Category Archives: cars

FSD update

cars

With v. 12.5.5.2 Tesla’s vision based autopilot (supervised) the emphasis is on driver supervision, still. While autosteer and adaptive speed control has become relatively polished and worry free there are edge cases where driver intervention is necessary/mandatory.

One might think that the driver, after having made a manual course correction or driving adjustment, would “teach” the car. Future revisits to that kink in the road would then go smoothly but Tesla’s system doesn’t work that way. At the end of the day, reams of video clips upload from the vehicle cameras to mothership — Tesla’s neural network.

Pulling data (last 30 days sample) from vehicle (as a network client) transiting my network to the NN uploaded: 71.6 GB

Autopilot behavioral changes and improvements come from Tesla’s end-to-end AI training system, code named Dojo, via versional updates downlink to the fleet. In this my car (and all cars) becomes smarter. What needs to be smarter? Safety, always of course, but key on the list is the ability to flow naturally with traffic and mimic driver behaviors good or bad. For instance:

The “Hollywood Stop” aka rolling stop might be considered bad but most people do it. Autopilot is not permitted such liberty. It comes to a dead stop before proceeding and this is slightly irritating but lawful.

Along this vein, the NHTSA stop, where one comes to a complete halt at the white painted marking at a stop sign before edging forward to see traffic and proceeding. Good form but unexpected in the common place sense.

While at the intersection, number 1 at the traffic light signal, turns green. Go. Defensive driving used to be taught but people seeing the green light step right out. Autopilot waits a split second before proceeding. It’s a minor irritation, and I haven’t had anyone from behind impatiently sound their horn yet (but I know in certain parts they surely would). The pause is good. Drivers with the green should look for oncoming. Autopilot has been hardcoded to know that cross traffic frequently will fudge run the red light.

Not wanting to be labeled as one of those Hypermiling roadie types, I still like to begin a coast down as I approach a traffic light that has turned red. Aggressive drivers usually charge right up until the last moment before braking. This avoids the dreaded cut-in and unfortunately autopilot mimics this. I would prefer that autopilot would rely on re-gen to decelerate and minimize use of the friction brakes. Conversely, without traffic, autopilot accelerates from stop signals like a jackrabbit — way beyond Chill.

v. 12 autopilot dictates what it considers to be a safe rate of speed based on conditions and environment. Previous iterations of FSD would rely solely upon and match posted speed limits. But now, the car dawdles on some stretches sometimes doing 10 mph under. I know other cars in trail are thinking “Sunday Driver” or little old lady that can barely see over the steering wheel. You can poke the accelerator pedal a bit to help with the confidence I suppose. So much for the annoyances.

Intervention: the need to take charge, assume control, over-ride the automation. Save the day!

More seriously, and Tesla does require users to acknowledge the need and requirement for supervision, there are edge case where autopilot has low confidence or encounters a situation for which it can not deal without help. In safety situations autopilot (hopefully) recognizes and alerts the driver with an audio chime and a take immediate command text exclamation. A situation is imminent (1-2 secs.) and if the driver has not been attentive there could be consequences.

The Handover: automation off, manual control on, is fine when the human operator desires but when it happens unexpectedly because of autopilot initiative the human can be momentarily caught off guard. A surprise transition is messy.

Luckily, these are now rare. The vast majority of intervention occurs when the driver’s comfort level is close to being exceeded. e.g. the vehicle cuts a corner, or threatens to curb a wheel. In such a circumstance driver takes control by disengaging.

Disengagement (driver induced): press the off button, tap the brake pedal, or for the most immediate action; yank on the steering wheel

A disengagement by manual steering will literally cause the autopilot to take embarrassed offense by posting a text to driver: “what happened/what went wrong?” Optionally the driver can respond with a haptic press on the microphone and give a short verbal complaint. e.g. “the car lost situational awareness in that turn and assumed the incorrect lane”. This, along with associated before/after video clip sequence capture, goes to the engineers for review.

