How to rationalize buying a Tesla

I’ve had many reservations about Tesla over the years and have never considered buying one. However now my wife and I are looking for an AWD EV that isn’t a ridiculous beast and can still tow a small travel trailer. We love our Bolt but it can’t do all of that. We’ve had our collective eye on a couple of EVs for this purpose and, now, the Tesla Model Y Long Range is in the running. Since we’ve started seriously looking at the Y I have to admit it’s pretty compelling. Efficiency, size, charging, comfort, safety, air filtration, range, cargo space, tow capacity, useful tow software, and no hard sell all combine into just about our ideal vehicle.

Here are some of the soul-sucking rationalizations we have to get past first.

Friday the 13th

Until January 13, 2023 none of these rationalizations mattered; with a roughly $65K price tag the Model Y never even made it on our list. But three things combined on that day (or maybe it was the day before…) to make the Y look a lot better:

  1. A massive price drop
  2. Potential for a $7,500 tax break
  3. The realization that there wouldn’t be a marginally affordable SUV-style vehicle with AWD and adequate tow capacity available for the foreseeable future. We had our sights set on the Chevy Blazer EV but they still haven’t released specs on the tow capacity and the pundits, including Chevy sales folks themselves, are saying 1,500 pounds. Unbelievably the same number that has been officially released for the smaller and later-to-market Equinox EV and not enough poundage for us. Our backup plan was the Fisker Ocean. Nice looking car, great towing capacity of 4,000 lbs, but much more expensive than the base Y when you add in all the options to reach parity and ineligible for the tax break until (if) they start making it in the US. Tesla scores some points here for a simple and sensible options story.

Even with all this the Y is an embarrassingly expensive vehicle. Now for the remaining rationalizations.


This is America, why worry about yet another rich, entitled oligarch?


I wonder what Nikola would think. Like Apple before it, Tesla has successfully transitioned from scrappy start-up “breaking all the rules” to bona fide cult. God save us from the rabid fanbois. I very much do not wish to be confused with one of Elon’s disciples but like any good Christian I can pick, choose, and manipulate to get what I want out of the Doctrine.

0-60 Time

No one needs to go from 0 to 60 in 3.5 seconds on a public US roadway, ever. Auto manufacturers, please stop leading with this ridiculous specification. And drivers, please get your adrenaline fix on a closed track where you can show off Track Mode to all your Mario Andretti wannabe friends. Maybe Tesla will sponsor you on the NASCAR circuit. And for Elon’s sake stop posting YouTube videos giggling into a GoPro camera with your cheeks pressed against the headrest.

For me, Tesla offers Chill Mode. I’m good.

155 MPH Top Speed

Really? Where?


Misleading marketing names like Autopilot and Full Self Driving combined with spectacularly and repeatedly overpromised / underdelivered FSD threaten to tank any hope we have of mainstream autonomous automobiles. That’s OK for now as long as adaptive cruise control works while towing. But I’m gonna be a (more) bitter old man if I’m not able to trade up for autonomy when it’s time for me to hand over the keys.

Phantom Braking

About that adaptive cruise. In a Chevy Bolt, less than half the price of a Model Y, I can set the cruise speed and the car will reliably cruise at that speed. Simple enough. If I want I can turn on adaptive cruise and it will cruise at that speed until I encounter a slower car in front, at which point it will follow that car at a polite distance. Perfect.

In a Tesla I have Autopilot. It might do everything the Bolt can do. Or it might decide to stop in the middle of the highway when it gets confused. There is no way to simply set a cruise speed without all the other automated nonsense that introduces phantom braking. My only recourse is to

  1. wait for Tesla to eventually fix the problem. But this has been going on for at least a couple of years now. Or
  2. drive on Autopilot with my foot continuously hovering over the accelerator pedal, always ready to override any phantom silliness. That sounds more exhausting than driving without cruise control. Or
  3. drive without cruise control.

I’d be satisfied for now if they simply added a cruise setting that does nothing more than stick to the set speed until I tell it not to. Like every other car on the planet. Let’s call it “Bolt Mode.” Maybe that’s too embarrassing for a company that’s been telling us for so long how genius their Autopilot tech is. Tesla is seriously screwing up on this one.

Proprietary Superchargers

Tesla’s charging infrastructure is impressive and orders of magnitude more reliable than the CCS charging mess we have today. But superchargers aren’t everywhere, especially not state highways in rural parts of the country. Our favorite route between Boise and Bend is a lovely drive in the Bolt but it doesn’t pass a supercharger.

I might buy that Tesla’s mission is to “accelerate the advent of sustainable transport” when my Bolt can charge at a supercharger just like a Tesla can now charge at any CCS charger on the planet. Now that Tesla has an opportunity to land some of that sweet federal infrastructure money it might actually happen.

Didn’t we already know how this was going to end? Not with proprietary Ford and Chevy filling stations. Thanks Tesla for throwing us backward a century. But while we wait for the inevitable, having elite access to every charger in the country is a sweet deal indeed.

Tesla Service Centers

It’s difficult to imagine Tesla can compete with ubiquitous legacy car dealerships when service is needed. While it’s encouraging to see more and more service centers pop up– they even have one in Boise now– reviews I’ve seen for them aren’t particularly good. To be fair, neither is my experience with most legacy service centers. I’m willing to roll the dice on this one.

