Edge 705 Users - Quick Question

[MattTuck]

I am considering purchasing a garmin edge 705, which seems to be a pretty full functioned uber cycle computer.

I have a question on how it works.

It seems the speed is generated by the GPS tracking where you go. Is this true? Is it accurate? It would seem to me that in trees and clouds, it could be a problem. My GPS in car unit seems to have a hard time sometimes when it first starts up, just curious if this is a problem.

My second question is based on this, some of the versions include a "cadence/speed" sensor. Is this is a sensor that can be either a speed sensor OR a cadence sensor? or is it a sensor that can measure both speed AND cadence.

Thanks for the help, the product descriptions didn't seem to specify.

[dave thompson]

The 705 GPS is quite accurate, within 30~40 feet accuracy IIRC. If the 705 cannot 'see' the satellite for an extended period of time, it won't register that portion of the ride. Mostly that type of problem occurs when you're in a 'canyon' situation (downtown with tall buildings, or in a deep, narrow canyon).

The cadence and speed functions are built into the 705 and to be used have to have a sensor which attaches to the chainstay and utilizes magnets on the crankarm and rear wheel spokes to get these readings.

[RBaker2778]

The speed/cadence sensor is also the only way you can use the 705 for trainer workouts. GPS will calculate speed outdoors, but the sensor can pick it up when paired. The first time you set it up, it compares rotation of the wheel with the GPS tracking, and calibrates automatically. No need to program wheel size.

[1centaur]

From my experience with the 305, when outdoors the speed sensor is designed to deliver a reading to the head unit when the GPS fails to (i.e., under heavy tree cover, etc.). In order for it not to be like every other speed sensor on bikes, it is not designed to be the primary provider of data all the time - it is a back-up. That's the problem. It takes too long to "admit" that the GPS is not delivering the speed data with sufficient consistency to make the switch to the speed sensor. So, if you are riding under patchy tree cover, you will lose speed accuracy repeatedly as it makes the switch. Further, I found the battery in the sensor to wear out quickly, and possibly even draw power if when stored the bike is put away with the sensor in range of the pick-up.

I chose thereafter not to have the sensor on my other bikes but to live with the GPS moment-to-moment speed accuracy being sometimes incorrect for short periods (i.e., I might never know how fast I plunge down a mountain under tree cover). The average speed calc remains plenty accurate (equal to traditional computers) for my environs because the unit calculates what it missed from a combination of one's position before and after the signal got spotty and the time passed. I'm not in a tight canyon for minutes, I'm under tree cover or otherwise getting poor reception for seconds at a time. Observing the readout, I see momentary speed gradually slow even as real speed maintains because the algo "assumes" I am slowing down when it loses signal strength. My mental rule of thumb is that my real speed is never slower than my readout, though I imagine there are microbursts of great speed as I catch up to real time that don't make it into the max speed calc at the end.