I recall getting the same avg power issue early on. I thought from memory that it was displaying my max power under avg. Would that make sense (I doubt it in your case though as your number would surely be higher than 248). I am pretty sure that after downloading the latest firmware that fixed it (don't have it here - at work - so cannot confirm).

Can't help on calorie thing because I never look at that.

Another thing that bugs me is that in Garmin Connect (online software) it would graph all of your stops. So for example if you had a 15min coffee stop in the middle of a 5 hour ride your HR graph would drop linearly from say 160 down to 80 over a 15minute period, then jump back up again. This could in part be linked to the avg power mess because the auto pause might not factor out the rest periods. Why anyone would want to keep this rest time (or light stops) in memory is beyond me? Perhaps do a quick check. If you haven't already download to Garmin Connect and see what overall time is says you took. If it says (say) 2:37 and you averaged 189 watts then this might compare to your actual ride time of 2 hours at 248 watts (189*157=248*120).

Can't offer any more at this very moment.
Mike

unfortunately the calorie counter is all but useless on the garmin products.
known and constantly moaned at feature.

apparantly IIRC, garmin reported this as being a copyright/patent issue, so they can't use the good calorie counter formulae.

best place for these comments is over on http://forums.motionbased.com/...index.php?board=86.0 there are very knowledgeable edge users there.

--

i don't have a PM hooked up to my edge yet, so can't comment on the other question.

I use a 305 and have some similar issues. The calorie counter is so wrong it's not even funny. It had me at 1900cal for a 20 mile easy ride last night...LAWL!!! Anyway, the Garmin showed an avg speed for my ride of 19.6mph over 100 miles on Saturday, and in TP it was 18.3mph! When I exported the file, cut out all the zeros and imported it back into TP my avg speed was 19.5mph, so probably what is happening is that TP counts the time spent at lights, refilling your drink bottle at a fountain, stopping for a sandwich, etc. This is EXTREMELY annoying for me in TP and I wish there was a way to tell it to NOT count zeros. You could try doing the same thing and see if this corrects it. Or you could do it the hard way and make a whole bunch of laps in TP and add up all the averages...ugh. IIRC you can look in the history in the 705 and I *think* it will tell you the "autopaused" time.

Edit: BTW 248W for a century is a LOT. If you think that's a real number (i.e. your FTP is in the mid 300s and you were pushing hard) then great job!

---

---

Ooooohhhhhhh........ so "autopause" just kills the timer, but keeps recordnig everything else? Hmph. What's the point then? Would hitting 'Stop' be more accurate?

---

__________________________________________________
__________________________________________________

Breakfast is for Closers

Actually on inspecting the data, what autopause does is stop recording. However, it doesn't stop the internal timer. So if you look at the data you'll see that the timer saves data on a regular ~5 second interval (registered in minute increments of 0.084) and if you stop for 2 minutes at a light then the time data would look something like this:

Time Pwr Cad Spd
4.2 245 90 21.1
4.284 180 87 19.5
4.368 50 65 10.2
4.452 0 0 5.1
6.452 150 61 6.5
6.536 210 89 19.5
etc...

The annoyance is that in the above example (slowing down to a light, waiting 2 mins and going again) TP will directly average in the 2 minutes of "stop time" into your total average speed. This totally f's up your average speed and power. I'm not sure if there is a way around this or not but I am looking! Maybe we should start a new thread on this specific question... Unfortunately hitting stop would be far more accurate, but 10000000000x more annoying.

---

---

I don't think hitting stop will make much difference. The rides I alluded to earlier I had actually pressed stop and start at the beginning and end of formal stops (mid ride pig outs) but left it just to auto-pause for quick 'traffic light' stops. Hopefully this link works:

http://connect.garmin.com/activity/426142

If you look at the 3hr 30 marks you will see a linear descending line on the graphs for time, heart rate and cadence indicating it has still taking note of this official time gap (it obviously knows it is like 10:35am when I stopped and started at 10:50am). A similar thing could be throwing out your data.

On a side note it's interesting that it's only with the recent update that Garmin Connect is allowing the use of power data. The ride in the link above doesn't show it as it was downloaded in a previous version of Connect.

I agree you think Garmin would have looked at (like) the PT computers and thought "You know I bet people appreciate the ability to not incorporate zero values into the ride, or at least have the choice! Let's include it in our high tech devices."