Rear Charge Port

Nothing says badass like backing in to a parking spot. But it was gonna happen on its own, did Tesla have to encourage this behavior with a rear charge port and short supercharger cables? I’ll just drop the trailer when the rare nose-forward charging stall isn’t available. I guess the prophets at Tesla missed that Nissan had already figured this one out.

No 360 degree Backup Camera

Both of our current cars, a Chevy Bolt and Nissan Rogue, have a fantastic 360 degree backup camera. Seems like Tesla could merge all of their Autopilot camera images into a similar 360 view. However for towing, having those two rear views on the Y available at any time is a huge advantage. And it may actually be better than a 360 view when backing, we’ll have to see.

Few Physical Controls

With voice commands and a reasonably well designed UI with quick response time on the center screen, it’s not as bad as I imagined. Far from our preference but we can live with it.

No Passenger-side Lumbar Adjustment

On trips together my wife will be in that seat. I hope she can find the right pillow.

No Driver Display

There is a reason every other car in production has a display in front of the driver. Our Chevy Bolt has a beautiful, functional, well designed display in front of the driver. Cars substantially less expensive than the Bolt have a display in front of the driver. The much more expensive Model X has a display in front of the driver. WTF Tesla? Of all the ways they try to be different for the sake of being different this might be the dumbest.

Cocking my head to the side to see basic information like my speed was annoying in the test drive. If it remains annoying I could always install one of several aftermarket driver displays.

Fake Wood Grain

My parents’ Plymouth station wagon in the 70’s had fake wood grain. I think that says it all. But I gather I can upgrade the dash to carbon-fiber if I begin driving like my dad. Better get that on order.

No Android Auto

Another in the category of how not to differentiate your product. Pay more for less is not a winning strategy in my book. Tesla appears to suffer the delusion that their infotainment system is so capable and the UI so stunning and performant that there is no rational use for something like Android Auto or Apple Car Play. Fisker went down this rabbit hole with a quote attributed to Henrik himself, paraphrased “…with such a beautiful and intuitive infotainment system on the Ocean, why would we bother with AA or Car Play?” Oh the arrogance. So to use my preferred EV routing system I have to fiddle with my phone while driving. So much for Tesla’s (and Fisker’s) claims about safety.

Still, I can always duct-tape my phone to the side of the Y’s infotainment display and dangle the power cord down to the USB port that’s buried somewhere inside the center console. I wonder how that squares with Tesla’s “minimalist design aesthetic”.

Fart Noises

Funny, whimsical, or just plain dumb? I can watch Beavis and Butthead at home. Or Elon’s Twitter feed.

No Physical Key

Most modern vehicles, EV or ICE, have some form of electronic key system and they all need a backup to get inside the vehicle in case the battery dies– a small physical key tucked stealthily inside a key fob. As far as I know only Tesla backs up their battery-dependent electronic key system with another battery. So what do you do when you need a jump because the low voltage battery is dead? Why you pull that spare 9 or 12 volt battery out of your back pocket to open the frunk. You still won’t be able to get into the cabin of the car but you can at least get to your emergency starter battery and attempt to jump the car.

Oh but you don’t carry a battery with you? Call someone with a battery. No cell coverage? Smash your $1,000 iPhone with a rock to expose the battery. And make sure you have the Tesla key card on you so you can start the car because your phone is now in pieces. I’m sure MacGruber would have a few more suggestions but realistically at this point you’re SOL until someone drives up with jumper cables.

Of course there’s a workaround: stuff a small a23 12-volt cylindrical battery inside the cover for the front tow hook, where the two leads live for the emergency frunk release. Rube Goldberg would be proud.

Tow Package

Finally, the rationalization that almost couldn’t.

3,500 pounds is enough capacity to safely tow a Safari Condo Alto that maxes out at 2,700 pounds GVW. So far so good. The Y is wired from the factory for a brake controller necessary to activate the trailer brakes on the Alto. Excellent. The factory tow package includes trailering software that detects trailer sway and activates those brakes accordingly. Fantastic.

Then I read the fine print in the owner’s manual, conveniently linked from the Model Y web page (I sincerely appreciate that).

Seems innocuous enough. Until you realize the receiver on the Y sits pretty low, around 14″ from ground to the top of the receiver. Many small travel trailers, including some Altos, need 20″ or more between ground and the the top of the coupler to keep the trailer level. When you add up the 14″ hitch height, 3/4″ rise, and 2.75″ ball height it’s still more than two inches shy of the mark.

That seems like a strange specification even for a sports car pretending to be an SUV. I checked the manual for the Cadillac Lyriq, comparable in many ways to the Y, and saw no such restriction for their 3,500 lb capacity tow hitch. (Prohibitively expensive and, well, Cadillac). So why the restriction? Can the tow bar really not handle the added rotational force of a rise beyond .75″? Is it a question of clearing the plastic bumper on the Y? Impossible to tell from a dumbed-down blurb like this in the manual. But some engineer somewhere presumably thought this was a necessary limitation. So I asked our Tesla sales representative and got this reply:

… like you referenced the manual states the rise limit is 3/4″. I understand if you need to look else where for a vehicle that can match your towing capacity needs. Tesla’s were not built to be an all inclusive towing car. They were built more to be a computer on wheels. If there is anything else I can assist with please let me know.

