficult because anyone is conspiring against you. Maybe it's more difficult because electricity is invisible, the techniques require new knowledge, or you're not familiar with it. Go get some CAN bus debugging equipment, plug it into your high-speed CAN bus, move the accelerator pedal, and you'll see the CAN messages.
Although, there's usually no debugging necessary for your accelerator pedal, because you no longer need to lubricate your accelerator cable; bits don't require lubrication.
From an engineering perspective, it is very convenient (and reliable) to have everything in the car just sit on a communication bus and write software to do the logic, rather than have dedicated wires or mechanical connections.
- jesterpm 5y agoBut it's not really that straight forward. You push the pedal, the computer runs an algorithm, and then throttle body moves. The algorithm is the part that you can't troubleshoot, mitigate, or fix. I recently started driving a drive-by-wire vehicle. A while ago I had an issue where the engine dropped to a near-idle (on the freeway) and the pedal was effectively non-operative. The diagnostics said the computer wasn't getting the expected results from moving the throttle body, so it went into some kind of safe mode. Now, there are plenty of ways that drive-by-cable could fail, but in this case I was slightly resentful because I could have mitigated the computer's loss of senses with my own.
- kube-system 5y agoThe problem there is that the software is proprietary, not that it has an electronic throttle pedal.
- thot_experiment 5y agoA million times this, I'm currently installing a hidden pass-through can-bus logger before I take my car to the dealership because of this insanity. So much of my life is wasted reverse engineering stuff when I could be doing shit that actually contributes to humanity. This is the cost of a broken IP system.
- kube-system 5y agoI agree with the sentiment, but I don't think IP law is to blame. I think it's consumer protection law that is the problem here. Magnuson–Moss should be amended to require warrantors to provide repair information. (edit: and tools)
- Nextgrid 5y agoIP law is the problem. The anti-circumvention clauses (that kinda make sense for media DRM) are used to criminalize tools that make unofficial repair (or just "repair", because official repair is just swapping entire modules) possible.
- kube-system 5y agoNobody would even need to make circumvention tools if the manufacturers are required to provide you a mechanism for doing what you want to do.
- kmeisthax 5y agoYes and no. We absolutely should be amending Magnuson-Moss, but the reason for why we need to legally mandate access to repair manuals is very much downstream of current copyright law. Companies realized that once you put software into a device, they owned the thing that makes the device useful at all, and they could then charge access to that software in the same way one charges access to a scummy mobile game. Furthermore, because the law surrounding creative works is extremely strict, any repair that might touch something that could be construed to be DRM protecting the manufacturer's software becomes legally dicey. You also can't tell anyone how to break that DRM, no matter how justifiable repair is. So if you somehow figure out how to reserialize new parts onto a locked-down vehicle, you're probably allowed to do that; but you can't legally sell that knowledge onto other repair shops.
- gameswithgo 5y agoThat you can't troubleshoot/mitigate/fix that is not an inherent property of such a throttle, but a choice by the manufacturer to not let you.
- willis936 5y agoNot so fast. It is emissions regulations that dictate it. I like driving cars fast, but locking down emissions to the chagrin of every car guy scratching their head online saying "why did it used to be better?" makes me happy. It's a small victory for the future of the species.
- kube-system 5y agoI am also a fan of emission regs. But, emissions regulations require that you leave emissions equipment intact, it does not require automakers to prevent you from touching those systems.
- NovemberWhiskey 5y agoYou feel slightly resentful, but if there was an incipient failure of the actuator and the next step might've been a stuck-open throttle then you should probably be a bit grateful too, right?
- spicybright 5y agoWould it have even failed if it wasn't drive by wire though?
- djmips 5y agoThrottle's did fail before drive by wire but I have no data to compare.
- kube-system 5y agoThrottles linkages are a regular maintenance item on vehicles with mechanical throttle linkages. And all (non-antique) engines with mechanical throttle linkages have return springs as a safety backup for when/if they do fail. Some of them also have redundant linkages on top of this. It is not uncommon for throttle cables to have their maintenance neglected, and eventually, they will wear out and break.
