Friday, January 22, 2016

Power struggle

A few months back I purchased a set of Garmin Vector pedals to add power data to my bike workouts. I have no regrets about this decision, but getting the system calibrated and working smoothly has not been easy. In fact, I would be kind in calling it a struggle.

Adding power measurement to a bike for the first time will be confusing for any cyclist. The watts display jumps all over the place, not at all like the familiar speed, cadence, and heart rate readings. Then comes the question, "What does it mean?" Dig a little deeper and you run into a whole set of new terms; Normalized Power (NP), Functional Threshold Power (FTP), and Variability Index (VI), just to name a few. But I am not talking about that stuff. What I'm talking about is getting the Vector system to do its job, smoothly and consistently. By the way, if you need help getting started with power I recommend Joe Friel's book, "The Power Meter Handbook."

My very first challenge was getting the pedals to clear the shoes. I have written about that previously and will not say any more about it other than the right cleats on the right shoes is critical for riders with luau feet like me. In my case, genuine Keo cleats work best.

Right up there with the shoe issue has been left-right balance. In the more extreme version of this problem one of the pedals will stop producing a power signal altogether. What should have been a reading of 100 watts is only 50. When such a workout gets uploaded to Training Peaks it appears I was out riding with a bunch of senior citizens on tricycles. Only slightly less troubling, but more common, has been a left-right offset of around 60%-40%. I am still not sure what causes this. At first I thought the most likely culprit is variation in tightening, but now I am not so sure.

I use a torque wrench. I wanted a nice, click type, and I was in a rush. The only place in Honolulu that had anything in stock was Sears. I bought their click type only to discover that it only works on right-hand threads. The left pedal attaches to the crank arm with a left-hand thread, so I had to exchange that wrench for a simple, bend-the-bar type. I actually prefer that the socket attachment does not ratchet, because these pedals require a crow's foot wrench and it must be at 90 deg. to the wrench to get a proper result. The problem with this wrench is that it is difficult to view the scale. The bike is on the ground, one hand holding the pedal, the other the wrench handle. I am working hard to make 25 ft/lbs while down on one knee where I can see the scale, all the while struggling to keep the bike from falling over. But there is more. The left pedal pod has developed a burr, just enough to catch on the pedal shaft and interfere with tightening. What happens is I pull the correct torque but I was pulling against the burr. A lot like going over a pot hole. The fastener did not really tighten, and when I let go the result is as if I only pulled 15 ft/lbs.

Even when the balance is right, the power zone readings would be sometimes off. Way off. One day I will be almost tracking the same zones for power and heart rate, and the next day, under the same conditions, my zones will be off by one or two. Going up Tantalus my HR will be 4Z and my Power up in 5Z or 6Z one day, then down around 2Z another day. The first time that happened I discovered that somehow my Garmin 800 power zone settings had changed. My guess is that I did not know I had to set them in Garmin Connect, and during a data sync the web values got pushed down to the head unit. I set my values in Garmin Connect and re-did the 800, and that problem went away.

I did my first time trial test on my new Cervelo P3 a couple weeks ago and was surprised at how much higher my FTP was. On the other hand, the value I had up until then, from the Merlin, seemed low. My first test on the Merlin gave an FTP of 124. Not long after that Training Peaks alerted me that my threshold power had gone up to 131. On my new bike I got 265. That is a huge improvement, enough to make me suspect something was amiss.

I always do the Tantalus ride on my Merlin road bike, and now I do everything else on my P3 to get used to the new position. On Tantalus last week my zones were once again way off. HR zone reading higher than RPE, power down in 1Z. Back home I did some recalculation and decided that an FTP of 170 was more realistic than that crazy 265. For HR, my LTHR is usually around 150 but that last test produced 140. Based on that I was climbing Tantalus in my sprinting zone. I put it back on 150.

These values seemed more realistic on my next Sunday ride, but I was paying more attention to form and watching the road. Then, coming back from Hawaii Kai, my power reading kept cutting in and out. No low battery warning, which I would expect -- perhaps the Edge 800 is too old to catch those? As I was changing the batteries in the peddle pods I decided to run through the bike settings. Along the way I learned something. Crank length, and some other settings, only appear when the head unit is in communication with the pedals. The pedals fall asleep when they do not move for about five minutes. I was stumped by the missing menu items until I Googled it. When I got my menus working I discovered that my crank arm length was set to 144. What kind of crazy value is that? My FSA cranks are 170. That might explain why X foot-pounds of force at the pedals produced 130 watts one week and 265 a few weeks later. Watts is like ft/lbs in that the moment (lever arm) length makes a difference. I set it back to 170. Then I went to Bike Settings and change to the Merlin. It has a crank length settings of 142.5. Where did that come from? (Off topic, the odometer reading is over 7,000 miles! Hard to believe, but I'm sure that number is right!)

One week later and I am still struggling. On Sunday's long ride the balance was off. On Monday (Martin Luther King Day) I did some easy riding with the goal of troubleshooting the left-right balance issue. The power reading was jumping all over the place so I went back home and re-torqued the pedals, smoothing out that burr while I was in there. I ended up worse off than before. Power balance would start off as before then fall away to 100%/0%, then no output. Rest a minute and it would do the same thing, over and over. This was on new batteries. I came home, Googled around and found there was a Vector software upgrade -- I was on 3.5, latest was 3.9. Tried to update but the updater program could not find the right pedal, or neither pedal. I let it all rest while I ate dinner. After dinner the updater connected just fine and the upgrade was successful. Go figure.

Sunday ride, no drop-outs but balance definitely off. (Click to enlarge.)

Monday, two rides, drop-outs on second ride after re-torquing pedals.

On Tuesday's Tantalus ride there was no sign of the drop-out issue but balance was still a bit off. More troubling to me that the power zone display on the Edge 800 was much too high - I was climbing steady state in zone 6 and 7 (if only that were true!). That night I fiddled with my FTP spreadsheet so that the zones moved down about one level, particularly so that what had been zone 6 would display as zone 5, which is what it felt like. The result was an FTP of 200. Okay, maybe that 265 was not a fluke. I edited the power zone settings on Training Peaks to 200, recalculated the Tantalus results and sure enough, most of the climb was in zone 5 where it belonged.  Last step was to change the FTP setting in the Edge 800.

Tuesday Tantalus ride, no drop-outs but balance still off.

Wednesday spin class was the next opportunity to test the new zones. This entailed moving the pedals from the Merlin to the P3.I moved the pedals Tuesday night, and when I set up the bike on the trainer Wednesday afternoon I had the same thing as my second Monday ride -- a minute or so of out-of-balance readings leading to a total loss of signal from the right pedal. I got off the bike, pulled both pedal pod connectors (this essentially removes batter power from the pedal), shut down the head unit, waited a couple of minutes, then plugged in. Set the crank angles (spin 70+ RPM) and once the power numbers came on everything was hunky-dory.

Wednesday spin class, including HR and cadence. Points high-lighted in yellow are slow pedal drills so expect balance to be off. Area circled in red was arms in the air balance drill, again balance wobbles are expected.

My theory now is that the initial configuration needs to be done soon after plugging in the pods. When I install the pedals but do not do the initial spin-up for some long period of time (8+ hours?) the system gets confused.

My next CP30 test to set power zones is Sunday morning. What I expect is perfect results from the get-go, because I will not have unplugged the power pods since Wednesday's spin class. If the balance is off I will go through the pod disconnect-reconnect sequence and see what happens.

In closing, I did my run LTHR test this morning and got 151. At least that number has been consistent.