I see, because a Tesla isn’t really a car it doesn’t have to abide by those pesky car things like a tow package that can actually tow something. That’s a $250 lesson I’ll never forget.

Then I contacted Tesla on their toll-free number. After the Customer Service guy lost his place in his script and asked again “who am I speaking with?”, he launched into his own mumbo-jumbo about Tesla’s strange spec. “blah blah blah user discretion blah blah designed for towing things like lawn mowers.” I pointed out that 3,500 lbs is an awfully big lawn mower and that nowhere in the manual does it specify what can be towed behind the Y. He agreed to put me in touch with someone who could answer my question. I’m not optimistic. Both of these folks were super polite. But c’mon guys.

After scouring You Tube, the Tesla and Alto groups, and various other places on the web I realize that the consensus is to just ignore this limitation. Which I get, what else can you do if you want to tow a travel trailer with your Y? And of course to a person they all enthuse about how great the Y is at towing their trailer. Which I also get. The thing about exceeding tow limits is that everything is fine until you have to do something unexpected, like stop in time to avoid that moose in the road. If we total our Tesla and trailer in a tragic moose incident will insurance pay out when they notice we’re using a riser that doesn’t even come close to meeting the mysterious specification clearly called out in the owner’s manual?

Turns out, for the Alto F1743 we’re interested in, the top of the trailer coupler reportedly sits at 17.5″ above ground when level. And 14 + .75 + 2.75 = 17.5. Even with some suspension squat this is probably close enough, assuming my Y and Alto numbers are correct. So we may have dodged a bullet here. We won’t know for sure until we have both vehicles in hand.


We’re beginning to see some competition in the US for a smallish electric SUV with enough capacity to tow a small travel trailer. The Cadillac Lyriq and Fisker Ocean come to mind. But in an apples to apples comparison including some of the features we care most about– range, AWD, tow capacity, adaptive cruise, winter package, and price– nothing comes close to the Y and I think won’t for the foreseeable future. However purchasing a vehicle from Tesla feels a lot like buying a car from Microsoft if Microsoft made cars. If you like what they’re selling, good. If you have any questions or problems, good luck. Wish us luck.

Checking In on Electric Trike Conversion

About a year ago I finished the second iteration of the torque arm for the electric motor on Lisa’s HDQE. Since that time the trike has seen about 500 miles including a trip along the Trail of the Couer D’Alenes. I recently removed the motor to see how the new torque arm is doing. There is some wear in various places but overall it seems to be holding up well.

My biggest concern was that the stainless steel pin would elongate the hole in the aluminum torque arm in which it rests. I see very little wear where the inner diameter of the pin meets the torque arm so that doesn’t appear to be an issue. However the outer shoulder of the pin seems to be carving a chunk out of the torque arm. This is happening because the shoulder protrudes slightly beyond the suspension arm. This wouldn’t be so bad except the end of the pin is very close to the brake rotor. In fact you can see where the pin is beginning to cause a very light scratch on the rotor. It looks to me like the wear on the torque arm slot is pretty much finished and the impact with the rotor is minimal, I don’t see this as a problem moving forward. If it does turn into a problem I’ll need to shave a few mm off the end of the pin.

The picture below shows the wear on the slot in the torque arm. The bottom arrow shows where the bolts for the brake rotor have been worn. I think this was caused by the previous torque arm when it failed.

Top arrow shows wear from pin, bottom arrow shows wear on brake bolts
Stainless steel pin that torque arm fits into, shows wear marks in shoulder
Pin comes very close to brake rotor

Another issue is that the torque arm is rubbing on one of the suspension arms. We knew about this problem, I had to do some filing to get things to seat properly after Terry handed the final version off to me. Maybe I need to file a little more because it’s obviously rubbing but I don’t think it’s causing any real damage. And I’m at about the limit of what I can file down, the tolerances are super tight in this area, so I think I’ll leave it alone.

Shoulder of pin protrudes slightly beyond suspension arm
Where suspension arm rubs torque arm (top arrow), marks on rotor from pin (bottom arrows)
Where torque arm rubs suspension arm

A final issue is that there appears to be a little more wear on the drive tire. It’s not enough to be of concern but next time I fix a flat I might swap the fronts just to even things out a bit. Maybe I need to re-adjust the toe-in, I’ll do that too. But my money is on Lisa’s affinity for throttle.

More wear on drive tire (right)

Schlumpf: Not dead yet at 10,000 miles

9,949 miles to be precise. I was on a short tour along the Trail of the Couer D’Alene’s in Northern Idaho when I noticed some odd vibrations in my drivetrain. Nothing audible but it felt like something was caught in my rear derailleur cage. It felt the same in either gear on the Schlumpf which reinforced the idea that the problem was elsewhere. I couldn’t find anything amiss anywhere in the drivetrain and eventually it went back to normal, smooth operation. This happened a couple more times but always snapped out of it and eventually stopped acting up.

