#15 The Deep Dive Fix Files: Week Ending 11 September 2026

This is a weekly roundup of significant sports tech problems that the endurance community has encountered and, where possible, resolved. Each entry covers the issue, the affected hardware or software, and the current best fix or workaround known to the community.

Garmin Edge 1050 randomly reboots mid-ride on firmware 32.20

An Edge 1050 on firmware 32.20 rebooted twice in the final hour of a long ride, roughly 20 to 30 minutes apart, with no error message. The ride had no navigation running, standard Garmin maps, and a cadence sensor, a heart rate monitor and a Varia RTL515 paired. Moving time was 8 hours 17 minutes across an elapsed 15 and a half hours. The unit had never done this on earlier firmware.

The boot sequence after the restart looks normal, with the copyright screen and the course loading screen. The activity then sits as though Finish had been pressed without being confirmed, showing a red outline and a chequered flag icon. A second owner on 32.20 found his unit came back paused and had stopped recording, and he lost a couple of kilometres before he noticed. Check the recording state on screen after any unexplained restart on this firmware.

That second owner also saw an out of memory exception, with the Komoot Connect IQ tool, Heart Rate Zone Chart and Power Zones Indicator installed. Removing third-party data fields is the one variable a rider can change without waiting for Garmin. Rolling back is not an obvious answer either, because one owner reported the same mid-course restart on 29.22 with no update installed, and a third rider on 32.20 hit it after about 160 km on a planned route.

Garmin has asked in the thread whether a course was loaded, whether the unit performs a full boot, and whether the restarts still occur with the Varia and other sensors disconnected. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Garmin Forerunner 970 screen goes completely black while the watch stays powered on

Forerunner 970 owners report the screen going completely black while the watch carries on running. Button presses still return haptic feedback, notifications still arrive and the watch keeps recording, but nothing is visible until the unit is restarted. The owner who raised it saw it most often with the always-on display enabled.

Recovery varies between owners. Some find the screen comes back on its own after a few minutes. Others get nothing until they force a restart by holding the power button. One runner lost the display at 9 km of a half marathon and found that holding the Light button brought up the shortcut wheel but would not return him to the activity screen. Another lost the display at the start of a marathon, rebooted, resumed the race and lost 250 metres of the recording.

Wet conditions came up as a possible trigger, since the marathon runner was running in rain with a wet sleeve over the watch and screen lock engaged as usual. Owners who see it indoors and at rest have ruled that out, and for them a factory reset and a full software reinstall made no difference.

Garmin has acknowledged the reports and is collecting details in the thread: whether the display recovers by itself or needs an action, how long it stays readable between events, how long it stays blank, and what other signs confirm the watch is still running. A video of the fault was posted a day ago. Garmin support has replaced units under warranty and has also asked at least one owner to send a watch in for inspection. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Garmin Edge 850 connects to the wrong bike’s SRAM AXS gears and power meter when two bikes are paired

An Edge 850 owner with two bikes, each running SRAM AXS shifting and its own power meter, finds the head unit connects to a random set of sensors at power-on. All four sensors wake at the slightest movement, and knocking one bike while lifting the other next to it is enough to wake both. Riding away from the second bike usually pulls the correct power meter across, but the gearing rarely follows, and the AXS shifting has to be connected by hand.

AXS components and most power meters drop out after about two minutes without movement, and speed sensors take five to ten. Moving the bike you are riding away from the others, waiting out that window, and only then powering on the Edge leaves one bike awake and visible to the head unit.

A more deliberate approach is to rename every sensor in the Edge sensor list so each entry identifies its bike, then disable all of them. Before a ride, enable only the two or three sensors belonging to the bike in use. The head unit cannot then attach to the wrong gearing, whatever is awake in the garage.

Indoor trainers create a variant of the same problem that the sleep timer does not solve. A Wahoo Kickr Move left switched on at the mains never sleeps, and one rider found his Edge still searching for it as a power meter several miles into an outdoor ride after a garage warm-up. A full reboot of the Edge reconnected the bike power meter, and disabling the trainer in the sensors menu before setting off would have prevented it.

Suunto Race 2 offline map shows double blue lines and rivers that do not exist

A Suunto Race 2 owner in the Czech Republic found two parallel blue lines drawn across the map with no start and no end, and after scrolling he found them crossing a second pair of the same lines. They are visible up to the 500 m zoom level and disappear once he zooms further out. He had updated the watch and the Czech Republic map on the same day. The coordinates are 49.57900 N, 18.63088 E, and the same location in Suunto RoutePlanner is clean.