I wish there was a way to similarly send kudos or an attaboy. Occasionally, autopilot does something unexpectedly brilliant. Recently exiting a parking lot with a construction barricade just prior to turning right onto the highway required a hard 90 degree turn which the self-driving accomplished with low speed steering full-lock all while maintaining lane discipline. I was primed to takeover because I was skeptical for the outcome but no need. Later, also a 2 lane road, I spied debris (a small branch and leaves clump) in my path and seeing it coming put my hands on the wheel when I felt autopilot do the nudging and offset ever so slightly right to avoid contact. Very subtle and I wanted to applaud.

There are other features with v. 12 that I am experiencing such as Autopark. More than a party trick (i.e. Summon) Autopark is cool and works well consistently backing into the white lined bounded box centered and without any involvement. When its time to leave the parking spot autopilot knows in which direction you want to go. i.e. if you backed into it you probably want to pull out forward when leaving it. Autopilot automatically selects forward ‘gear’ saving you the effort. (As supervisor, verify and confirm the selection of course)

There are times of despair when you say don’t take away the steering wheel just yet, FSD won’t be happening anytime soon, and not ready for prime time. You feel like a beta tester and wonder how they get away with selling this Full $elf Driving option. The manufacturer should be paying ME to be the early adopter helping with the machine learning. I do feel like I’m riding herd with a 15 yr. old with a learners permit coaching / anticipating everything.

The next FSD version 12.5.6 and within a short time (Elon time) v. 13 there are to be significant changes.

With FSD version 12 my driver job is quite secure. It is more fun to drive than to watch it being done and less stressful. I’m always comparing technique and so far I can still say I’m much smoother and more efficient at it. Never the less Technology is a wonder and it is exciting to be a first hand observer of this innovation and progress.

Ace – photo detective

cars

From Aunt Jayne’s dusty family archive comes this forgotten 1910 photograph of a first car in Iowa…

somewhere in Iowa…

What car is it? A major clue is the marque logo on the facade. Star Cars. The emblem, while somewhat faded, is a match from the Durant Motors Company, manufacturer of.

Model C Runabout

A 1923 Star Runabout roadster is an obvious preliminary guess but this automobile using a catalog picture above is not a match for Pop and Mom’s car. Here is a better one:

The trunk bustle has a reverse sweeping curve up to the convertible top whereas the Star (above) maintains a constant arc to a level body top edge line. The side by side cars are much sportier looking especially with the rakish mountable spare tire. The visual comparison on the right is a [1918-] 1923 Dodge Standard A Roadster. Tell me if you think it a ringer for our original car on the left…

As usual more questions than answers. Grandpa’s first car is showing some age so perhaps the photo was snapped sometime after the mid ’20s. There isn’t a numbered registration plate on the bumper. The place looks closed; out on a Sunday drive? They could have acquired their car from the Star Cars Auto Repairing or had it serviced there however It seems an unlikely photo venue.

We are on the wrong track. The object of the photo is quite possibly of the building or property itself which just happens to include a random car in the foreground. The original photoshoot could not happen before 1922 because this was the inception of the Durant Company’s Star Car. Based on the weathered appearance of the paint logo on the building face, I’d have to say that it was a least a few years further along.

This car pictured, having no backseat, would have been suitable for 2 or 3 but hardly large enough to be used as a family car. I count Grandpa, Grandma plus 4 children by that time. A roadster would have been impractical and the folks in Iowa were a sensible and pragmatic sort.

Here’s what we know: The caption date of 1910 is bogus and we are pretty certain that the car is not a Star; furthermore we are not swayed by claims of the featured car’s provenance. The significance of the building we have yet to learn. My Aunt’s pen & ink caption is fantastical — or is it…

Benz-Jewelry

cars

A broader flattened and padded safety hub was introduced with the 1959 W111 ‘fintail’ model. This engineering effort was years ahead of other makes. The padding may have been a false sense of security but the expanse would improve chances in case of accidental impact. Previous hubs were decorative and could be pointy (harpoon spear shaped).

Donor hub

Serving its purpose for years after having been liberated from Dad’s 1967 200 D (W110) which was pretty used up, I’m certain that this relic piece survived the car. It now lives on the shelf having been replaced with a reproduction part that is a near match to the original ivory one.

The wheel rim has stress cracks so it too will have to be restored at some point but Mercedes bling is like jewelry $$$. So, methodical pace.

These early cars had primitive seatbelts (if that) so any passive protection from the steering wheel is appreciated.

Hood Hinge Procrastination

cars,

I’d released the counter springs on the hinges decades ago and so the hood would no longer hold itself up; not without a prop rod. The springs were tensioned to their weakest setting and I knew not how to increase it. Tools on hand at the time were a pair of vice grips with bailing wire; good enough to release a spring but clumsy. Coil springs have stored energy and can cause injury when mishandled. Uncertain about how to [properly] go about it and a bit apprehensive the job was put off. Fast Forward today to the Internet Age; a Youtube hit gave inspiration.

If the springs could be extended and placed in the hinge’s more extreme setting slot then the increased force under tension would correctly balance the open hood. Unfortunately, the size of each individual spring (one per hinge) was such that mere muscle power would not overcome the difficult flexing required to get that job done. Leverage was needed.

rope line loop used to coax the spring

A couple of Ratchet Tie-Down Trailer Straps provided the persuasion. An concrete post in the garage floor lent an immovable anchor point. A second strap maintained a proper pull angle since the post was off center.

ratchets serve as a Come Along

The ratchet acting as a winch provided the grunt and the ear of the spring was positioned adjacent to its proper location. With care and finesse the spring could then be hooked over the tang and into the correct slot on the hinge mount.

I have to give thumbs-up acknowledgement to the Youtuber DIY; he sure made it look easy. Well, it’s never easy but a least now finally — no more prop rod! A bucket list project perceived as insurmountable all these years gets resolved.

Student Driver

cars

Trust (but verify)

Self driving cars are a few years away. Letting Autopilot take control is like watching an inexperienced beginner on a learner permit. Fortunately, as a monitor (with skin at stake) you can gently intervene or abruptly take-over if the performance is not to one’s liking.

Current state of the art tech with Hardware 2.5 does quite the job centering the vehicle in the lane and maintaining proper distance from the car ahead with adaptive speed. Today, on an extended drive, I was able to sample test the latest enhancement: Navigate on Autopilot (beta) which further extends Autopilot.

Following GPS guidance Navigate on Autopilot suggests lane changes required for the route. Using cameras and forward radar it also recommends an open lane when overtaking a slower vehicle. You, as second pilot, acknowledge the suggestion and give permission with a tap on either stalk on the steering column (ignore is the default). In actuality you are babysitting the proceedings — and like a hawk.

There isn’t, nor should there be any trust expectation between you and the machine. The manufacturer (and his lawyer) feels the same way. Beyond any disclaimer you need to be attentive. The reason is called the automation hand-over. The autopilot can, after the briefest of warning, signal bailout and return control. If your head is not in the game this transfer could be messy. So, autopilot continually confirms that you are standing by. If you’re not detected actually holding the steering wheel you are politely prodded before it becomes serious. The system shuts off / locks out an abuser ignoring any cautionary nags.

Proper technique: the weight and friction from one hand is sufficient to apply “slight turning pressure” which is the assurance that Autopilot is looking for

The autosteer function continually saws back and forth ever so minutely seemingly testing to see that it has not been abandoned. A sensitive passenger may perceive the motion but in comparison to earlier iterations it is fairly smooth going.

Hardware 1.0 was primitive as autosteer would ping pong on secondary roadways hugging the inside line or wandering inebriated opposite. Still, it was exciting to experience this alpha software. Realizing that the future had arrived inspired anticipation for FSD

Lane keeping on today’s trip was quite good during the Interstate portion. There were some, shades of yore, one spurious blip of regen, a wander noted in the tight radii of an exit ramp. A little faith was required but no boundaries were exceeded. I did wonder what any closely following driver might be thinking.

whoa boy (too wild)

This happened… A mowing tractor pedaling the shoulder as fast as he could go — hazard flashers blinking but protruding somewhat into the right lane. Autopilot decelerated for the detected obstacle as it should do. I switched off auto steering and moved to the left lane to pass which was occupied by a truck just ahead. During the maneuver, just as I started over Autopilot saw daylight between Mower and Truck and (adaptive cruise control still engaged) accelerated harshly. Splitting lanes like some deathwish motorcyclist would have been a bit too sporty.

push nudge (too timid)

Autopilot slows as necessary for a vehicle ahead which is turning right onto a crossroad. Ordinarily, a driver can safely resume speed accelerating in anticipation of the other vehicle clearing. Autopilot delays this for a count of 3 before getting on with it. It seems like an eternity because know that the driver(s) behind have lost patience.

advanced nav (passed)

One part of the route today is considered tricky even for an experienced operator. I have traversed it many times and decided to let the student driver have a go. The course begins with an off ramp exit, followed by a double merge, another exit, a sweeping loop where traffic sometimes can crawl or even halt, and then yet another merge — all rapid fire. Autopilot has to steer navigate and sequence with the merging vehicles. It’s the full deal. Tight choreography. Autopilot will slow to make a merge opportunity happen but doesn’t offensively overtake for a cut-in. 

As blue traffic autopilot performed a weave merge with cyan traffic. Green dots depict where autopilot asks permission to exit route leg or for the navigational lane changes.

Eventually cars will computer interface with each other and the dance sequence will be safe and carefree. Until then, it’s student learning so; mind how you go. It’s early days yet.

Ponton Door Sill Restoration

cars,

Old cars had running boards. A vestige from this bygone era, the modern car door sill is nothing more than a decorative piece to prevent one’s shoes from scuffing the body painted entry threshold. This plate is likely to be of plastic assembly with the make or logo embossed. On the exterior side is the tarmac and the other side is the carpet. When shut the door covers the sill plate.

The door sill/scuff plate on ‘Ponton‘ era Mercedes-Benz were not that far removed from the old running boards. They could be described as internal running boards as they were enveloped by body when the closed door overlapped them. They are really not wide enough to stand upon nor were they so intended but there they are welcoming gateway to driver and passenger. From the illustration below observe that the vertical chrome trim plate and an aluminium surround piece serves to offer scuff protection. The flat sill spans between them.

perished mat

The original rubber sill (above) is aged and brittle. It has seen better days. The refurbishment project was initiated some years back (pre-internet) where it stalled due to non-availability of replacement parts. Recently, I began the search anew. The problem is that there is no longer much demand for this style ribbed rubber material. What I could find was in black only or some egregious industrial offering. The pursuit picked up when I found “grey mat for Mercedes”. ‘Imported’ from the limited description sounded promising but the bulk size and color were not and were likely for some later model.

new mat in place

Finally, success! An outfit in Germany called Niemöller Ersatzteile für Mercedes-Benz Veteranen seemed to have what I sought. I ordered their self-described hellbeige ribbed rubber with high expectations and was not let down. The material was rolled in bulk so a paper template had to be drawn and the mat trimmed. The new piece was cemented down after the fasteners removed and the shiny bits cleared for access.

Factory labor intensive, I counted 15 individual screws for each door entry (for which holes also had to be drilled). I don’t suppose that they had bots or even electric screwdrivers to speed that process. A contrast in manufacturing technique, today’s sill plate application must take the assembly line person mere seconds to snap in to place.