I’m pretty judicious about lubricating my Schlumpf SpeedDrive using their lube though I gather most any kind of lube will work. If anything I over-lube. When I got home I squirted in 2 ml more lube, I was pretty close to the 6 month mark, and went on a local ride. Climbing the first hill, the scuzziness in the drivetrain was back. This time when I took a look there were metal shavings all over the chain. As I limped home it never “snapped out of it” again. On closer inspection at home I noticed a whole bunch of lateral play in the Schlumpf. I’m pretty sure it’s done.

I contacted Utah Trikes who had installed the Schlumpf about six an a half years ago, hoping that they were a repair shop for the hub. They’re not. They told me a new hub would cost around $800 and is currently on backorder, it might take several months to get one. I emailed the factory in Germany and so far have only received the “we’re really busy, expect delays” reply. So much of that these days.

I was joking just a few weeks ago on this blog about how much I liked my Schlumpf but that my tune might change if it failed tomorrow. Tomorrow has arrived and I think I’m done with Schlumpf. It bothers me that a piece of equipment this expensive apparently can’t be repaired. At least not in a timely matter. 10,000 miles would be a decent lifetime if it wasn’t a throw-away when it finally died. Maybe it’s not as bad as this, I’m still investigating options, but initial indications are not good.

It doesn’t help that this is the last in a long line of issues I’ve had with internally geared hubs (IGH). After initially loving DualDrive, Sturmey-Archer, and Alfine hubs I eventually gave up on them all. In the case of Alfine it gave up on me. I’ve slowly come to the conclusion that the relatively minor advantages of an IGH compared with a DI2 externally-geared system are far outweighed by their cost and reliability issues. I think at this point the only thing that would bring me back to IGH would be belt drive and that’s just not a thing for recumbents. If somehow that did happen, Rohloff is about the only hub I’d consider. And with that I’d be going from zero to two mechanical cables which would be a pretty tough sell for me.

For now I think I’ll slap a conventional bottom bracket on and ride with my single chainring for awhile. The gear range is fine for my commute without overdrive and gravity will suffice for now on downhills. Once removed I may try to disassemble the Schlumpf myself and see if I can fix it (but can I get parts?) or maybe I’ll send it to Utah Trikes and see if they can get it repaired through the factory. I don’t know.

Maybe I’ll add a motor– I’ve been kicking that around but was stuck because with the Schlumpf I can’t add a torque-sensing bottom bracket for pedal-assist. Seems I don’t have that constraint any more. With a motor I’m not as concerned about the low end of the gear range so I could add some teeth to the chainring to improve my top end without resorting to my least favorite piece of cycling technology short of chain tubes: the front derailleur.


The problem wasn’t the Schlumpf. It was my idler that is now getting new bearings. But it wasn’t even the bearings though they are a little rough– the chain was getting wedged between the cog on the power-side idler and the outer plate of the idler. This has never happened before and I’m not sure how it’s even possible, will have to investigate. In the meantime I think I’ll send the Schlumpf off for an “overhaul” since I’ve already removed it from the trike and it is pretty sloppy at this point.

The good news is that my fears about servicing the Schlumpf have been allayed. The Schlumpf factory got back to me and offered to fix or replace the hub. They sort of implied for free (aside from shipping) but that may have just been a bad translation. They also gave me the name of a shop in Seattle that services Schlumpfs, thinking I may send it there.

So for now the Schlumpf remains my only success story with internal gears. But it’s a huge success: two front gears with near-perfect spacing for my needs, crisp shifts, no clumsy front derailleur, no cables. And hopefully still going strong at 10,000 miles. I’ll report back on the outcome of the overhaul in a future post.

E is for Electric

Behold the HDQE. The HD refers to the current-model ICE Adventure HD frame. The Q refers to its origin as a 2008 Trice-Q of which only the boom, suspended rear section, and rear derailleur remain. E is for Electric.

This trike has undergone many changes over the years but this is by far the biggest. The result is perhaps the ultimate electric pedal-assist trike, a bold claim considering I could have simply purchased a complete electric-assist trike from ICE.

Better than ICE E-Assist?

ICE’s integration of the Shimano STEPS mid-drive motor into their trike line is impressive. STEPS itself is a proven geared mid-drive motor with integrated torque sensor yielding smooth, natural pedal-assist and ICE has integrated it into their boom in typical elegant fashion. Another big point in its favor is that STEPS integrates with Shimano Di2 electric shift, currently my favorite piece of bike technology. I’ve been running it on my Sprint for years. I’m still sad that ICE doesn’t offer a retrofit kit with STEPS. It would be so easy, just swap out the boom. And the combination of STEPS and Di2 provides automatic shifting, a perfect solution for my mom who has never mastered that art. But alas their web site makes a point of saying it is not available as a retro fit kit. When I inquired with the factory about this a few years ago they said simply “We only obtain Steps for OEM fitment” and suggested I try Bionx. Which I did for a different trike and it turned out great. There is much not to like about Bionx– a dealer is required to make even the simplest changes to the system. Oh and they don’t actually exist anymore. But my experience with Bionx sold me on a direct-drive hub vs. a geared mid-drive motor. So why do I think the HDQE is better than an Adventure with E-Assist?

