Speed Dongle Results - SCC Dongle
Background
I have a 2019 US-spec Orbea Gain M30i, with a maximum assisted speed of 20 mph. This is the original X35 system, not the newer X35+. I purchased an X35 tuning dongle from Sunshine Coast Carbon repair. Per
their web site, “The dongle works on both USA and NZ restricted bikes. You can expect the same improvement as the 25kph models which is up to 17%”. Assuming best case, 17% against 20 MPH would result in a new speed cutoff of 23.4 mph.
Brief Summary Results
The dongle does enable power to continue past 20 mph, but full power doesn’t continue for long. I get full power from 20 mph to 21.0 mph, after which power drops quickly. At around 21.2-21.5 mph, only 60-70% of the expected power (based on the power level) is being provided by the controller. For example, at the 60% power level (orange level, out of the factory, for the Orbea) only 35-40% is being delivered. Above 21.5 MPH the power continues to drop and then level off, and fluctuates in the 10-15% range. Overall I’m very disappointed. It is clear that the motor can provide more power (see below for details) but there is some other parameter that the dongle is not compensating for. Instead of 17%, I’m getting a 5% increase, from 20 mph to 21 mph.
Detailed Results
Motor and Controller and how the dongle should work
Like all X35 systems, the Gain has the controller and the battery mounted in the down tube, and the motor in the rear hub. All components of the system communicate to each other using the CAN bus standard. There is a cable that exits the hub motor and is connected near the back of the left chain stay to a cable that is routed inside the frame to the controller. This cable supports communications between the motor and the controller, and also carries the current that powers the motor.
One of the parameters that the hub transmits to the motor is speed. More accurately, it transmits the number of rotations per second, which the controller then uses to calculate speed. The controller uses this to cut off the power to the motor when the speed exceeds the programmed threshold (20MPH in the US, 15.5 MPH in most other jurisdictions). Ideally, the dongle, inserted in-line, tricks the controller by reducing or fixing the speed sent to the controller.
The controller, as its name would imply, controls the voltage to the motor. The hub motor is a DC motor; an increase in voltage results in an increase in torque and (at a given load) speed. Although the eBikemotion app has a “power” reading, that number is really just the percentage of the total power available, so 100 is the max. The Green/Orange/Red levels of the X35 system dictate the max power provided per level.
Bench Test
I put the Gain on a bike stand to see the results of the dongle with zero resistance. To measure what the controller thinks the metrics are, I used the eBikemotion app on my iPhone. For speed, I used a Garmin detector connected to the rear wheel linked to a Garmin Edge 520.
The most obvious thing is that the eBikemotion app shows that the speed is 10.4 mph at all times. If the bike is sitting in the garage: 10.4 mph. If the bike is bombing down a steep hill: 10.4 mph. So at first glance it appears the dongle is doing the job.
With the bike in the air, when I start turning the pedals, the motor starts up immediately. With no resistance, the speed also ramps up quickly, even at the lowest (green) power level. So quickly, that I can’t tell exactly when the “new cutoff” occurs, or how it occurs (abrupt or ramp down). But it was clear that above 21 mph the power was significantly reduced.
Also, no matter how fast the rear wheel spun, the power never cut off completely. Above 25 mph, it hovered around 10%.
Road Test
For the road test, I attached my phone (running the eBikemotion app) and the Garmin 520 side-by-side on the handlebars. In this fashion, I could see the true speed (on the Garmin) and the power (% of available power) on the eBikemotion app. I rode 4 times with routes from 15-22 miles, three solo and one with a friend.
On the road, the bike feels normal at speeds under 20 mph. I supposed the start-up is a little different, since the controller thinks that the bike is already moving 10.4 mph, but it is not obvious enough to tell.
Above 20 mph, full power continues from 20 to 21.0 mph. It doesn’t matter which level (Green, Orange or Red) the bike is in. At around 21.2-21.5 mph, only 60-70% of the expected power (based on the displayed power level) is being provided by the controller. For example, at the 60% power level (orange level, out of the factory, for the Orbea) only 35-40% is being delivered. Above 21.5 MPH the power continues to drop and then level off, and fluctuates in the 10-15% range.
To be clear, the power never fully cuts off at any speed with the dongle. Hammering behind my friend at 27 mph on the flats, I was getting 10%-15% (and suffering greatly). Soft-pedaling down a steep hill at 35+ MPH, 10%-15%. And there was no change based on the Green/Orange/Red level; at 27 mph, I got 10-15% in Green, in Orange and in Red.
I should also note that all of the above was reflected in how the bike felt as I was riding. It feels good to blow through 20 mph, but then almost immediately you feel the drag of having to provide more leg power. It does seem that it helped me having some power, even if only 10%, at high speeds (every bit counts I suppose).
I also see faster battery depletion, which is to be expected given that as long as I’m pedaling, some power is being provided regardless of speed. Nothing dramatic, and of course it is highly dependent on the route and my pedaling style, but on a route where I would generally hit the 75% battery threshold at 12 miles, I hit that threshold at 10.5 miles.
Conclusions
This dongle is doing something, as power continues past the 20 mph threshold. It does trick the controller into thinking that the bike is going a slower speed (10.4 mph). However, there is still a power decline (not cut-off) starting at ~21 mph and then a leveling at 10-15%.
There is no doubt that the motor could provide more power about 21 mph. This is not a case of the motor “spinning out” or running into heat, hysteresis or eddy current losses. The controller, for some reason, is limiting the voltage provided to the motor.
My belief is that the motor is sending another metric to the controller, over the CAN bus communications, that the dongle is not intercepting. Most likely this is the RPM of the motor. And that the controller has a failsafe threshold to cut power when the motor exceeds that RPM.