Headlight Cover Restoration

cars

Full disclosure: The Before picture was taken indoors whereas the After is using natural light, thus the paint color variation.

Opaque / Clear

The Zuffenhausen factory specified a tough clear UV protective coating which has performed well for the past 13 years of exposure. However it’s a tough environment and these lense covers are only plastic afterall.

A polishing kit from 3M was purchased to rejuvenate the current cloudy finish. Using my low speed electric drill, the supplied buffer wheel attachment and abrasive media, pits and scratches were incrementally removed. I deviated from the prescribed and used some 320 grit to begin with. This made quick work of the original factory coating, totally removing it. Naked uncoated, stages of number 500, 800, 3000 (wet) progressed. Finally the kit included a liquid rubbing compound that completed the goal of clarity. The surface is now quite smooth.

Routine application of a formulated polish with water resistant polymers such as PlastX or toothpaste (!) should offer protection from this point onward.

PU

cars

Ordinarily the N2, O2, H2O, CO2, CO, NOx… and other gases exit the tailpipe but driver complaints indicated that exhaust smells were entering the interior.

It was somewhat of a mystery because noxious odors would only manifest after the engine had been operating for awhile (warmed-up) and only when the vehicle was stationary (idling at a stop sign).  Common deduction was that the emissions were drawn in through the fresh air ventilation system; so possibly emanating from under-hood? DIY Forum searches revealed no symptomatic similarities unfortunately, but it was revealed that not all exhaust is directed to the tailpipe at least not immediately.

Inner workings to help achieve automobile emissions standards include things like exhaust gas recirculation (EGR) which is exactly as it sounds. A portion of spent gas is returned to the combustion chamber to act as absorbents of combustion heat.  Another subsystem is called secondary air injection wherein fresh air is injected into the exhaust stream to allow for a fuller combustion of exhaust gases.

This is a good detective starting place because of their attendant plumbing and routing all of which is located under-hood making them suspect. The aforementioned DIY search revealed the secondary air injection as a typical trouble spot and prone to early fan blade  failure so the investigation invited focus.

An electric air pump (2) that operates only after an engine cold start and ceases at warm-up (fitting the driver complaint scenario) draws under-hood air from behind its finned cover and pushes it under pressure through rubber hose (8) where it meets one-way valve (10). From there it directly enters an exhaust manifold depicted in the diagram by grey outline silhouette. Bolt (12) attaches the one-way valve to this manifold. The purpose of the valve is dual purpose: to allow air to pass into the exhaust manifold to mix with the exhaust therein but also to prevent the exhaust gas to reverse flow when the air pump is switched off. Hence one-way.

Remove the finned cover revealed a ‘smoking gun’ find. Underside was heavy moisture (a by-product of combustion), black soot, and grime — a dead giveaway. Further inspection with removal of the check valve revealed that it was possible to blow air through the valve in either direction without impediment and was therefore faulty.

Forensic Display: The valve body housing has been destructively separated to make visible the valve head which should ordinarily be in the closed position for this state. It nearly is, but observe the small piece of plastic debris (a single fan blade) wedged between the valve and the valve seat. This is the culprit. It is holding the valve head off of the sealing body in a stuck open position.

A new assembly has been sourced and ordered. This fix will restore operation and prevent the toxic stink.

Why was this only noticeable when the vehicle was motionless? The answer is that, thanks to aerodynamics, there is high air pressure at the base of the windshield. This is also the location of the cabin fresh air intake.  When moving at speed it was enough to eject the underhood fumes harmlessly beneath the car. When pulling to a stop warm air rising and depending on prevailing breeze, brought the odor topside and directly into the vent. PU

What to do you see here

cars,

This photograph is hand dated 1910. The Randrup family lived in Copenhagen (København) Denmark The letter K on the registration plate so designates. The owners hand draped over the front seat armrest and his shoe on the running board seem to be rather crudely inked out. (early days for photoshop)

The carbide headlamps used acetylene gas but the side marker lamps were oil burning and that may explain why the lense on the passenger side is blackened. It would have been used, as required, for street side parking during the night.