  • I could upgrade my wife’s existing trike so it didn’t cost more than $8,000 to electrify.
  • Direct-drive hub motors are nearly silent, significantly quieter than geared motors. It’s difficult to explain why this matters so much but it does to me.
  • Direct-drive hubs are simpler than geared motors with fewer moving parts. The bearings are the only parts particularly prone to wearing out and they are easily replaced.
  • Hub motors don’t use the drivetrain. This has a number of advantages: Less stress and wear on the drivetrain, no interaction between shifting and the motor, and propulsion that is independent of the drivetrain. And with the hub motor on a front wheel, increased traction from propulsion on two different wheels.
  • Direct-drive hub motors allow regen. In addition to extending range it’s a fantastic way to slow the trike when braking. Especially when you have crappy Tektro brakes that squeal embarrassingly.
  • Ability to choose any battery and PAS sensor
  • Ability to configure the system any way I want, whenever I want.

Of course there are down sides to a direct-drive hub motor:

  • It is heavier and bulkier than an equivalent geared motor. I ride a trike, my mantra is that weight is the enemy unless it adds something useful enough to justify the weight.
  • It adds more drag when “off” because the motor can’t be mechanically disengaged.
  • Together, these two are a bit painful for me. They effectively mean that once electrified I’m obligated to ride with the motor on to offset the added weight and drag. But I’m not sure that isn’t just as true for a mid-drive system. Anyway it’s a moot point for this project. This is my wife’s trike and she’ll always ride with assist.

Here’s a more thorough discourse on the advantages and disadvantages of hub vs mid-drive and geared motors. For me, the short answer is: The HDQE is better than E-Assist because it’s cheaper, quieter, and field-programmable. The challenges were integrating this tech as cleanly as ICE has integrated STEPS and getting the pedal-assist working smoothly so that it feels just like normal pedaling but with more power. I think I succeeded on both counts.


I’ve been toying with this conversion idea for some time, much to my wife’s dismay. I got my experience with Bionx by converting my mom’s ICE Adventure to an e-trike with a Bionx rear hub a few years ago. The conversion was straightforward and the result was quite good. If Bionx hadn’t gone out of business I probably would have done the same thing again. But throughout all of this I’d been eyeing some other options that would give me more control over the system. One was Falco which made some impressive-sounding claims. 5 phase drive technology! Ultra-reliable! Ultra-fast Operations! (I have no idea what that means). But at the time a couple of years ago I couldn’t even figure out how to order one and decided it was more marketing than product. They still have a web site, maybe I was wrong. After my mom’s Bionx trike was stolen (from the rear patio at her retirement home, WTF?) we got her into an upright electric trike that uses an EBO hub motor. It’s a reasonable option and the company operates near home in Denver but the options seem limited, especially for pedal-assist (PAS) which appears to be limited to the basic approach based on crank rotation rather than measured power to the pedals. Recently a friend turned me on to Golden Motor which also seems reasonable but similarly limited.

What I kept returning to over these years of mulling and brooding was Grin Technologies out of Vancouver, BC. This is a very different type of operation than any other I’ve run across. It appears to be run by a small but talented band of technical geeks. Instead of marketing propaganda you’re bombarded with detailed, accurate technical, educational, and support information. As an engineer myself I’m always drawn to thoughtful engineering over marketing BS and these guys deliver. The amount of information was initially a little overwhelming even for me and I was put off slightly at the idea of having to grock it all when I could much more easily just go buy a Bionx (incidentally also based in Canada before their demise). But the benefits are enormous. It’s an insanely open approach to electrification– they produce controllers and displays that work with virtually any components (motors, battery packs, etc.) available. They stock a curated set of these components married with their controllers and glue technology into complete retrofit kits. They have staff that understand the products they sell and the market they’re in. But what finally sold me was the second iteration of their own Grin All-Axle Front Hub.

While Grin sells and supports all the usual suspects in Chinese hubs, I’ve been slightly troubled by reports of reliability and fit-finish issues. In general I think they’d work fine and the price is comparatively right, but this Grin hub is impressive. Even if it had no real advantages (it does) it would be worth the higher price just for the aesthetics. I’ll let Grin articulate the advantages, they do a pretty good job without resorting to hyperbole.

Asymmetrical front drive, Disc brakes, Front suspension

ICE Suspension Kingpost

Grin designed their All-Axle hub to work with virtually all axle types and routed the power cable through the axle so that it can be mounted single-sided as the front hub of a tadpole trike. They even talk about mounting two motors, one on each front wheel, but that’s overkill for my application. I’m looking for pedal assist not a race car. So I decided to mount the motor on the right wheel. This required building a new right front wheel with the motor. And replacing the right drum brake with disc. And since it would be pretty wonky to have one front drum brake and one front disc brake, I rebuilt the left wheel with a disc brake hub. And I’d been itching to replace the mirrored Tektro hydraulic brakes on my trike with Shimano hydraulics so I moved the Tektros to the HDQE. And the mirrored Tektro brakes happen to be compatible with ICE’s front suspension kingposts so, why not, I upgraded to front suspension as well. Whew!

Left and Right Hubs


I’ve built quite a few 20 inch wheels now in my trike career. Grin made this build the easiest yet with well-placed drillings on the hub and excellent tools and advice.

And now I will continue to barf out various artifacts from this effort with little explanation because in the nearly two years I’ve procrastinated in finishing this blog post I’ve forgotten most of the details. I blame the pandemic. And the ensuing supply-chain issues, everyone’s favorite excuse for… everything.

