Thursday, October 23, 2014

Old La Honda: always calibrate Powertap after battery swap

Despite an early afternoon meeting for which I risked tardiness if anything went wrong, I felt a strong need to test my fitness on the Noon Ride, Wednesday edition, which climbs my favorite climb anywhere: Old La Honda Road.

After the cheap LR44 batteries I'd last installed in my Powertap gave up the ghost a few weeks into my Basel Switzerland experience, I found some superior 357's (silver oxide) in a local combo department store / food store. This should have had me up and running but I didn't have the tool to remove the cover on the hub. I eventually brought it to a local shop, to see if the guy there could remove it with an open-end adjustable wrench, but it was too tight and the metal wrench risked damaging the flats on the hub cover. So I decided I didn't need power all that much in Europe and to wait until I got home.

Indeed, the cover was on quite tight for some reason, and after applying some Tri-Flow to the interface between the cover and the hub, applying the plastic tool, carefully pressing my body weight against the tool and turning slowly, I was bit by bit able to get the cover off. Weird. I then replaced the battery and retightened the cover, albeit perhaps less tight than it had been.

Riding into work from the train I came to a traffic light which I wanted to make, so I sprinted. These sorts of sprints are substantially less than the sprints I'd do in a race or sprint workout: those are all out maximal efforts, this is a "speed up but stay aware of what's around me and in control" sprint. Yet afterwards my Garmin read 656 watts max power. That's not an uncommon number for me to see during a sprint workout. I was very pleased with this number, preferring to be very pleased to assessing if perhaps there was a measurement issue.

When it came time to leave for the Noon Ride, getting to which takes me up to approximately 30 minutes depending on my luck with the inhumanely long Peninsula traffic lights, I noticed I was cruising along at 280 watts without much trouble. Wow -- impressive power. Riding at over threshold hardly feels like work.

The ride went as usual, around the Portola Valley Loop and to the base of Old La Honda. Before we even hit the bridge, Chris Evans took off in close to a full sprint, easily double the speed of anyone else. Obviously he wouldn't sustain that, but the "start hard then manage the crash" approach is one I've seen before from riders with a strong top end. They want to make sure to empty the tank on the climb, and the best time to do that is when fresh, they feel. So blast off, then try to cut back to near threshold the rest of the way, holding on to the time boost from the first minute or so, This works on Old La Honda because the effort is short, around 16-17 minutes for good climbers. For a longer effort the penalty for the early anaerobic indulgence would be payed over a proportionally longer period.

However, not blessed with much top end I've always preferred to ride the climb more aerobically, going out at an optimistic pace, cracking off from that a bit during the ride, then in the final minutes ramping it up for the finish. Anaerobic efforts at the end of the duration don't contaminate the rest of the effort, unlike those at the beginning. But it's harder to make sure you empty the tank this way.

Consistent with my pacing strategy, I didn't want to see more than 300 watts on my Edge 500, which shows 3-second power on my lap page. So I try to keep that nice and steady close to but not more than 300 watts. Doing this will result in an average power less than 300, since when the grade transitions from steeper to shallower, there's a tendency for the power to sag a bit when spinning up the pedals. So I never average my target, as long as I treat the target as an upper bound.

I was feeling strangely good, spinning my 36/23 up the climb. Normally I'd ride a 36/21 or even 46/19. I set a PR with a 36/18. So 36/23 is low for me. But I found with a higher cadence I had no issues with my 300 watt target. Wow -- I'd really gotten more fitness than I imagined in Switzerland!

I finally started to feel the climb approaching the finish, but then it was time to ramp up the effort. I didn't look at the power meter here, rather focusing on spinning my 36/23. Then I was at the top. Looking at the lap timer approaching the finish I was dismayed to see I was over 19 minutes: 19:08.85 I later determined from the FIT file. But the display showed the laps's average power had been 289 watts. What??? That would be among my best-ever powers up Old La Honda. How was it possible I got such a good power with such a mediocre time, especially when I'm relatively light right now (56.9 kg when I weighed myself this morning).

I immediately turned around, headed back down the climb (normally I descent nearby Highway 84, but I was in a rush to make my meeting), then back around the loop the way we'd come and from there rode back to work. Along the way I rode a bit with Chris Evans, who caught up to me when I stopped to get some water. I mentioned the mystery of the high power + long time. "Time never lies," he responded, "your brakes were rubbing or it's power meter error." After we split up, he heading to his job at Stanford, me going further to Mountain View, I stopped to check for rubbing brakes. Nope. I realized it may have been related to the battery swap in the morning.

Eventually I got back to my office where I managed to get some lunch, upload my data, take a shower, get dressed, and make the meeting with a few seconds to spare, albeit with somewhat wet hair, eating some yogurts during the meeting.

Afterwards, I looked at the FIT file and was impressed. I'd gone out at close to 290-300 watts and essentially held that, average power dropping into the 280's but then rising up with my unusually strong late-climb surge to that 289 watt value. My average cadence was 83, which is good for me when climbing.

Tom Arnholt of AlphaMantis responded to my tweet by noting that after a battery change, it's important to do a manual zero of the Powertap: it may not be able to re-zero after this using just the automatic zero which occurs when coasting. This had been my mistake. Not only had my battery died, but I hadn't used the wheel in over a month. So I did the Garmin 500 "calibration" step. It said "calibration successful" and reported a huge offset. So hopefully things are better now.

Here's my data for the ride (yellow solid line) and the data scaled to give an average of 261 watts (estimated by Tim Clark's power estimator (dashed yellow).


I was disappointed with the time. But then I probably shouldn't be. I was targeting the wrong power, which I was largely able to sustain, and I finished with a lot in the tank, since I was able to surge a lot more than normal even after adjusting the power. I obviously could have gone faster. How much faster? Oh, I don't want to guess. Maybe I can try again next Wednesday.

No comments: