Possibly, but temp affects air density the wrong way: a 9 degree F. drop in temperature should *increase* air density a little. So I'd be looking at tires or your power meter.

 
You could do a couple of things. I would try to get hourly weather data and see if things are changing over the 2+ hours. Look at barometric pressure which will greatly impact air density.

Also look at wind. You may have more of tail wind contribution in the 2nd part of the ride. I am not sure what your ride look looks like. Is it out and back or some kind of loop ?

Maybe look if there has been a big change in zero offset of the Quarq at the beginning and end of ride. You used to be able to do this with a Quarq, not sure you still can. Can you zero during the ride ?

Is this pattern reproducible ? Knowing that would help narrow things down.

That is a low CRR you are using.

Im using a 3 mile loop with decent tree coverage. I shorten them to half laps when I do proper testing.

I'll try to do ABA with my pedal power meter added to compare with my Quarq and see if I can tease anything out. I wish it were reproducible but some days I get really clean data and other days the last position is ~10 watts faster.

The CRR was just a random number. I usually use around .005 for my training setup.

I am not certain if this really matters but I notice you have constant elevation velodrome checked as on. What happens if you check the auto correct and the constant elevation check boxes to off?

Robert, do you use Golden Cheetah for you analysis? I can't get it to run on my MAC and I have the operating system updated, but just won't load. Any suggestions other than Golden Cheetah?

What version of the OS (Catalina, Big Sur...) and what version of GC ?
When you say "won't load", what happens ? Do you get something like the GC icon appearing as if it was running but nothing on the screen ?

I am running Big Sur 11.4. I am using the recommended GC that it detects the 64 bit one. I download it and open the dmg file to open it. I click on the icon and click on open. The GC icon just bounces on the left hand side and then after about 5 minutes it says it is not responding.

 

I build GC myself, but did have some problems on Big Sur which I never fully investigated.

Maybe try the 3.6 development build ?
https://github.com/...ses/tag/v3.6-DEV2101

lonniecdams wrote:

I am running Big Sur 11.4. I am using the recommended GC that it detects the 64 bit one. I download it and open the dmg file to open it. I click on the icon and click on open. The GC icon just bounces on the left hand side and then after about 5 minutes it says it is not responding.

I tried the 3.5 and 3.6 builds on Big Sur and they did not work. Same thing as you, bouncing of icon.

I went to the google group (golden-cheetah-users) where all the smart guys hang out.

Ale, one of the main developers put a link to this build
http://goldencheetah-binaries.s3-website-us-east-1.amazonaws.com/

I tried it on Big sur and it worked. It did the standard apple bitching about malware.

There is talks of versions of Qt (one of the main components they use to build). The conversation matches my experience. The new QT required a new XCode. The new Xcode works on old machines, but the old Xcode doesn't work on new machines........

Thank you so much. What was weird is last night it bounced and then about 15 minutes later it just opened. It seems to sit there for a while saying, "Application is not responding". Then it opens right up, but I will try those other versions. You guys are amazing thanks for all of your help.

No problem, glad it worked

And to answer your original question, maybe it's an urban legend, but Robert does not use GC/Aerolab. Us mere mortals can do mental addition and subtraction. Robert does mental VE.

Very strange. I went back and retried 3.5 versions that did the same "bouncing" as you. They opened up immediately this morning. As if there is some type of "initialization" that had to be done, and now that's it's done, everything is fine.

Other than the train option not working Golden Cheetah seems to be working just fine. Question about the VE testing. Can I just use any home weather station that measures barometric pressure? I am thinking the home ones that you see that measure inside and outside temperature and then humidity and pressure. Those that do a lot of VE testing thoughts?

I'm hoping to get a little help troubleshooting my protocol. I am currently setup using a wahoo elemnt bolt as my ride recorder, garmin vector 3s as my power meter, and a garmin speed sensor 2 as my speed sensor.

My course is an out and back with an initial climb and then a short downhill and immediate uphill, both of which I coast to lose enough speed to turn around on the uphill without using any brakes. I've got my start/finish position chalked on the road as well as the start, apex, and finish of my turnaround point. I use the lap button on each start and finish and use that to cut out the turnaround back (in red) to the start as I have to break hard to turnaround on the start/finish portion of my test area.