Actually I think it's a problem with TrainingPeaks and not Garmin. I looked at some older Ergomo data and it had the same graph appearance...blank gaps 1:00 to 2:00 where I stopped at a stoplight. If I drag an interval over those gaps the average power and speed gets hosed too. I don't recall if it did this in version 2.1 but it definitely does in the new version...now I am going to have to go back over some old graphs when I get home and see if they get screwed up by the gaps. I have data typed in by 50 mile intervals in some long rides and if the avg speed/power no longer matches then they changed something in the latest version.

---

---

Although I don't use TP, I use SportTracks and it has a similar inaccuracy with the time/distance imported. Usually short distance and time in the automatically filled blocks and lower avg speed as well. However all I have to do is click "enter manually" to make those blocks editable and the numbers automatically swap to the accurate ones without me actually entering anything. Not sure if this helps but it works for me.

I agree with Triguy42: if your power is lower than expected in Training Peaks, it has to be averaging in some low values that the Edge is not. Stops are the obvious candidate. That's the main reason for different output on the device vs. on Garmin Connect vs. TrainingPeaks or whatever: they all have the same data, but they use different algorithms to calculate things.

You can look at the output from the device directly if you want, it's just an XML file (if you can read those.) It's a bit tedious, but if you're really really curious, you can figure out a way to get the data into Excel, and see if you can figure out how your software is calculating avg power from the data.

And on the calories, yep, fuhgeddaboudit. You're not going to get good data from Garmin on that one unless they buy the patent. They're really conservative about issues like that from what I've observed. They seem to prefer owning a company with IP than have any risk of being sued over infringement.

The funny thing is that they could just do the simple thing and use a fixed calculator based on distance traveled and the weight you enter. This wouldn't be highly accurate but would at least be in the ballpark unless you were climbing the alps. On flat roads calories per mile aren't highly related to speed unless you are going extraordinarily fast...you burn more calories per hour going fast but not more per mile. Whatever formula they are using is so far off it's pathetic.

Anyway, I will check some data at home to see if it really did handle the Ergomo differently than the Garmin 305. This thread indicates to me that they are thinking of putting in a check box for "include zeros" or "don't include zeros" in there...

http://www.trainingpeaks.com/...id=37282&posts=4

---

---

Here's another wiki that I found:

http://wiki.peaksware.com/...urer.27s_Software.3F

It seems that they are aware of the whole distance thing. My understanding of the 305 vs 705 is that the 305 uses the GPS by default for distance and speed, not the speed pod. This means that under trees and things the data could be very erratic...which I have experienced. Garmin Training Center may recalculate by interpolation, where WKO+ does not! You could try exporting from GTC to WKO+ instead of downloading directly into WKO+...I'll try this also tonight when I get back from my evening TT.

---

---

Well I figured out something very disturbing last night when looking through TrainingPeaks. I think TP has a MAJOR problem with Garmin data and the way the Garmin records it:

1) It appears that TP is not handling the variable rate recording correctly. If you take the individual data points and average them over the actual time period in each data point (which in the intelligent recording is an average of about 1 point per 7.5 seconds but can range from 3 to 12 seconds per point) it doesn't match what TP says. So if Garmin records 250W for a 3 second data interval and the next point is 200W for 12 seconds, TP reports the average to be (250+200)/2=225W. The real average is actually (250*3+200*12)/15=210W. I *think* this is what is going on...but I didn't have time to try it out on a big section of a file. When I exported it to try it here at work the exported CSV data was on a 5 second constant interval, which means the variable time basis is now gone. I should have exported as a text file...oh well.

2) It also appears that TP is averaging in the stopped datapoints as a single data point. This supports the above guess on how they are doing it. In the Ergomo data I recorded every 1 second, so on a ride that I rode a consistent 200W for 5 minutes and then stopped for 2 minutes, it would have 5*60=300 data points at 200W and 1 data point at 0W. The average according to their calculation method would be (300*200+0*1)/301=199.3W. This seems reasonable until you look at the extremely low data collection rate of the Garmin. Over that same 5 minute period the Garmin only records 40 data points!!!! So TP would calculate it to be (40*200+0*1)/41=195.1W which is a dramatic change.

If the above 2 items are actually true then TP has a major problem with the way they calculate power/avg speed/avg cadence/etc. Combine the two error sources above and you could end up with an average power that bears very little resemblence to reality.

---

---

FWIW I think the Calorie counter on the FOrerunner 405 is pretty accurate.

Are you sure you set up your information on the Edge 705 correctly?

Also, I know that PT users hjave an issue with displayed and downloaded power data becuase the display can show non-coasting inclusion. Is this an issue with the 705?

Bob