Deleting the map from the watch and downloading it again left the lines in place. A second user confirmed the fault on 7 September and narrowed it further. The lines appear in offline maps only, and online maps at the same location and the same zoom render correctly. He also deleted the offline maps from his phone and downloaded them again with no change, which points to the offline map data as the source.

Some of the artefacts carry names. At one location a watercourse labelled Studeny potok is drawn as a straight line, does not exist on the ground, and terminates at an arbitrary point. Nearby streams labelled Osoblaha and Blazicky are real. Anyone navigating a trail in poor visibility should confirm an unfamiliar blue line against the online map before treating it as a feature.

Zoom level decides whether you see the problem at all, because the artefacts appear at certain zoom levels only and vanish above and below them. The original report dates from April and the confirmation arrived in September. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Garmin Edge 850 workout beep loops endlessly when paused in the last five seconds of a step

An Edge 850 owner on firmware 32.20 found that pausing a workout inside the last five seconds of a step sends the step change beep into a continuous loop. The tone repeats without stopping until the workout is resumed. He had not seen the behaviour before 32.20.

Garmin asked whether the fault was consistent, whether it occurred across more than one workout, and whether restarting the Edge changed anything. He reproduced it with other workouts, which rules out a single corrupt structured session, and restarting the unit made no difference.

Resuming the workout stops the tone immediately, so the loop lasts only as long as the pause. Letting the step counter run past zero before pressing pause avoids it entirely. Riders who stop at junctions during interval sessions cannot always time it, and for them the alert tones can be turned down or off in the sound settings for the duration of the workout.

Firmware 32.20 reached the Edge 850 about three weeks ago. The report is open and Garmin is still collecting detail. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Garmin Fenix 8 Pro Portal watch face not dimming with always-on display since software 22.35

Since software 22.35 the built-in Portal watch face on the Fenix 8 Pro no longer enters its dimmed always-on state. The outline of the hour that used to appear in the dim version is absent, and the face runs in its full brightness form continuously. The owner had always-on display enabled, watch face seconds set to wake, wake on gesture on, an eight second timeout, touch on and touch lock on. He noticed because his battery was draining faster than usual.

He later measured the cost. Running Portal for a full day with no recorded workouts consumed about 18 percent of the battery. A second owner saw the same behaviour on the beta and on the public release of 22.35.

Other watch faces are unaffected. Built-in alternatives and well built third party faces still enter the dim state correctly on the same software, so changing the watch face restores the expected drain. That is the only fix available at present.

Software 22.38 did not restore the dim state, and a comment posted on 6 September confirms that 22.44 still does not have it. Garmin has not replied in the thread in three months, and the constant full brightness output raises a second concern about the display being driven harder than intended over long periods. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Garmin Fenix 8 Coach workouts logging as a generic run in Garmin Connect since software 22.38

A Fenix 8 owner on software 22.38 found that planned workouts from his Garmin Coach plan no longer log correctly in the Connect app. He selects the planned session on the watch, which is synced and shows the workout in the calendar, and the activity comes back logged only as a run with the run type, not under the name of the base, threshold or sprint session he selected. Workouts recorded on 22.35 and earlier were logged correctly. A second owner reported the same mismatch between the training plan and the watch ten days ago.

Garmin listed five things to check. Update the watch to 22.44. Sync after creating the Coach plan and after any change to it. Sync again before starting the workout. If more than one device sits on the account, confirm the one being used is set as the Primary Training Device. For an Expert Coach plan, start the workout from the Coach glance on the watch rather than from the activity list.

The reporter came back a day ago running 22.44. He syncs daily and starts his runs from the run activity, selecting the planned session as it appears in the calendar on both the watch and the app, and the naming is still wrong. His activities do appear in the Coach plan after syncing back, so the session is being credited to the plan even though the log entry carries the wrong name. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Garmin Edge 1050 Saving Diagnostics reboots when stopping route navigation or saving a ride

An Edge 1050 owner on software 32.20 gets the Saving Diagnostics message constantly on a unit in factory state, with standard maps and one Connect IQ app, Shimano Force Vector. It appears when he edits the data screens on his cycling profile, which then cannot be saved or restored to their previous settings. It appears on the Save or Cancel screen at the end of a ride. Loading a route created in Connect and sent to the device rebooted the unit continuously for 10 km, with the route calculation resetting to zero on each reboot and never passing 39 percent.

Garmin pointed him to the support article covering an Edge stuck in a boot loop. He had already run that reset on 31.xx, and running it again on 32.20 made matters worse, with the message appearing often enough that the reset could not complete. He also observed that the message does not appear at startup but when the software calls on the hardware. Garmin then asked him to uninstall the Connect IQ app to rule it out.

