Total consumption was 15% over a 2 hour ride. I don’t have the Roam 3 but the Bolt 3. We seem to share the exact same problem after the firmware update though. I can’t tell if it was warm but it was actually quite cold that day so I guess strong sun wouldn’t be the issue.
Hi Jakub,
So, if you consumed 15%, Bolt 3 is displaying 85% remaining after 2h? On that basis I don’t expect running warm in cold ambient temps. That consumption is similar to my experience with Roam 3.
I’ve never fully discharged my Roam 3 and my longest ride time was about 8h.
Assuming displayed discharge % is linear (probably not), then 7.5%/h extrapolates to 0% at 13.3 hours. However, the displayed % is unlikely to be a linear match to battery discharge characteristic and shutdown probably occurs around 5 to 10%. Accurate tracking discharge of Li and other batteries very difficult
I guess you need to Start a ride, in a room in daylight or with light on (keep Bolt backlight on) and see how long it lasts, check and record %?
Personal view - I think quoted runtime is with default setting basic Auto backlight, Auto Max probably reduces runtime (but I’ve not felt the need to verify).
Two days ago, I rode 3h15 minutes following a gox track … I started with 100 % charge and ended up with 51 …
my roam 3 isn’t 3 months old … hopefully they’ll fix it soon.
Mind you, DC Rainmaker also showed how the battery percentage is not ‘true’ in the first place. The battery percentage is fixed for the first hour or so. In the video he says that according to Wahoo this was a deliberate choice to reduce battery anxiety. So extrapolating battery consumption from shorter rides does not really work.
Interesting. I suspect the reason why % doesn’t change much at first is technical / physics rather than deliberate design. Fully charged LiPo battery measures 4.2V and for light discharge for a long time, for less than 0.1C (10 hour life) discharge. It then has a gradual discharging voltage down to 3.7V.
Capacity greatly reduces at low temperature. So, accuracy of % display all depends whether Wahoo coding compensates for temperature, discharge rate and with accurate A to D for voltage measurement. Then they have to map voltage 4.2V to 3.7V (or their critical cutoff) into % scale.
Check out the first screenshot though. This behavior apparently doesn’t just occur around 100%.
Updated to latest Roam 3 software with ‘battery discharge fix’. Today, I did a 40 mile ride, riding time 3h 55m, let’s say 4h. Actual runtime from 100% was 5h 2min (say 5h). End of ride 48% remaining.
To summarise, from my past experience, I think it’s got worse!
Battery consumption continues, probably same rate, when paused (lunch, coffee, pre ride set up etc).
For a scientific investigation, I’m going to record a simple metric, %used per hour (%/h) based on total runtime. My ride today as example 52/5 = 10.4%/h battery used
My rides are with the following config:
Auto Max backlight, ride today dull/ overcast, so I gueess backlight power less than max.
Live tracking, Varia radar, ebike TCU, SRAM etap, cadance/wheel speed and Tickr heartrate. Running route guidance with voice prompt.
I’m going to record %/h and log date + conditions for the next 5 rides. It would be great if others could log their results ans share summary after 5 rides. We then present our ‘group’ results to Wahoo?
Investigation update, using USB power analyser:
CHARGING from the 48% battery remaining state, after last ride:
- Tested with 3m USB lead known to have higher resistance - would not take charge. Roam didn’t fire up Wahoo and charge indication after 5 min wait, voltage at Roam 3 was 4.7V. [I’ve seen reports of this problem on Wahoo forum]
- Tested with fast charger 22.5W max charge capability, 2m lead rated for high-capacity charging. Charged at 1.3A to 1.1A initially. Observation - when Roam 3 display reached 100% charged, Roam continued to consume (still charging) 130mA, after 30 min reduced to 30mA and remained at this level after 3h, then disconnected for discharge tests.
DISCHARGE ANALYSIS - note this test done after full charge, i.e. 3h at 100% displayed. The following readings are averaged as the activated Roam 3 consumption varies with processor activity demand, so the following are roughly indicative only and after 10 min of stabilisation:
Auto Max backlight: Bright torch on sensor - 0.6W power consumption. Finger over sensor for 30 sec - 0.8W
Auto backlight: Bright torch on sensor - 0.45W power consumption. Finger over sensor for 30 sec - 0.6W
Backlight ON setting: 0.45W
Backlight OFF setting: 0.32W
CONCLUSIONS:
A. Roam/Bolt 3 requires a good quality charger and low resistance (1 or 2m) USB C lead.
B. When Roam / Bolt 3 declares 100% charged, this may not be true, another 30 min on charger adds at least another 130mWh (ok probably only a few minutes extra runtime)
C. Auto backlight is counter intuitive. The backlight increases for dull days, power creases when it’s sunny. It’s a transflective display, only needs backlight at night or dull days. For sunny days backlight dims slightly. Auto may be better on sunny days than Auto Max, certainly reduces power consumption. With backlight OFF only 0.32W consumed, with Auto Max power consumption increases significantly (more than double).
