Tried again with backlight off.
18.0 hours, down to 19% battery life left
Ritchie,
Interesting result at 4.5%/h, Looks like your 18h would amost extrapolate to Wahoo’s advertised 25h. What was your experience of readability and weather / sun during the ride?
The 5 second ON mode should have similar result and allow brief illumination when needed.
LOL the hour I did at night with no backlight was not good for readability. Otherwise it was actually pretty decent. The Roam screen is physically way better than the Ace screen, but not as clear as the Gen2 screens.
I also was using plenty of sensors and LT the entire time
Ritchie, I think we have proven, with current soft/firmware, that Wahoo’s claim of 25h for Roam 3 is (almost) achievable with no backlight. Also, that backlight level is the primary consumer of battery capacity. Other factors such as number of sensors and consequential comms activity will affect consumption but to a lesser extent. Whilst Wahoo released a UI software version to improve consumption, there is no information on change made and I personally didn’t notice any difference under the same use conditions. Also, I should comment that our tests are on Roam 3 and Bolt 3 users have reported short runtime so other factors such as smaller battery, supply batch etc could cause a different experience.
Interesting comment about Ace since I had considered upgrading from Roam 3.
I think Wahoo support are remiss in not publishing a guidance note on use of backlight settings and implication on battery consumption. Also, in respect of improving screen readability in sunlight - I found much improved after changing incident angle to my eyes. I found flat (horizontal) with 30 to 60 degrees view much better than when I’d angled for perpendicular (0 degree).
When I get a bit of time, I plan to produce a PDF Guidance Note to summarise findings and post here as a concluding contribution to this thread.
Yeah backlight is the big draw.
Some more thoughts on this:
All the Gen 3 devices share the same UX but the Max is really intended for Ace due to the crummy screen, but is less of an issue since Ace has a lot of battery life. That really shouldn’t be a choice most people make on a Roam or Bolt.
Auto backlight works pretty well IMO, and gets most of the gains of no backlight. Honestly the only reason I use always On is over the winter commuting when I’m in the dark 75% of the time and just want it on. I have consistently forgotten to switch it back to Auto in the spring. That should be people’s choice unless they’re going for maximum battery life on a long ride or always in full dark.
I also found easy to accidently touch the right-side brightness vertical bar (with light bulb icons) - Display/Sounds menu page.
In ‘Backlight ON’ mode, three different intensity levels.
In Backlight Auto and Auto Max modes, if you catch the mid touchbox of the bar there is immediate selection of Backlight ON – mid intensity. In good sun / daylight this will consume more power than Auto with backlight off / dimmed whenever sufficient light encountered.
I accidently did this some time back, took a while to realise as I don’t visit that menu often!
So, just a possible that some of the excessive battery drain reports from other users could be down to finger trouble?
Glad I didn’t go for the Ace now that you mention the screen.
I agree with you, after switching back to Auto, only changed to Auto Max after a grim overcast UK day last winter.
Any news regarding these issues? I have a new Roam 3 and power consumption is inconsistent. One day it’s as expected (4-5% per hour), next day it’s really bad (more than 10% per hour). I can’t pin down any specific reason that causes it. It’s definitely not the backlight, in one case the battery drain continued with the backlight entirely disabled mid-ride. It might be sensor related, but I’m not sure (it seems to have started after I paired additional speed and cadence sensors).
Wahoo support hasn’t been very useful. They acknowledged the problem based on my recorded activity data, but basically ignored the strange inconsistency and that this is and ongoing issue for lots of people and very likely a software issue. They want to open a warranty case and replace my device.
The kicker is that I specifically got the Roam 3 to replace my aging Roam v1 whose battery life has deteriorated. The new unit is now pretty much worse than the old one.
I’m getting consistent results around 5.5%/h now after:
- Top up charge for 1h before ride, even though Roam 3 displaying 100%.
- Riding with AUTO backlight.
In respect of topup charge, past 3 rides, 1h 15 min before capacity started to fall below 100%.
Whilst there may be other factors affecting discharge rate, the state of charge at start could cause inconsistent results, especially when comparing short and long duration rides.
Other devices, like my Samsung phone, also report 100% well before full charge. I tested with a USB power analyser. Lithium batteries exhibit a plateau Voltage near 100% so difficult to determine fully charged. NOTE for long life, best to not over / topup charge every time.
I’ve been using the same charger and I also leave the Roam on the charger for an hour or so after it hits 100%. I know the effect you are talking about. I can’t imagine the root cause is a badly calibrated battery gauge anyway.
The difference in consumption is extreme: In one case, a long ride with around 11 hours total runtime resulted a drop from 100% to 49%. This was with display on auto backlight. A week later, a somewhat shorter ride with around 9 hours of total runtime resulted in a drop from 100% to 7%. I turned the backlight off after a few hours to save power. Then a few days later, power consumption on a shorter 3 hour ride appeared normal again.
Wahoo now definitely wants to replace my unit. They don’t acknowledge or deny that there are firmware issues. This seems so stupid to me…
Discharge down to 7% after 9h on basic AUTO setting and fully charged is extreme, so there is some other factor related to firmware and your configuration of sensors and phone / Wahoo App connection.
