"Rail Rider" - a light(ish) weight e-bike legal on Amtrak trains

[ncmired]

I got the light and USB-A port wired, and put in a hot day ride up to the supermarket. On the ride (90+ F and humidity) I pushed the motor hard both ways (2kw power setting) and didn't exceed the indicated 180F core temperature. While taking the following pictures after stopping, I watched the core temp rapidly drop to 140F before turning the motor display off.

The aluminum-cased light cost approx $12, and is rated for 12-100V D.C. I'm going to lighten up the light bracket, as I overdid it a little, and also move it a little more out of hand-grab reach:

The USB-A port is likewise wide-voltage rated, and I underslung it under the bar (on the left side). The light and the USB port are connected to the 52-volt battery via a 5-amp fuse. For the main wire harness run I used a 22 gauge 4-conductor cable (two hot, two ground) from the fuse up to the bars, ending at two 2-pin JST RCY connectors. At the bars, the USB port uses one connector, and the light the other, via a bar switch.

For long-trip navigation in unfamiliar areas, I picked up a Garmin Edge Explore 2 with a 1/4-twist mount for it (or my phone in a pinch). Under normal conditions, it'll be powered off of the bike. The stand-alone ride time battery life, depending on power-saving settings, is a couple of days, and the display can be read in direct sunlight:

I chose the Edge Explore 2 over the other Garmin GPS devices, as it seems more tailored to turn-by-turn navigation and has a slightly larger screen. It doesn't accept some of the more advanced Garmin rider monitoring pickups. Screen visibility in full sunlight seems fine - I wish the motor displays were as good.

[ncmired]

On the display - there may be a slight bit more data sharing happening with the controller. I noticed that when I changed the wheel size on either the display or the controller, both sides were updated. And maybe, software updates, hopefully, makeable via the smartphone app, will update both. And new features, such as display voltage calibration - please.

Another app feature is a throttle voltage "calibration" tool, which reads the off and full throttle voltage output and sets the controller values appropriately.

[ncmired]

I rode up to get some groceries, on a route I know well, to get some initial BBS02 versus Photon impressions.

Gearing-wise, the bike is now 2 teeth lower, 38 versus 40, otherwise the bike is the same, and I tested with a full charge 52-volt 21AH battery I also know well. I've set the initial motor programming so that the motor power is "unlimited" (2000 watts) for the throttle (three levels, set at 500/1200/2000) and left the torque sensing levels stock (400/800/1200).

On the slight altitude climb way to the market, I throttled along in level 2, with a few "yeah, torque sensing still works" tests. I think the Photon motor makes more full-throttle speed than the BBS02, but it also turns quite a bit faster (190 versus 140 RPM/cadence or so). It may make speed more efficiently, but my GPS battery conked out so I couldn't verify the display speed. There, the battery voltage read about a volt better than I remember other rides, and that was after mentally adjusting for the verified 1.3 volts the display reads low. Throttling almost all of the way, and in the mid-twenties-ish, the motor core temperature read 155F, the controller 96F. Surprisingly I was able to read the fine text on the display, in slightly more time than a quick glance.

On the slight altitude drop home, I torque pedaled for a bit, realizing that the bike gearing is about a gear and a half too low (rear sprocket change). On this route are some short hill climbs, so that's where I fiddled, shifting gears on the fly, relying only on the shift sensor. On the last mile, I kicked the assist to level 3, went for top speed and to see if I could trip the temperature cutback. As said, can't yet trust the indicated speed, but it was faster than this bike should be pushed on a regular basis. During the speed run the motor core temperature climbed to178F, the controller 99F, and I didn't notice a cutback. Once I stopped riding and rolled the bike back to put it away, the motor core temperature dropped to 134F.

My initial, very preliminary take is that full-throttled speed-wise it can approach what a stock BBSHD can do. For how long I don't know, but it will at least protect itself - not something the stock BBSHD can claim, even though the motor core has temperature sensors that Bafang elected not to monitor (from what I've read). We'd need true head-to-head BBSHD versus Photon, like bike setup, rider weight, and ambient condition tests to know for sure. I suspect its sky-high throttle RPM helps the speed, but I didn't get the sense it has the big-block low-end grunt the BBSHD can muster. Bolt on the Luna's BBSHD V-2 controller and the Photon eats dust. But without torque sensing, for those that want it.