There’s a folding top but it’s a fair weather open car without windshield or side windows. I would say very low mileage based upon the state of clean and shine. See the messy bits on the pristine roadway substrate inside of the front year tire. I hope that this automobile wasn’t blamed. Good reason to check your shoes before coming indoors!

The 1910 Randrup family photo op featuring the  brass era touring car above has been identified. It was a difficult task because I only had the first 3 design letters visible on the radiator: “Ber” and come to think of it; that longhand script style was pixelated and no doubt in a foreign language to boot.

Typing just three letters into a Google Search did not result in a handy autofill suggestion! I did however manage to find a comprehensive list of makes whose name began with the letter B. Matching up the next two letters narrowed that list considerably and voila. After selecting a few of these near matches, comparing images of early cars, fender and radiator shapes I found a match. The dead ringer confirmation is the unique “B” signature script of the name on the radiator which was clear in other comparison photo records.

It’s a French made Berliet motorcar and as a company went on to a long history but one not well known in the USA — except for one obscurity. The Berliet design and manufacturing rights were licensed to the American Locomotive Company of New York. They attempted to capitalize on an American perceived exotic european excellence theme but note that the eye catching brass Berliet logotype on the radiator was struck.

1907 American Berliet

Marine Carburetor Labyrinth

boats, cars

You’d think that a brand new marine carburetor would bolt on and function perfectly out of the box but I’m finding out that you have to fuss with it. I’m focusing on the Idle circuit because turning the Idle Mixture Screws is having no effect on my less than smooth running engine.

I’ve learned a few things from Randy, a good ol’ boy sharing his sense of experience on a Youtube channel. Studying the intricacies of the casting; all facets to figure the fuel trail and how it gets from here to there, I now know that there are dual routes. See if you are able to follow:

Fuel arriving from the Primary Float Bowl is flow limited by the Idle Feed Restriction (solid yellow arrow). The fuel is drawn upwards through an internal passageway (solid turquoise line arrow) to the top of a parallel downleg (dashed turquoise line arrow) to be split off to the Idle Feed Port and to the Idle Transition Slot. Small air is also  introduced via the Idle Air Bleed (solid yellow circle) to mix it up with the fuel traversing the downleg.

The dual pathway ports (dashed yellow arrow lines) deliver the the air/fuel emulsion to the Idle Feed Port and the Idle Transition Slot  in the throttle body below. The throttle body contains large butterfly valves that allow significantly more big air to mix and swirl making a combustible mixture. This [ stoichiometric ] ratio is roughly 15:1

What’s happening with my application is that the throttle valves are exposing too much of the Idle Transition Slot (lower image detail). This has the undesired effect of creating an excessively fuel rich mixture.  The Idle Feed Port is not in the game passing little or no air/fuel. The Idle Transition Slot is doing it all.  Allowing the throttle to close down further will hide most of the slot (upper image detail) The Idle Feed Port (the small black dot) will resume its function with the Idle Transition Slot properly obscured. The Idle Mixture Screw can then be brought to bear allowing for precise tuning.

Final consideration: Closing down the throttle valves will by nature reduces the air volume. The engine will stall. In order to restore adequate airflow, a hole (~.080″) must be drilled into each butterfly. It will be trial test with hole size until the ideal diameter is reached. Starting small, I want to get the engine idle RPM into a ballpark range. Fine RPM adjustment can be made with the Idle Speed set screw at the throttle linkage.

That concludes the custom setup mentioned in this post intro. Idle Mixture screws will be controlling at slow idle. The Idle Transition Slot will become effective once the throttle is part open. Keyword is transition.  A fuel main circuit gradually provides even more fuel as demand calls in a seamless progression from idle to full power. Keep it smooth.