This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Forerunner 645: GPS and HR inaccuracy

I've been running with my 645 for nearly one year. I actually had it replaced in October due to HR and GPS inaccuracies. Unfortunately, it seems the new replacement has the same problems. Namely the following:

1. GPS: Of course I wait every time until the GPS is locked and the ring turns green before I start a workout. I understand that GPS accuracy is limited in cities with high-rise buildings, narrow valleys or dense forrest. Yet when I run in completely open space, e. g. on a trail in a flat field, sometimes GPS goes crazy and some data points up to 1 km away from my actual position appear. Sometimes the GPS simply does not update and I see my current pace going slower and slower while I run at steady space, eventually, the lap pace is still correct because all of a sudden it jumps. So is this a general Garmin problem, a software problem or am I just unlucky and even my replacement 645 is simply a bad one.

2. Optical HR sensor: I understand that optical HR is not very accurate, in particular for high-intensity workouts. Thus, for temp and interval session I use an additional chest-strap for accurate HR data. For easy runs I normally rely just on the internal optical HR sensor. Many times though it shows complete non-sense numbers, way too low or way too high. On a recent run when I saw HR on the watch going crazy, I actually took it off and held the sensor against the sun. I kept running at steady pace and the watch would still display the same, very-high HR after 5 minutes running with the watch off my wrist? How is this possible? Did smart Garmin programmers just modify the algorithm a bit and in case of bad quality sensor data just correlate it with the cadence? I also noted that my HR went up when I was running down, so higher cadence but I could simply feel that my actual HR was going down since I was more relaxed and breathing frequency went down as well.

So I'm really wondering if I just received another poorly manufactured 645 or if this is a general problem? Having it replaced is annoying, too. It takes at least 3 weeks for the replacement to be sent and all the fitness data is lost. So it takes another month until you have more or less accurate training status on the watch. I must say that I really like the design, handling and features of the watch. If a GPS running watch is unable to deliver accurate HR and/or GPS data, it's absolute useless though!
  • Screenshot of Garmin Connect with actual route added
  • 1. I use GPS+GLONASS to increase he accuracy. I am surprised that this would still be insufficient in open field in western europe. At least I never experiences similar issues with my car GPS or cell phone. I'm wondering if it is a simple software problem or down to some real hardware issues.

    2. I think you misunderstood me. When I saw the watch showing unrealistic HR, I actually took it off my wrist. I still kept running and held the watch with the sensor directly against the sun. So I highly doubt that any optical sensor could still record any kind of meaningful data. Nonentheless, the watch kept updating the HR and showed similar strange values as when I was running. And as I said before, it pretty much corresponded to my cadence.

    Bottom line. Forerunner 645 is a compact, good-looking GPS running watch with nice functionality. Unfortunately though it fails to measure accurate HR data and GPS location, the two core parameters to determine the runners fitness. So all the nice analytics by FirstBeat are meaningless if the recorded HR+GPS data is faulty. And with a USD 500 price tag, Garmin seems to address serious runners, which actually rely on accurate data.
  • I would say optical HR inaccuracy is "normal" and due to physiological limitations.
    However, the GPS recording you posted by far is not normal. Never seen such inaccuracy with my 645 and normal GPS mode in open space. I fear you had bad luck two times with GPS? Or could it be you carry any other electronic devices with you disturbing your GPS? Or wearing some high tech clothes with metallic parts in it?
  • I've been running with my 645 for nearly one year. I actually had it replaced in October due to HR and GPS inaccuracies. Unfortunately, it seems the new replacement has the same problems..
    >
    >
    2. I use an additional chest-strap for accurate HR data. For easy runs I normally rely just on the internal optical HR sensor. Many times though it shows complete non-sense numbers, way too low or way too high.


    1. My 645 also got replaced end 2018 for horrible GPS tracking, the replacement one is behaving the same. (Garmin has got an issue with antenna / receiver architecture which is especially quite prevalent when running, but will not admit it. Not talking about accepted drifting etc)

    2. Just take note to (sanity) check your HRM strap readings as well, in my case it seems the watch will get 'stuck' on a reading for quite a while before 'jumping' again to the new value after some time, ref to picture. (3 batteries and two straps, still same behavior.) I switched to oHR approx. halfway through. Take note of the flat lines and sudden jumps (steps) up to before the switch. This is quite critical as I need to monitor my HR very carefully for certain reasons.

    I am still maintaining what I posted months ago - Garmin still can't get the basics right, it IS after all a GPS Sport watch - failing miserably on both aspects.

    ciq.forums.garmin.com/.../1462070.jpg
  • 1. I use GPS+GLONASS to increase he accuracy. I am surprised that this would still be insufficient in open field in western europe. At least I never experiences similar issues with my car GPS or cell phone. I'm wondering if it is a simple software problem or down to some real hardware issues.


    Definitely it is down to real hardware issues. Your car GPS and cell phone have between 10 and 100 times the power envelope, antenna, and probably GPS hardware capacity to work with. I agree your signal should be good enough though in an open field in western Europe; mine has never shown an error like your track in northern US or equatorial Africa.

    2. I think you misunderstood me. When I saw the watch showing unrealistic HR, I actually took it off my wrist. I still kept running and held the watch with the sensor directly against the sun. So I highly doubt that any optical sensor could still record any kind of meaningful data. Nonentheless, the watch kept updating the HR and showed similar strange values as when I was running. And as I said before, it pretty much corresponded to my cadence..


    I understood you. The sensor flashes a light and looks for a signal bouncing back at a certain wavelength. It has no mechanism to detect if it is against your skin or not, it assumes you are using it in a way it was designed for. It makes a best guess out of noisy data coming in rather than simply not working.

    Bottom line. Forerunner 645 is a compact, good-looking GPS running watch with nice functionality. Unfortunately though it fails to measure accurate HR data and GPS location, the two core parameters to determine the runners fitness. So all the nice analytics by FirstBeat are meaningless if the recorded HR+GPS data is faulty. And with a USD 500 price tag, Garmin seems to address serious runners, which actually rely on accurate data.


    Serious runners use external HRMs, and the FirstBeat analytics mostly work with that data. Any market analysis (looking at reviews) reveals this, the wrist based HR is not for tracking serious runs. It would be nice if the GPS worked better (yours definitely should), turns out the market wants lighter, smaller good-looking watches and most of the buyers aren't serious enough to notice the inaccuracies. Sadly Garmin indeed seems more interested in selling only their stuff / listening to their marketing execs' views instead of actually optimizing their customers' fitness tracking as we were expecting.