- NovemberWhiskey 5y agoObviously not in the same way. Overall system failure rates comparing cable vs digital throttle control - I have no idea. I do know that I can’t remember the last time I drove a car with a bad idle, which is much more than anyone who lived in 80s can say. Also: throttle by wire permits simpler cruise control integrations, better fuel economy and also facilitates the operation of safety systems like traction/stability control - so it’s not exactly a like-for-like comparison.
- smolder 5y agoThis is a nit, but idle control was automated well before drive-by-wire took over. Idle could be on closed loop control, but if the idle solenoid failed, it was less catastrophic than losing throttle control.
- thereisnospork 5y ago> But it's not really that straight forward. You push the pedal, the computer runs an algorithm, and then throttle body moves. The algorithm is the part that you can't troubleshoot, mitigate, or fix. You can, actually, and its fairly commonly done[0]. Mapping pedal to throttle position is trivial, as is intercepting and modifying the pedal output signal. Its also not that hard to rip out the throttle body and replace it with a programmable unit. Imo new cars aren't so much harder to fix and modify as they are different to fix and modify.[1] As Socrates rise from his grave and say: “The cars now love luxury; they have bad throttle control, contempt for steering; they show disrespect for elders and love economy in place of horsepower. Cars are now tyrants, not the servants of their drivers. They no longer rise when the key turns. They contradict their owners, turn off at stoplights, gobble up electricity at the plug, beep incessantly, and tyrannize their seatbelts.” [0]Pedal boosters or pedal tuners. [1]Not excusing the myriad of insipid design choices, of course.
- steelframe 5y ago> I was slightly resentful because I could have mitigated the computer's loss of senses with my own. For me this is the crux of the matter. As the person in the driver's seat, I demand that I be the ultimate arbiter of what the motor(s) and wheels ought to do, regardless of what any set of sensors and/or set of computer programs thinks is going on. I once had a car refuse to release the emergency brakes when it was time to disembark a ferry because the firmware misinterpreted the natural rocking of the vessel (and hence the car itself) as some kind of suspension fault. It left me more than "slightly resentful." I was absolutely furious at being stuck in a really rotten situation where my car was actively causing problems for those around me (people behind me had to drive off the ferry) and wouldn't even let itself go into a mode where it could be pushed off. The dealership's service center simply said, "Uh, sorry. Your brakes are locked. You can't override it. You need to get it towed in." I don't care if the car thinks it's keeping you from rolling off a 500-foot cliff. If I say, "No, release the brakes, I mean it," the car had better do exactly as I tell it.
- gameswithgo 5y agoI grew up as a car enthusiast during the transition to things being computer controlled, and for car modifying purposes it was fantastic. Relatively simple to plug my laptop in, monitor air fuel ratios, make adjustments as I see fit. Didn't have to rely on any third party or anything. However more rececnt trends see OEMs trying to prevent you from doing this kind of thing, in the same way that Apple, John Deere, and others try to prevent you from repairing or modifying or repairing your own hardware. This makes me very sad, because the future should be a utopia for people who whish to repair or modify their own hardware and instead we head towards dystopia, for no good reason!
- petre 5y agoThe problem is that your average Joe now needs an engineering degree or at least understanding of how a CAN bus works, how to properly terminate it and basically use a CAN to serial convertor to read tge messages which aren't always standardized. I agree that it's more flexible but you need a highly qualified professional and the right tools to work on the vehicle. I'm working with CAN networks myself but I still need an experienced electronics engineer when I get in trouble. In addition, most manufacturers add proprietary extensions and most of the work involves a fair amount of reverse engineering or NDAs.
- kube-system 5y agoI have a long lost relative who quit their job when their employer upgraded from horses to automobiles. They knew how horses and wagons worked, but working on what is now very simple mechanical engines, was way beyond what they were willing to learn. Technology changes and you can either learn it or not. Specialization of labor does make everything more complicated, but I'm glad I'm not having to grow my own vegetables or pump my own well water, despite the fact that I have no clue how modern farming or water treatment work. I suspect transportation will continue to become more specialized just like everything else.