Grin dish offset to match original ICE contact patch
Motor wheel (right)
Left wheel
Left wheel (Spocalc)

Buyer’s remorse and a custom axle

The ICE OEM axle is 12mm so I ordered Grin’s 12mm axle inserts. However the dimension of ICE’s suspension kingpost and Grin’s motor are such that the alignment of the disc brake rotor on the motor winds up about 10mm outboard of where the kingpost puts the centerline of the disc brake caliper. 10mm doesn’t seem like much but it’s way more than I was willing to shim the caliper outward to center it on the rotor. I fully expected to be testing Grin’s return policy until we discovered we could make up the 10mm by removing the 12mm axle inserts and sliding the now 20mm axle opening of the hub over the shoulder of the kingpost. Genius!

The “we” I’m referring to and the genius in this story is my friend Paul, a consummate tinkerer and fabricator of all things mechanical and electrical. And now Paul had a new project: fabricate a 20mm axle and end cap that allowed this little slide-over maneuver and prevented the hub from sliding outward or (more interestingly) inward. But hey this was a much easier lift than the complete tadpole trike he fabricated from scratch for his first aluminum welding project. Indeed after numerous Signal calls and mailings of parts back and forth (Paul lives about a thousand miles away) I had a perfectly milled 20mm axle and end cap in hand. This turned out to be the easy part…

Grin hub in position with 10mm overlap over kingpost shoulder
20mm axle, axle bolt for kingpost, endcap and bolt, some shims used to figure out exact length needed for axle

Tortuous torque arm

…now for the hard part. A drive hub needs a mechanism to prevent the stationary portion of the hub from rotating when power is applied. I ordered the Greenspeed Single-Side Adapter from Grin because it looked closest to what I’d need and they didn’t yet offer an ICE adapter. Maybe I’d get lucky and it would just work. I wasn’t lucky.

Grin adapter w/ custom plate and cutout to clear kingpost suspension

The Greenspeed adapter is built on Grin’s universal adapter that includes a machined aluminum part that mates to the spline of the hub, allowing us to remove the attached metal plate that picks up the caliper mount of a Greenspeed brake and replace it with our own plate that picks up the ICE caliper mount. However reaching the ICE brake mount would have required a decidedly long and unwieldy plate that wrapped most of the way around the hub to reach its attachment point. Instead we decided to pick up a pivot point for the kingpost suspension which was closer but still a long way from where the connection to the adapter needed to be. We could do better.

I contacted Grin and they graciously offered up the mechanical drawing for the spline assembly of their universal single-side adapter. We’d make our own! Oh the folly of engineers with a CNC machine. The idea was to mill an adapter that picked up that same suspension pivot point much more directly than our attempt with the universal adapter. The result would be a compact torque arm that was more solid and aesthetic. Yeah I said it, more aesthetic.

Pivot point where we installed a stainless steel pin

Now Paul had a new, humdinger of a project. Build up a 3D model in Freecad of our own adapter design from the dimensions on Grin’s adapter drawing, use that to program his CNC machine, and pop out a perfectly milled custom torque arm. What could be easier.

After months of Signal calls, more mailings back and forth, and prototypes first in wood and then aluminum, we couldn’t crack the code for the spline. Curse you crafty Grinners! The spline pattern always came out just a little too small or with some infinitely small misalignment at some location that we could never quite identify. Could have been due to some alignment issues with Paul’s CNC setup, incorrect Freecad dimensions, or some combination of both. I took a file to our final attempt; it was so close! After a whole lot more filing than expected I finally got the adapter to mate to the hub. But my ham-handed filing introduced a bunch of slop; it felt like the adapter would wind up harming the splines on the hub after a bunch of acceleration/deceleration cycles.

Our final attempt

Installation of Grin components

While Paul agonized over the torque adapter I installed all of the Grin components. First I made some modifications to the V3 Cycle Analyst: I removed cabling that wouldn’t be used for this install, added a cable to drive the headlight/taillight, and rewired it so the menu button on the auxiliary input and power switch activates the e-brake instead. The menu buttons on the Cycle Analyst are adequate for me and I thought it would useful to instead have a regen “paddle” like I have on my Chevy Bolt. It works quite well for this.

I installed the controller and battery mount low on the trike with a Terracycle battery mount. I routed wiring for all of the handlebar-mounted controls inside the handlebar tubes, multifunction controller on the left, throttle and e-brake tripwire on the right. This signal drives the same input as my modified “regen paddle” input so that either will initiate regen.

Because we use this trike for guests in addition to my wife and I, I have a chain easy-adjust kit installed that allows the boom to be adjusted for leg length without changing the length of the chain. This proved a challenge for the cabling running between the Cycle Analyst and the boom (to torque sensor and headlight/USB charger) because I had to add a big service loop to accommodate the longest leg length. I ran these cables inside a nylon sheath that keeps them together and in the proper shape when the boom is adjusted. It also protects them. It turned out quite nice. In fact all of the cable routing, much of it inside the handlebar tubes, is pretty clean.

Beginning Assembly
Headlight and USB charger wiring
Cycle Analyst wiring
USB Charger
Installing USB Charger into Boom
Left handlebar with multifunction switch and throttle. I later moved the throttle to right handlebar.
Right handlebar tripwire for e-brake
Modifying Cycle Analyst
Torque sensing bottom bracket
Service loop, Cycle Analyst, external cabling

Component costs

More incomplete pandemic-addled notes.