This bike has specific goals to meet - primarily size, weight, range, and true exercise. In the next few years, I'll see if it can meet them. Hopefully, whatever weak points the motor has will also be revealed. And I need to become a pedaling rider again, lighter by about another 80 pounds.

Here are some current pictures - the bike with most of the storage in place (missing the triangle bag):

The front bag is a wheelable, wearable backpack, with a telescoping handle, that anchors to the rack via a center bungee cord and some side buckles:

On the back, for small quick-grab stuff, is a Topeak RX BeamRack. With the quick-release front and rear storage removed, the bike has several grab points for loading/offloading from cars, the train, etc. On the top bar is the, needs to be charged up, GPS. The soon-to-be reorientated seat bag has tools and flat-fix stuff.

Here's the bike, unbagged, as it would be loaded on the train, and with the racks providing lifting/grabbing hand-holds:

To hold the backpack load, I doubled up on the fork bolt brackets and the struts:

The rear Topeak rack simply clamps to the seat post, and has no struts:

[ncmired]

Dunno yet Retro - I'm hoping that it will more smoothly contribute to real effort pedaling, as I find real effort cadence pedaling (set at 9 levels) very clumsy and awkward. Some love it, but I don't. But after 7 years of cadence (but honestly mostly throttle) e-biking, I've got some "getting used to" to do.

One immediately noticeable behavior difference is off of a stop. I always preset my right leg/pedal at about a 1 PM clock position and hold the position with the brakes, to aid smoother launches. With TS, there's little/no motor spinup delay after releasing the brakes, unlike with my BBS motors.

[Retrorockit]

Did you find the TS feature of much benefit?

[AZguy]

I probably would have done the same thing and pondered for a good long while before trying the dragging brake maneuver LOL

Good to hear and interested to see how it goes!

[ncmired]

I screwed up while testing the bike on the stand.

• initial config, via the app ✓
• run the motor via the throttle ✓
• check brake and shift sensor inhibits ✓
• rotate the pedal forward ... ruh-roh - nothing

So I checked the config screens - the torque sensing is enabled and set to various increasing power levels. Tried a factory reset. Crap - nothing.

Then I contacted the vendor, who asked if I had actually ridden it. Err, ah ... no.

Yeah, the torque sensing works and the motor runs while actually, you know, riding the bike AND applying some force to the pedals. What a dummy. The vendor said, and I quote, "With a torque-sensing motor, you cannot activate the motor on a bike stand without wheel resistance. You will need to ride it. This is a common misunderstanding for those with Bafang experience." Yupper, that's me - red faced and hand raised.

Maybe, just maybe, if I had dragged the rear brake caliper hard by hand with the bike on the stand, I would have set off the torque sensing. Took a short road ride and it indeed works.

More test riding and pictures to follow.

[AZguy]

It is good looking

[ncmired]

While waiting on the additional cables, I powered the system up, mainly to check for smoke and to gander the display. I chose the bigger DS103 display versus the Eggrider-style one, which I already know I can't read. There's also a horizontal 750C-like display.

The motor is very quiet on the stand and makes a muted whirring sound. I've not gone into the settings yet, via the app.

The DS103, like many of the newer displays, has a separate but hard-wired button panel with no connector in between - potentially a problem for some builds, like recumbents and bikes with long wept-back bars. As discussed in the Photon-specific post, the display is chock full of info - most of which will likely not be glance-readable underway or at all in direct sun:

Fortunately, it can be set to white text on a dark background - maybe that'll make it a little better. I think I want to be able to glance-read just:

• the speed
• the power draw when I'm striving for range
• if I'm pushing the temps, a warning (how about a submarine klaxon noise!)

But, this is my first exposure to torque sensing, so maybe more info will end up being useful.

[ncmired]

The Amtrak bike project now continues, as the CYC Photon motor kit has arrived. The motor and related mechanical bits are installed. I need to finish up the electrical stuff which is on a slight hold, waiting for a brake "Y" splitter and short extension cable due today.

The vendor kits the motor assembly and the other related parts separately - which makes sense with the way the motor is designed to accommodate varied BB widths.