- Spooky23 5y agoSometimes folks like that actually aren’t Luddite’s, but deeply understand a process. Case in point, I have a former high school friend who runs a farm that’s about 50 acres as his primary business - without major machinery. They have quads and pickup trucks, but most applications where one would use a tractor are now done with animal power. The result? It’s actually cheaper to operate as grass is cheaper than diesel. He’s not getting rich, but makes as decent living.
- floren 5y agoI'm really interested in learning more about this. Does he have a website? I've sat down and run some numbers before and it definitely seems feasible, but that's just back-of-the-napkin estimation. I grew up on a farm, but of course we were fully mechanized.
- babypuncher 5y agoI always chuckle when people make these arguments about cars being better when they were easier to repair. Sure, diagnosing and repairing a problems like the throttle control example is more difficult and expensive today than it was in the '70s. But cars are also so much more reliable today that the frequency of these repairs is considerably lower. I have a 2009 Toyota Matrix that I have owned for 12 years now. It has nearly 140,000 miles on it. Last month, the vehicle finally had to get a repair of a non wear item (failing O2 sensor), which cost me $300. In 1975, how many cars could be expected to last 13 years and 140,000 miles before needing a single repair?
- hef19898 5y agoDiagnostic computers are cheap enougj if your car is old enough. Thise for new models are expensive. Self repairabiliy stopped being a thing in the, IMHO, mid 80s. For a lot of stuff. Adding diff locks and the like, especially those not controlled by the onbord computers is actually pretty trivial, regardless of model year. What's said about EVs can be as much true about ICE powered cars, the computer controlled nature is the same for both. I appreciate my 1982 Range Rover more every day so. Besides a treadful fuel consumption. After all it is running on 3 fuses, perfect for an electricity and electeonics idiot like me.
- Nextgrid 5y agoKeep in mind that third-party diagnostic equipment is almost always based on reverse-engineering and may lack some features, and original equipment from the manufacturer is rarely available.
- hef19898 5y agoI haf this discussion with mechnic friend a while ago, obviously not not a 40 year old car so. Our true second car is now, I have to guess, 15 years old. Diagnostics are no problem, computers with the correct software are around 400 bucks. Cheap enough for either an enyhisiast or free mechanic. The main car is 4 now, diagnostics equipment is only OEM, and goes for 5 figures. Go figure, but it explains why diagnostic jours can be so expensive.
- burntoutfire 5y agoThe electronic-based acceleration pedal started malfunctioning in my car merely 4 years after purchase, and required a fairly costly repair... Maybe I was unlucky.
- outworlder 5y ago> From an engineering perspective, it is very convenient (and reliable) to have everything in the car just sit on a communication bus and write software to do the logic, rather than have dedicated wires or mechanical connections. I would say that some repairs are _easier_ now. The computer can tell you about many issues it knows about. You can see if a cylinder is misfiring. Or if the O2 sensor is bad. Etc. Just have to plug in a reader. It would be great to have some standardization and open documentation on this. But it beats dismantling stuff and trying to figure out by trial or error (or by listening to sounds)
- antiterra 5y ago> Until the late 1980s, a cable on the engine throttle connected directly to the accelerator pedal, Was this really just until the late 80s? I’m pretty sure I had cars with a traditional throttle cable with model years in the 90s or later.
- greyskull 5y agoThe Honda S2000 had a throttle cable until... the 2006 model year I believe. I don't know how prevalent this was in the early 2000s.
- narimiran 5y ago> The Honda S2000 had a throttle cable until... the 2006 model year I believe. Mazda Miata (MX-5) had it in the NB model (until 2005) too.
- warble 5y agoMy 2002 Tacoma has a throttle cable.
- stavros 5y agoCan I also send CAN messages to manipulate the throttle programmatically?