Shimano SLX M7100 Disc Brake Set (left and right)Cambria Bike$260
Shimano XT SM-RT86 Rotor – 6-Bolt 160mm (Qty 2)Competitive Cyclist$100
ICE Kingpost RIGID pair for IS disc brakes, Left side onlyRecumbent Bike Store$99
Parts list for upgrading Sprint brakes
ICE Retrofit Front SuspensionRecumbent Trike Store$954
ICE Front wheel hub ICE for disc brake 36 holes stub axle silver
Width: 65mm
Flange Width: 42mm
Flange Diameter: 71mm
Pitch Circle Diameter: 60mm
Axle Hole Diameter: 12mm
Axle Length: 100mm
Spoke Hole Diameter: 2.5mm
Recumbent Trike Store$57
ICE Headset Assembly, 2010 and onward frame (qty 2)Recumbent Trike Store$92
ICE Seat Clamps (Lower Front) (Pair)Recumbent Trike Store$44
Tektro Volans Mirrored Hydraulic Disc Brake (Qty 2)ICE
Shimano Disc Brake Rotor, 6-bolt 160mm (Qty 2)ICE
SUN RINGLE Rim Sun 20 406X18 Cr18 32 SpokeEbay$37
SUN RINGLE Rim Sun 20 406X18 Cr18 36 SpokeModern Bike$42
Fully Configurable Battery MountTerracycle$128
Front All-Axle motor, fast winding, silverGrin$5504 kg
Greenspeed Single-Side AdapterGrin$100
12mm thru-axle adaptersGrin$10
Sapim Strong 13-14g Butted Spoke 93 mm x 32Grin$48
Sapim Strong 13-14g Butted Spoke 187 mm x 36Grin$54
Baserunner L10 ControllerGrin$220
Motor Extension Cable L10_extGrin$10
B3620Li-DT 36 Volt 19.3 Ah Downtube Battery w/ mountGrin$510
Satiator ChargerGrin$250
V3 Cycle AnalystGrin$125
Auxiliary Input and Power SwitchGrin$20
ERider68_4B 4 bolt Torque Sensor for 68mm BBGrin$155
T-Lever ThrottleGrin$15
TripWire_push e-brake lever adapterGrin$39
Anderson TapGrin$20
TTL-USB Programming cableGrin$14
Parts list for transforming HDQ into HDQE

Plan B

This project started sometime in late fall. By now June was approaching and with it an inaugural ride along the Mickelson Trail in South Dakota that my wife and I were planning. Paul and I ditched the custom CNC’d torque arm and slapped on the one we’d fabricated with Grin’s universal adapter.

And it worked! I had a week or so before the ride to run the assembled trike through its paces– my work commute and some of my recreational rides that incorporate the steepest hills around Boise. I watched motor temperatures on the climbs on some pretty warm days and they never got close to anything concerning. The trike performed flawlessly and the subsequent Mickelson trip was fantastic. Using the Grin simulator I had sized motor and battery to achieve at least 50 miles on a charge. The first leg of the Mickelson trip, on crushed gravel from Deadwood to Hill City, was about 50 miles with a significant climb in the middle. The HDQE tackled it with battery to spare.

Acceleration steer

Perhaps my biggest concern with this project was the possibility that mounting a motor on just one of the front wheels of a tadpole trike would introduce “acceleration steer”, the motor equivalent to brake steer. Brake steer, i.e. steering induced by braking on only one wheel, is something trike manufacturers solved many iterations ago. However even with the best trike designs, ICE included, there is still a hint of brake steer. Or maybe it’s my imagination. But would this turn out to be a noticeable problem when the motor is accelerating?

Turns out, no. There is not even a hint of steering when you romp on the throttle. Not even in my imagination. It just.. goes. I attribute this to a solid trike design and our strict attention to maintaining the same location for the tire’s contact patch as a stock ICE wheel.

Warranty: void

It’s worth noting that when I approached ICE about this project– I needed some dimensions on the suspension kingpost– their response was rather disappointing. Instead of giving me the information I requested they told me that such a modification was not engineered or tested by ICE and would void the warranty. Seems like classic CYA, I can’t imagine a motor puts more stress on the frame than a disc brake. But this frankentrike is so far beyond any warranty anyway that I couldn’t care less.


After a season of riding that began with the Mickelson trip and otherwise consisted of occasional rides, I noticed a bunch of “slop” when engaging the motor. Turns out one of the screws on that torque arm we had slapped on at the beginning of the season had sheared off. Our instincts about that solution were confirmed. And I had a crippled trike on my hands.

Time for Plan C

Thankfully Paul still had his Freecad model. Over the winter I contacted a local machine shop that took on what amounted to a pro bono project– I didn’t pay Terry nearly enough for the time he spent. But after a couple of prototypes and several tweaks on the final version, he nailed it. It’s been working flawlessly all summer, including a recent trip along the Trail of the Couer d’Alenes in northern Idaho. But now that the season is winding down I’ll remove the wheel and see how the new adapter is faring. Terry promised me one more pass at the adapter if I needed it.