His national service team replaced the unit and reported that the RAM in the original had failed, on a device four months old. The replacement arrived on software 13.80 and worked correctly, then updated to 32.20.

Three days ago he reported that the reboots have returned on the replacement, less often than before. They occur when he stops route navigation part way through a ride, on the save screen at the end of a ride, and when he powers the unit on indoors, where it reboots repeatedly. He reads the diagnostic messages as GPS related, and the behaviour started two software versions ago on a device that had worked from purchase. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Wahoo ELEMNT Roam v1 switching itself off mid-ride while showing plenty of battery

A Wahoo ELEMNT Roam v1 has started switching itself off during rides. Switched back on, it reports plenty of charge. The owner describes no gradual decline in capacity beforehand. The unit went from completing all day rides without trouble to shutting down after less than an hour.

A reply attributes the shutdowns to battery end of life. Internal resistance rises as a cell ages, so the unit can read as fully charged while a high current draw pulls the terminal voltage low enough to trigger a shutdown. The charge indicator therefore does not clear the battery. A Roam v1 bought near launch is around five years old now.

A factory reset was suggested on the basis that a corrupted file can produce the same symptom. He ran one and reported today that the unit still switches off after about five minutes.

Wahoo support is the first step for anyone seeing this. For those prepared to open the case, iFixit carries a screen and battery replacement guide for the Roam WFCC4, written because there is no service documentation for the unit from Wahoo.

Garmin Forerunner 970 triathlon brick workout starts the transition automatically when the bike timer ends

Garmin Coach triathlon brick sessions on the Forerunner 970 start the transition automatically the moment the planned bike time runs out. The owner who raised it had a three hour ride followed by a 55 minute run, and the watch transitioned while he was still 3 km from the point where he wanted to start running. He had Transitions enabled and Auto Sport Change disabled, so neither brick setting prevents it.

The thread has been open for over a year with no change in behaviour. A second owner posted three months ago that he had deliberately ridden longer than the suggested time, ended up with a 20 km transition segment, and could not undo the transition once it had started.

Pressing the lap button to enter the transition early looks like the obvious workaround. One owner reported eighteen days ago that doing so breaks the file, so the workout stays on the watch and never syncs to Garmin Connect.

For sessions already recorded with the transition in the wrong place, thesplit.fit was posted in the thread seven days ago. It moves transitions backwards or forwards inside a multisport FIT file after the event. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Garmin Fenix 8 XC Ski Dynamics stopped recording with an HRM-Pro Plus after software 22.35

XC Ski Dynamics stopped recording on the Fenix 8 after software 22.35. The owner who raised it uses an HRM-Pro Plus, had the data on 24 May 2026 and nothing from 31 May onwards, and reset the cross-country ski profile and ran two separate sessions without getting the metrics back. A Fenix 8 Solar owner reported the same loss with both an HRM-Pro Plus and an HRM 600, naming software 21.39 as the last version that recorded the data.

Garmin asked for the usual sensor checks: confirm the strap is paired and transmitting before the activity starts, check its battery, confirm a supported cross-country ski profile is in use, remove and re-pair the strap, and update both devices. A later reply asked whether FTP and power zones had been set, since XC ski power depends on them. Neither owner recovered the data after moving to 22.41, and running dynamics from the same strap kept recording normally throughout.

The only workaround offered in the thread is to record the session as a run and change the activity type to cross-country ski in Garmin Connect afterwards. Garmin – PW posted on 9 September that an update to resolve it is being worked on, and asked owners to bookmark the thread for further news. This issue remains unresolved — if you have encountered it and found a fix, the discussion is open on r/the5krunner.

Wahoo KICKR Headwind minimum and maximum heart rate will not change in the Wahoo app on iPhone

KICKR Headwind heart rate control can stop responding in the Wahoo app on iPhone, leaving the fan tied to a minimum of 115 bpm that is too high to start it during a Zwift race. The owner who reported it has a Garmin chest strap paired to the Headwind, and found that the settings menu opens but the arrows that raise and lower the thresholds do nothing.

The prerequisites were all met. The Wahoo app was connected to a powered-on Headwind, which is required before the settings menu becomes available at all. The strap was worn and transmitting. The firmware update that added adaptive cooling, which lets the fan respond to heart rate and power together, was already installed.

Another owner tested the same sequence on his own Headwind and saw the arrows fail in the same way, which rules out a mistake on one unit.

Selecting forget device in the Wahoo app and pairing the Headwind again fixed it for him, and for the original reporter, who confirmed the thresholds were adjustable straight afterwards.

More from last week’s Fix Files roundup.

Browse every issue in the archive at the Deep Dive Fix Files hub.

Last Updated on 11 September 2026 by the5krunner