- wffurr 5y agoYou sure can. In fact, at least for my car (2018 Chevy Bolt), it's possible to find detailed instructions on how to wire up a throttle interceptor and hook it up to a Comma2 unit to add adaptive cruise control when it wasn't even available as a factory option. I even briefly considered doing so after reading a favorable review of the Comma2's driving algorithm by Consumer Reports, but decided too much could go wrong with messing with my car's throttle wiring.
- stavros 5y agoHmm, that's interesting... Why is the throttle interceptor necessary? Is the CAN bus jack read-only somehow?
- kube-system 5y agoThere are other devices on the bus looking for messages specifically from the throttle address. The idea is that you want to change those messages, but only sometimes (i.e. when cruise is on), so MITM'ing the messages is the easiest way to implement this.
- stavros 5y agoAh okay, I assume you can't just spoof the address then, thanks.
- kube-system 5y agoYou could, but you'd have two devices reporting different values for the same address and the devices reading that value wouldn't know what to do.
- 5y ago
- Reubachi 5y agoYou're intentionally leaving out the fact that a greater and greater number of manufacturer/OEMs are making necessary tools to properly debug proprietary and locked behind contracts. Or, you have a newer car and haven't had to diagnose yet a myriad of growing issues that you can't check in any way shape or form. Want to link into any car in the last 5 years can-bus for anything beyond simple air-fuel mixture issues? You can't. You can't even reliably go to a mechanic down the road, as they don't carry the 500k a year license from GM for the proprietary OBD2 scanner. This problem is 10 fold with electric cars, and 100 fold with Tesla or boutique EV manus.
- jaywalk 5y ago> Want to link into any car in the last 5 years can-bus for anything beyond simple air-fuel mixture issues? You can't. Huh, I guess I was just imagining it when I plugged into the OBD-II port on my 2020 Ford and reprogrammed a bunch of engine and transmission parameters to improve performance. I certainly didn't have any proprietary Ford software or hardware.
- markandrewj 5y agoIt is possible for a lot of vehicles, but at the same time there are companies have been trying to make self maintenance and repair difficult. I am particularly thinking about John Deere. Although not specific to Electric vehicles. https://youtu.be/EPYy_g8NzmI https://youtu.be/EPYy_g8NzmI
- markandrewj 5y ago*electric
- speeder 5y agoBe happy with your Ford. I had a Peugeot that required proprietary cable and software, and that software works only on Windows XP...
- shortstuffsushi 5y ago
- pengaru 5y ago>> Until the late 1980s, a cable on the engine throttle connected directly to the accelerator pedal, giving drivers total control of their engine speed and power. Late 80s? No, that's at least a decade off. OBD2 wasn't even made mandatory in the US until 1996. That's when things started to go off the rails.
- jjav 5y agoYes, that date range is way off. My mid-2000s cars still has a proper physical throttle cable.
- jjav 5y agoA throttle cable has one trivially diagnosable and fixable failure mode, it might snap. Although in practice they rarely do. A highly complex interconnected message bus has hundreds of bugs (guaranteed) and many unpredictable failure modes. Also diagnostics is potentially impossible if the manufacturer hides the details. Simpler is always better.
- nomel 5y agoThere are two failure modes. It can also stick. I've only experience a stuck throttle cable, not a snapped one.
- Rayhem 5y agoTrivially, simpler is not always better. A bike is simpler than a car, but it is not at all what I want if I have to go to the hospital.
- rasz 5y agoIf it was better you would see it in motor sport.
- jjav 5y agoNot at all comparable. Motorsports does not need to optimize for longevity or ease of maintenance over the decades. Both factors are irrelevant to race cars.