I had problems last year and contacted Wahoo support, after months of emails, repeated questions were asked. I eventually came to the conclusion I was corresponding with an AI bot (the ‘person’ was Cai - just thought it could be an acronym for computer AI). In the end the final response was that it would be reported to software dev team. Cai never asked logical questions relating to the problem, only advised the standard responses, reset, update firmware etc. I suspect that replacing Roam 3 is Wahoo support fallback action when they run out of standard responses and don’t know what the problem is. Given that your unit is OK on some rides, I don’t think it’s a hardware issue, so I suspect the replacement may exhibit the same issue - but let’s hope it doesn’t.
I solved my Roam 3 problem last year by testing different connections, e.g. discovered problem didn’t arise if phone/Wahoo App not connected during ride. The problem I experienced was Roam 3 functional latency in longer rides e.g. after a few hours ride, everything lagged progressively up to 15 seconds (Varia radar alerts long after car had overtaken me). I found the Bluetooth (hidden) system App on Android with ‘Optimised’ power setting, changed to full power, problem solved. Given the apparent Roam 3 processor overload (BT retries?), battery consumption increased. Just worth noting that a config on the phone can have a dramatic impact on Roam 3 functional performance. In my opinion poor software design by Wahoo as comms issues should be flagged (we get too many Radar Error reports, so they can do alerts!).
In summary, many Roam 3 users out there finding consumption OK (within limitations of backlight setting). See how you get on with the replacement Roam 3 but if the problem continues, you may have to experiment with your particular setup.
Yes, it seems obvious to me that there are specific conditions or triggers for this power bug, which means that many users probably have no issues. I just cannot find out what it is, and it is frustrating to see Wahoo not interested in debugging this either. I never have the phone connected to the Roam, so I can rule that one out, at least.
After you receive and test the replacement Roam 3, hopefully you will post an update, to advise whether random battery drain solved or remained.
Several other Bolt and Roam 3 users have posted similar battery drain experience, not heard if they’re getting replacements though!
On the rennrad-news forums (German), one affected user reported that they already got a replacement, and as expected (this is very likely a software issue) it changed nothing:
I’ve now requested a return with the seller but I decided to do some testing in the coming days anyway. Maybe it was a good idea.
It needs more data, but it looks like the Bluetooth connection with my phone (FYI, in my case it’s a Google Pixel 8a) triggers the high power consumption. As soon as I connect to the phone after booting the Roam, power consumption is elevated. It doesn’t matter if I kill the Wahoo app or disable Bluetooth on the phone later, the initial connection seems to be the trigger for high power consumption until next reboot. I have now unpaired phone and head unit and for the last couple of rides it all looks good.
If this truly turns out to be the root cause, it isn’t really acceptable of course, but at least there’s a reliable workaround and it is something I can report to Wahoo. Maybe there are compatibility issues with some phones or something like that.
My phone is a Samsung S20, your recent Pixel 8a will have a later Android version, so possible differences from commemts below.
Check the Bluetooth system App, hidden from user Apps:
Select ‘Settings’
Scroll down, then select ‘Apps’
When Apps page shown, look down page to find text “Your Apps…” To the far right of this text, there is an icon with three horizontal line. Press this icon.
A page appears, at the top enable the slider ‘Show System Apps’
You now get to see all Apps, scroll down and select Bluetooth.
Select ‘Battery’, it must be set to ‘Unrestricted’
I’ll give this a try maybe later, but like I said, power consumption is still high after I disconnect the phone, so I can’t imagine that will do anything. I don’t see any lags either with the phone connected. Pixel phones generally lack aggressive battery optimization anyway. Samsung OTOH is really bad here: Samsung | Don’t kill my app!
Whatever the case, the previous generation units don’t have this problem and Wahoo should fix it.
BT comms with Roam 3 should use BT low energy (BT LE) mode which only uses short bursts and small data packets. BT LE uses minimal power.
Classic BT uses continuous comms for large data exchange. If somehow comms is running BT classic then power consumption higher. However, I don’t know if classic BT is supported by Roam 3.
Wi-Fi is used for large data transfers and WiFi consumes more power than Classic BT.
Wild thought really but could you be running a WiFi hotspot on the phone?
I suspect once Roam 3 makes a connection, it retries to reconnect hence contined power drain?
No WiFi hotspot. Besides, I think WiFi is always disabled on Wahoo head units while activities are being recorded.
I suspect the power consumption bug is only indirectly related to Bluetooth communication. Some specific aspect of communication with the phone is triggering misbehavior in the head unit, and this continues until you turn it off. Maybe after the trigger it spins the CPU pretty hard or something like that. Unfortunately Wahoo disabled ADB support on newer devices so as a user you can’t really debug this.
Do you have (Android) Developer Options enabled on the Pixel? It should be possible to log on the phone.
In Developer Options, in the Bluetooth section, there is an option to turn on ‘HCI Snoop Log’. I set this function to ON but never worked out where to find the log file, although there is info in the link below:
If we can find the log file, I guess we could compare my log file with your log file?
I’m talking about ADB on the Wahoo Roam side. It’s Android based, too.
Anyway, I don’t really want to do low-level debugging like analyzing Bluetooth communication. This is Wahoo’s job. We’ll see if my hypothesis about phone communication triggering the high power consumption continues to be true for the next couple of rides, and if so, I’ll report that to them.