Plan C
Plan C
Terry at his CNC
EV1 and EV2

Touring with Tech

Zoe and I just finished our annual overnight trek to Givens Hot Springs.  50 miles from our doorstep to the springs.  The weather this year was spectacular:  endless sunshine, highs in the upper 70’s, though a bit brisk in the early ride out from Boise at about 45 degrees.  Temps change fast in the high desert.

As usual, I had a mildly ulterior motive for the trip:  see how the latest technology pans out in hopes that one of these days I’ll be on another big tour.  This time around the tech consisted of the following:

  • Nokia 6.1 Android phone
  • Strava app running on phone to measure total distance, for upload to LoveToRide web site
  • CycleTrip running on phone to provide route guidance
  • Anker PowerPort solar charger
  • Anker PowerCore 10000 (10000 mAh Li-Ion battery pack)
  • Schmidt SON dynamo hub
  • Sinewave Cycles Reactor dynamo charger
  • Shimano Ultegra Di2 rear derailleur

The general idea is this:

  • Charge the battery pack from the sun during the ride and at camp
  • Use the phone throughout the ride, extending its life via the dynamo hub / Sinewave charger
  • Overnight, top off the phone from the battery pack
  • Repeat
  • Top off the phone and battery pack whenever the 115 VAC power grid is available
  • Occasionally charge the Di2 battery from the Anker battery pack or 115 VAC grid (if necessary)

Here’s how it all went down, without tapping into the power grid after initially charging everything at home:

Trip Waypoint Phone Battery Anker Battery Di2 Battery
Leaving Boise 100% 4 out of 4 bars 2 out of 4 bars
Kuna (20 miles) 86% 4 bars 2 bars
Arrive Givens Still working (can’t remember %) 3 bars (after charging Zoe’s phone in Kuna) 2 bars
Leave Givens 100% 2 bars (after charging Android phone and downloading and watching an episode of “Anne with an E” in our tent, and charging Zoe’s iPhone) 2 bars
Arrive Boise 5% 4 bars 1 bar
Charge Di2 n/a Still 4 bars after charging 4 bars


  • The phone was always displaying the CycleTrip app with adaptive brightness (meaning 100% brightness in these sunny conditions).  The display never turned off because I had CycleTrip Display mode set to Always on.
  • Turning off display except during turns should dramatically improve phone battery life during ride.
  • CycleTrip (an app I’m writing) performed well.  Not quite ready for prime time but close.
  • Charging Di2 from the Anker battery appeared to work, though I’m a little suspicious that the Anker battery didn’t lose a single bar of battery life  (still 4 out of 4 after charging).  Di2 says 4 out out of 4 bars though.  And the Anker battery has way more capacity than the Di2 battery so probably makes sense.


When the CycleTrip app intelligently turns off the display (on the list of feature enhancements), I should be able to run the phone and Di2 indefinitely with only occasional access to the power grid.  Di2 has worked so well that I don’t even think I’d pack a spare battery.  As always, here’s my battery performance since installing Di2:


Bike Routing, Tricycles, and the Cinema

When my family gets together we like to see movies and ride our trikes. Sometimes we combine the two, assaulting the trails north of Denver with as many as five trikes hell-bent on making it to the movie on time. We take advantage of some excellent bike paths to avoid automobile traffic in a very congested part of town, but I have to admit I’d be hard pressed to find my own way. It takes a local guide with tons of experience riding in the area to get us through the web of interconnected bike paths. That role falls to my brother Mark and it’s about time he got some help. I’d offer but I don’t want to upset the delicate family dynamic. Better to leave it to Mark so that when we get lost I can point out his silly mistake with righteous indignation. Recently I’ve been exploring some technology that might help us avoid family counseling. Spoiler alert: it isn’t Google Maps.

Continue reading

Sawtooth Loop: The Third Shoe Drops

Day 2• Sept 4, 2016 •  Celebration Park to Bruneau Dunes •  65 miles

Day Two of the tour dawned with the threat of a recurrence of last night’s rain and no small amount of apprehension.  Rationalization was over:  we were now committed to what was most likely a criminal act.  While I stand by the convictions expressed earlier, in the grand scheme of things it’s difficult to make the case that this act of defiance rises to the level of, say, civil rights or women’s suffrage.  We were a band of middle-class white people trespassing on the property of another band of middle-class white people.  If anything, this act was embarrassingly trite.

Continue reading

Di2: One Bad Shift

After 4,863 miles of shifting nirvana (5,217 total miles on my Di2 drivetrain), today I had a bad shift.  I’ve grown so accustomed to absolutely perfect shifting with Ultegra Di2 that this was a rather soul-wrenching experience.  The reason I don’t have 5,217 perfect  miles is that it took me a little over 300 miles in the beginning to figure out that Di2 is less forgiving of lateral chain sway than a conventional drivetrain.  Once that was solved, Di2 has been a dream.  What follows is the reason my record of perfect shifting ended today.

Continue reading

Ebikes: A pedal is a pedal

A bill was recently introduced in the Idaho legislature that would treat a certain classification of electric bike as a plain ol’ bike.  This action spurred much needed debate about the legal status of ebikes in Boise, where their current classification as motor vehicles makes them illegal to use on the greenbelt.  I’m convinced ebike naysayers have no pedal to stand on.

Continue reading