A concern I've had is the chainring, which appears to be one aluminum cast and machined piece - something that I suspect Lekkie could replicate quite easily and I hope they do, filling out the tooth choices (from the current 34, 38, and 50) and perhaps even a generic spider.

CYC's online user manual lists out the complete and very usable dimensions, such as the chain line for various BB widths and even the Q factors:

Yep - this is a wiiiiide Q setup. My install is on a 68mm BB with no chainstay interference, so I didn't use any drive side spacers - I'm aiming to be close to the 48mm Alfine offset sprocket chain line. I'm also using a 1/8 chain - for which there's a lot more motor housing clearance than on the Luna 42 Eclipse./Lekkie 40 and BBSHD/BBS02 combinations. I've had to space out the chainrings and hurt the chain line some, to stop the 1/8 chain cutting gouges on the BBS motor castings.

My rough measurements are 47mm-ish chain line (which I doubt is that good), and a slightly more, to-the-left pedal offset. On narrow-frame installs like mine, a shorter spindle would be nice, but there isn't one. Here are some photos - first of the left side and all of the spacers required. I'm betting the Photon, unlike the BBS02, wouldn't be easily usable in the stoker position on a tandem - it's just too wide:

If you're gonna have fins, you might as well put 'em out in the breeze. Meanwhile, on the drive side, the secondary housing is so small it barely extends out to the chainstay, and is tucked away from the chain:

It's hard to show how much smaller the Photon's secondary housing is, compared to even the BBS02.

The motor bracket and screws are positively puny, compared to the Bafang parts - aluminum versus steel, and M4 versus M6 screws. The motor HAS to be butted up against a stop of some sort.

On the electrical side, the wiring comes out of the front of the motor. It does make the power wire length requirement that much shorter on most typical installs, which is a good thing. I've had to shorten the too-long stock power wire on all of my BBS motors, as cramming the excess up behind the motor housing is kinda clumsy. Not to mention this is a wire set you want to keep as short as possible.

There's only one wire going backward, that being the speed sensor cable (no shift sensor cable), and there's plenty of room between the motor housing for it, as well as for the normal shift/brake and whatever cabling - maybe even for carbon fiber frames. The speed sensor is attached to the frame with two wire ties and contains a Bluetooth antenna. The motor's firmware is a VESC spinoff, and there's a phone app to configure and monitor the motor - no more USB/Serial cable.

The main harness motor plug and thankfully now has both an alignment keyway and some waterproofing - up at the bar end are the four, familiar plugs we all know. And that's where I've stopped.

[Retrorockit]

I have a 650 that I would be using except working on my car I left it sitting on the exhaust manifold for a couple days (turbocharged) it sill works but he battery won't hold a charge.
NR soldered some wires to the battery. My big soldering gun is broken. I'll fix it at some point. Same strobe as the 350.

[pbreezy]

Nite riders are good lights, I ran (I think the 600?) for about five years and it was a great light. It’s cool that they’re a local company to me as well, so I could in theory walk into the actual Niterider store/help center if I need anything.

[Retrorockit]

I already committed to NR years before E bikes. It just turns out that for me the first one is the best.But not for it's intended purpose!
In SoFla bright sun, tinted windows texting, other random distractions.The Lumina 350 head on or in the rear view mirror can hold it's own. I turn my bars to the side at stop lights, but when the light turns green it's ON.

[pbreezy]

The Lezyne lights have a “Race mode” that only cycles between high and low, and the bontrager has “trail mode” which is the same thing. You access these modes by holding the power button for some number of seconds as specified in the instructions. I use it quite a bit on the Lezyne lights, but with only three modes on the bontrager I usually just cycle through the strobe.

[ncmired]

Thanks guys, especially on the front light thoughts. I know what you mean Retro, on the cycling through the modes nonsense - I have a no-name light that has that annoyance. I think mode-wise, for the front headlight I won't use during the day, off-high-low-off is enough - if there is such a light. And, since I don't shop locally, there will likely have to be some purchase "experiments" (returned via Amazon or the like). On the rear, I use the low-end Cateye blinky red lights, which seem to last years on two AAA batteries. Front light power-wise, I'm still undecided - a deciding factor may be that I'm probably going to have USB power available on the bars.