I've been out testing a single position on 3 different days so far and have yet to come up with consistent results from any of them. Each day of testing has had very little wind. I recognize that I'm on a single sided powermeter which isn't the best, but is I would think that having a muscle imbalance would skew the cda results and that I wouldn't have an inconsistent enough imbalance to make the results as unreliable as they are.

Any ideas on what could be making my lap not symmetrical comparing pre-turnaround to after turnaround? My first guess would be maybe some wind but each lap being non-symmetrical in almost the same way makes me doubt it. Could I have missed a recording setting that smells like smart recording? I've got my speed sensor set to use the specific wheel circumference but haven't found any settings for the powermeter that would suggest fishy recording. Or does it all come down to using a single sided powermeter?

Any help would be greatly appreciated. Happy New Year.

Happy new Year too.

In you precise description, you didn't mention if you ride on the middle of the road, which is a necessary condition (but potentialy unsafe one) to have symmetrical elevation profile.

I also have non-symmetrical elevation profile here (red curve: use of a barometer with a +/-10 cm precision) during a low wind day (virtual elevations corrected by airspeed from an anenometer). It may be explained cause a road is not perfectly level from a side to other (or some other reason that I am also interested to know too).

---

Blog | Twitter| Bike CdaCrr app

Just to clarify, it looks like the 2nd image was a Golden Cheetah screenshot? What was used for the first image?

The first one was mycda.app. Golden Cheetah doesn't seem to import my laps in the correct spot and mycda.app does while also automatically isolating the laps. Using mycda just makes it so I don't have to manipulate my data to remove the between laps section.

Oh, I didn't realize that I was supposed to be going center but that makes sense for the asymmetry, thank you. As far as the lack of precision in results, I am beginning to suspect either my speed sensor accuracy or how the elemnt bolt uses the speed sensor. I have the elemnt set to use the wheel circumference instead of the "auto" setting which still uses some gps. However, my lap distances still vary by up to 26 feet in a set. What sensor do you use and/or is there a gold standard that people tend to use for aero testing? I've seen a lot of mentions for the gsc 10 but was having trouble getting my hands on one so I had settled for the garmin speed sensor 2.

I personally could never get remotely decent data with a single sided power meter, but that may just have been me!

What speed sensor are you using?

• Do not use magnet-less speed sensor.
• PowerTap G3 hub may have some (rare) oddities (velocity spikes).
• Gsc 10 is an old speed sensor.
• I am using this cheap one (magnet sensor) with good success: https://www.amazon.com/...aps%2C163&sr=8-3

---

Blog | Twitter| Bike CdaCrr app

Unfortunately I have been using a magnet less sensor. Garmin Speed Sensor 2.
https://www.garmin.com/en-US/p/641230

I’ve now got a magnet one in the mail. Arrives Tuesday. Thanks for the help. Hopefully the next round of testing has me closer to quality results.

All happy aero tests are alike, all unhappy aero tests are unhappy in their own way. There are so many ways that failure to fit can occur that you need to dive into the actual data to see what happened. It's hard to diagnose just from a VE profile. That is, VE is a very sensitive diagnostic but because so many things go into it, it's hard just from the profile alone to figure out which of those many things went wrong.

Usually when there's a lack of fit, I have to go through a checklist of things that may have gone wrong. Unfortunately, right now, diagnosing lack of fit is sort of like drawing an owl.



At some point those of us who know what to do have to fill in a bit more about step 2.

Robert, tow quick questions for you.

1. Can I use a hub mounted speed sensor. Some have said yes, some have said no.
2. I have a Garmin Edge 130, I did a route with the GPS and I just got a straight line for elevation. There was about 6ft of gain/loss in the 250ish Meter loop I did. Is this enough elevation or do I need more? Will that computer work?

Thank you.

I'm no Dr. Chung, so take what I say with a grain of salt, but I would try multiple loops of that track. You should see the elevation rise and fall periodically. The 130 has a barometer for barometric elevation so should work OK. As for the hub sensor I have used one and it worked for me but make sure you set the correct wheel circumference (use a weighted roll out meaning sit on the bike full weight and roll out 1 wheel revolution measure the distance).