Colorado timer inaccuracy

[+g-o-cashers]

I've been running some tests to try and narrow down what causes the Colorado position error issues. While I was running these tests I had my 400t next to my 60cs and noticed that the "Total Time" field on the odometer page was consistently higher on the 400t as compared to the 60cs.

 

I added the "Time of Day" field to the odometer page so that I could compare "Total Time" to the GPS accurate time of day to see if there was a drift. After having the unit on for about 30 minutes I noticed that Total Time advanced about 14-15 seconds faster than Time of Day (0.7% error). I see similar results on 2-3 other tests. At least on my unit the internal timers seem to run about 0.5%-1.0% faster than "real" time. I see no such drift on my 60cs, or if there is, it is much smaller.

 

I ran another test where I started the stopwatch on both my 60cs and 400t and compared them after 15-20 minutes and I see similar results, the 400t gains time at the same rate as compared to the 60cs.

 

While I don't need these times to sub-second accurate I was surprised at the amount of drift. Seems to suggest that Garmin is using very low accuracy oscillators (5000ppm-10000ppm?) in the Colorado. I thought even inexpensive oscillators were 100ppm-200ppm. Or maybe it is just my unit.

 

Has anyone else seen a similar issue?

 

GO$Rs

[+XopherN71]

I'll have to take a look, what I do know is that the time remaining to next turn/destination seem to be dead on.

[+Timpat]

I ran another test where I started the stopwatch on both my 60cs and 400t and compared them after 15-20 minutes and I see similar results, the 400t gains time at the same rate as compared to the 60cs.

 

Has anyone else seen a similar issue?

 

GO$Rs

I saw your post and since I'm in the office today I tried your stop watch test. I used my 60Cx and my 300. I see a similar advance amount of time on my 300. I ran both units together on the stop watch for roughly 20 mins and my 300 showed about 9-11 seconds more than my 60Cx during 3 test runs. It is pretty tough to start both at precisely the same instant but this is close for these crude tests. It sure shows the 300 is indicating more elapsed time. Hmmm, what's up with that?

[+g-o-cashers]

I'll have to take a look, what I do know is that the time remaining to next turn/destination seem to be dead on.

 

I'm guessing these timers won't be affected because they are computed every second or so based on the speed data along your route and your distance remaining to dest/next.

 

 

I saw your post and since I'm in the office today I tried your stop watch test. I used my 60Cx and my 300. I see a similar advance amount of time on my 300. I ran both units together on the stop watch for roughly 20 mins and my 300 showed about 9-11 seconds more than my 60Cx during 3 test runs. It is pretty tough to start both at precisely the same instant but this is close for these crude tests. It sure shows the 300 is indicating more elapsed time. Hmmm, what's up with that?

 

Software timers like the stopwatch and total time are usually implemented based on some sort of hardware timer/counter. The hardware device receives a periodic clock from a crystal oscillator. The accuracy of the oscillator (usually expressed in ppm) determines how accurate the hardware device is and, hence, how accurate the software timers using it will be. It is normal to see inexpensive oscillators used in electronic devices with accuracy in the 100-200ppm range which would yield an error of about 10-20 seconds over a 24 hour period, worst case.

 

However we are seeing something that is almost an order of magnitude higher, nearly 500 seconds in a 24 hour period. Originally I thought maybe Garmin is really using a very poor oscillator, but now I'm wondering if maybe there is some sort of software bug or hardware issue because the error seems to beyond what would be caused by oscillator inaccuracy -- especially if multiple people start seeing nearly the same error in the same direction.

 

GO$Rs

[Didjerrydo]

I'll have to take a look, what I do know is that the time remaining to next turn/destination seem to be dead on.

 

I'm guessing these timers won't be affected because they are computed every second or so based on the speed data along your route and your distance remaining to dest/next.

 

 

I saw your post and since I'm in the office today I tried your stop watch test. I used my 60Cx and my 300. I see a similar advance amount of time on my 300. I ran both units together on the stop watch for roughly 20 mins and my 300 showed about 9-11 seconds more than my 60Cx during 3 test runs. It is pretty tough to start both at precisely the same instant but this is close for these crude tests. It sure shows the 300 is indicating more elapsed time. Hmmm, what's up with that?

 

Software timers like the stopwatch and total time are usually implemented based on some sort of hardware timer/counter. The hardware device receives a periodic clock from a crystal oscillator. The accuracy of the oscillator (usually expressed in ppm) determines how accurate the hardware device is and, hence, how accurate the software timers using it will be. It is normal to see inexpensive oscillators used in electronic devices with accuracy in the 100-200ppm range which would yield an error of about 10-20 seconds over a 24 hour period, worst case.

 

However we are seeing something that is almost an order of magnitude higher, nearly 500 seconds in a 24 hour period. Originally I thought maybe Garmin is really using a very poor oscillator, but now I'm wondering if maybe there is some sort of software bug or hardware issue because the error seems to beyond what would be caused by oscillator inaccuracy -- especially if multiple people start seeing nearly the same error in the same direction.

 

GO$Rs

 

I got it on a very good source that the time errors you are seeing here is a result of Garmin's ongoing effort to somehow correct the presistant clock problems with the Colorados. It seems as in tweaking them to keep the clock from totally quitting, they've somehow slightly over-corrected and as a result, the onboard clock now is gaining a little time as a result!

[+g-o-cashers]

I've played around with this all week and on my unit it is very consistent. My timers gain about 20 seconds per hour or around 500 seconds per day. I checked another 400t running 2.51 and it seemed to gain time at about the same rate.

 

Another call to Garmin, I was hoping that I just had a bum unit but it seems to affect all (most?) units about the same.

 

GO$Rs