- dan000892 5y agoIn 2004, I found myself breaking waves at the helm of my land yacht of a 1991 Crown Victoria LTD on the rolling hills of an upstate New York forest road when I came upon an inferior vehicle which required overtaking. Alas, once a rare passing zone appeared and I positioned to overtake and put my foot down, the pedal limply dropped to the floor and I began to slow. (Of course it was once I was right next to the guy who rightfully gave me a WTF look but I digress.) After finding a good shoulder on which to beach my LTD and popping the hood, I found the problem: the socket side of the ball-and-socket throttle cable linkage had broken and would no longer hold together. Looking in my (voluminous undergrad compsci-student) trunk for a solution, my eyes landed on the -shiny- beigy PC tower, specifically the floppy drive. A couple strands of the disused 34-pin floppy cable well tied proved to be strong but flexible enough to hold that joint together and get me back to school. Despite my college having a robust automotive program (which I even had a suitemate in), I never replaced the part and that hack held until I got rid of the car 10 years later (though I kept the other 30-odd strands of the cable in the glove box just in case). While nostalgic, don’t read this as supportive because today I’m far more likely to find a CANBUS interface in my trunk than a floppy cable.
- abruzzi 5y agoreminds me of my 1963 VW Bus. The throttle pedal pressed down on a 90 degree arm/bellcrank which pulled on the throttle cable that ran to the rear engine. Because the throttle cable was old it had more friction internally, and the undercarriage of the vehicle wasn't in the best of shape. So the flat piece of metal on the body that the bellcrank attached to started to rip, so when I pressed the pedal down, the assembly would deflect and very little motion would be transferred to the cable. I took a small needlenose nose vise grip and clamped it over the rip in the metal, and then continued to drive it like that for the next 5 years.
- lisper 5y agoMy 1970-somthing VW Dasher wagon wouldn't go into gear one day. Turned out that the metal lever that the clutch pedal was connected to had developed a fatigue fracture and broken and could no longer disengage the clutch to get it into gear. But it was easily replaced (it was, after all, just a hunk of metal with some holes in it). Good times.
- joe_the_user 5y agoIt isn't more difficult because anyone is conspiring against you. Nothing in the passage you quote indicates this is a conspiracy, merely a change. Maybe it's more difficult because electricity is invisible, the techniques require new knowledge, or you're not familiar with it. Or maybe it's as the article suggests - complex electronic machinery creates the possibility, not the necessity, of an manufacturer sealing everything from being repaired by individuals. Is it beyond comprehension that a new technology can give many advantages and offers a number of problems? That people who criticize completely sealed systems aren't always "back to the good old days" Luddites (notably, even the historical Luddites didn't conform to what Luddite has come to mean now).
- jka 5y agoIs the source code for electronic throttle control systems typically made available to customers?
- upsidesinclude 5y agofrom a computer engineering perspective would have been more appropriate. As a mechanical engineer, I disagree with most of your comment. You've added sensors, electronics, servos... all to position control something that's an arms reach away (technically...). Lubricating a cable, which happens almost never in this case to no ill effect, directly compared to bits takes a few leaps over the reality of hardware.So the complexity is far higher and the control system can now be compromised, will be compromised by design and you are not writing the firmware updates.. Maybe the car you were taking about is in the metaverse though? edit: Perhaps far more importantly, the concept of features as a service has suddenly caught the eyes of vehicle manufacturers, see Toyota latest debacle with deactivated remote starters. The more digital control they have.... This is really a question of when. People are rightly assessing the sale/lease of software by the second in the very near future. Imagine driving down the highway on the first of the month and your pedal deactivates because you forgot to update an expired credit card autopayment. Future's looking bright, can't wait to get my toilet on CANbus
- rasz 5y ago>You've added sensors, electronics, servos... all to position control something that's an arms reach away There already were airflow/pressure and throttle position sensors and idle control servo in the throttle body unless you mean carb antique. Removing direct mechanical linkage decouples raw air flap control movement from driver intend (/1), allows things like dynamic throttle response, rev matching, better cruise control, better fuel efficiency by precise control over fuel/air mixture, lower emissions, and simplifies throttle body by removing a ton of mechanical junk. /1 If you really think you need that in your life try flying older helicopters where You the pilot must act as an engine governor because thats how it worked in the 60s and thanks to insane aviation safety regulations everything got frozen in time with 1060s designs being still manufactured up to 2010.