Now the sun has finally arrived in UK, I’ve set the Roam3 to ‘Auto’ backlight’, may return to Auto Max for those darker days in the winter!
BTW: I recently changed the angle orientation of the Roam 3 on the out-front mount. Previously, was set to be perpendicular to my line of sight. Now changed to be near horizontal (flat), so probably 30 deg away from my line of sight. It seems more readable in sunlight and this may be a feature of the transflective display and anti-glare screen cover?
I got 14.22 hours elapsed this week with 19% remaining.
Backlight always on, live track going, lots of sensors.
I’ll try next week with backlight off.
Richie,
Your battery consumption rate 5.7%/h backlight ON. My last ride 10.4%/h backlight Auto MAX.
Oh wau, great job! Big kudos to you 
Do you think Auto mode (not the max) is better than for example Medium or Low brightness?
Options:
Auto Max, Auto, ON, 5 sec ON with button or OFF.
My understanding, subject to testing:
- Clear sunny day, not riding under shade, then OFF, or 5 sec for short tunnels
- Auto - best all round but for really dull days screen may not be bright enough. Possibly ok in full darkness at night
- Auto Max - for really dull / half light rides, but this mode consume battery capacity.
- Many sensors incding running with phone / Wahoo app increases battery drain.
- Lots of sensors, Auto Max in dull conditions, live tracking, many Ant+ sensors could lead to 10 to 14 hour runtime.
BTW - not tested yet but I believe possible to get >24h. I have a Towild CL1600 front light and USB C right angle each end short lead. This torch in medium ON runs almost a day, daytime flash probably 48h. Towild charge port is bi-directional can charge Roam 3 and run the light.
Hi Jakub,
I just added how I understand best way to set backlight. Also, photos and description how to have bike front light and Roam 3 work for a 24 hour ride.
Regards
Steve Radford
Did a 7h 41min elapsed ride today with backlight off, navigation and strava segments on (mostly on stats page, no map usage). Went from 95% to 4% almost died on the way back. I did the latest firmware update on the Roam 3 device just before the ride. Device is almost brand new.
Bolt 3:
2h7m. 100% → 74% road bike with power meter, heart rate strap.
Same settings, 1h 74% → 69% mountain biking in the woods, same power meter type, same heart rate strap, rear hub mounted speed sensor.
Both rides 26-28 deg C.
Hard to reproduce indeed. My device is on Auto Max brightness. In my case it could be the difference in ambient light response (the road ride was significantly brighter), or the fact that one is trying to use GPS to measure velocity and the other isn’t.
I guess the most obvious, well described battery consumption factor would be display brightness settings.
Beong:That’s 15.2%/hour, far above the other short runtime reports like my 10.4%/h. I suggest you charge ti 100%, leave it charging another 30min before ride. If you get similar result again, report to Wahoo. Advise backlight setting, sensors and firmware. Though with backlight off and minimal sensors should be closer to advertised 22h.
Your result could be due to a faulty battery if same in next ride?
Mry:
Your result is more in line with other reports from Auto Max users. Longer runtime / better in sun, worse when dull/ shaded when backlight increases intensity.
Good point i didnt know what this thread was about i just googled why my roam was dying and found this thread and posted without checking exactly. I just did, it was 7h 41min elapsed which with 91% battery loss comes to about 12% per hour loss. I have a heart rate monitor, powermeter and di2 sensors. I turned the backlight off after about half the ride because i noticed the battery dropped so fast.
Backlight is a huge draw I hear, so depending on what the light sensor is seeing maybe the auto count be drawing more. I’m going to try next week with backlight off.
@vjakub Auto should have better battery at the same level (auto on better than on, auto max better than max) but it will depend on what the light sensor is seeing.
Bright sunshine? More battery gains from auto.
Riding in the dark? No battery gains from auto.
If it works for you I’d use the Auto. Works fine for me on Roam, but the finish on the Ace screen is bad enough I need the backlight all the time.
almost identical to my bolt 3 today, turned backlight off, disconnected phone and di2 and battery still dropped like a stone