Consistent power drops while cresting rolling hills

I was riding the Sand and Sequoias route this evening and on the rolling hills I kept having my power drop to zero every time I would pedal through the resistance of the next incline at speed and in a high gear.

Most of the troubleshooting advice seems to suggest BLE issues, but I would suspect that would be much more random. This was very consistent. Every single time I was cruising downhill on Titan’s Grove and keep pedaling through resistance of the next hill my power would rise to 250-500 watts then suddenly drop to zero watts for a couple seconds. It seems unlikely to be a tire slippage issue either as the flywheel would keep supplying at least some watts against the resistance (also, I didn’t feel or hear any slippage, subjectively it felt like I was applying similar levels of power to the pedal throughout the events).

Is it possible that this is just Zwift trimming what it sees as outlier power readings? My FTP is pretty low (mostly because I just started this week and am out of shape but) partly because I’ve been sharing the bike with my wife and kept forgetting to readjust the seat before riding. My ride this evening was one of the first where I both had the seat adjusted correctly/could really apply full power to the pedals and was also hitting some inclines/resistance on the hills.

Activity is here. I can provide the log file if it would help, but I don’t see any messages that coincide with the power drops.

I’m using a 26" MTB with city tires on a Kurt Kinetic Rock and Roll Control. I’m trying to not get too crazy with my expectations from a wheel-on trainer before the trainer tire I ordered arrives, but this doesn’t seem to be a tire issue from what I can tell.

Thanks in advance for any help.

I’d see if it was repeatable and also if changing computing device changes anything (ie try on your phone or tablet if on a computer or vise versa).

Did you run the log through the zwiftanalyser or do you know how to look for drop outs?

I did run it through Zwiftanalyser and didn’t see anything unusual, just the normal connection issues at the start of the ride and a couple places where FPS dipped by 1 or 2 FPS–and no events in the middle of the ride when the issues were occurring.

That said it’s probably accurate to say that I don’t know how to look for drop outs, I’m just following my intuition and looking for anything suspicious in the logs correlated with when the issues were occurring.

If this is a smart trainer then see if it happens at both 0% and 100% trainer difficulty.

Okay, thanks for the advice. I’ll try seeing if I can replicate the issue tomorrow and with varied trainer difficulties. I believe it is currently set to 100%.

So both myself and my wife were able to replicate this today, using different phones to run the Zwift companion app (she has an iphone, I have a pixel).

After replicating it (at 3:36 in this activity), I turned the trainer difficulty down and the issue seemingly disappeared–although it’s a bit of apples and oranges comparison as the issue only appeared with 300+ watts on 5%+ inclines and with MTB gearing and 0% trainer difficult I was well below 300 watts for the most part. I did really spin up my cadence as much as I could at one point (around 7:18) and got it up above 400 watts for a second or two without any power drop. Then as soon as I turn trainer difficulty back on, I get a power drop pretty much right away (8:56 - 390w, 8:58 - 0w, 9:01 - 217w). I was then able to once again replicate the issues later on the rolling sections at the end of the course.

Based on this, it seems like it is the combination of resistance from inclines and higher power than is causing the power drops. I wonder if it might be something that only happens when the stepper motor controlling the magnet is trying to make quick changes (as would happen when taking rolling hills at speed or when switching from 0-100% trainer difficulty).

Any ideas, or is this something I should contact Zwift support or Kinetic support about?

I don’t think it is a zwift issue.

It seems to always happen when there is a sudden surge in power read. It looks like the trainer over reads the power and then to compensate to drops it to zero for 2 seconds. You could be correct about the stepper motor, it might be just a limitation of the design. I am not sure.

Can it be that your wheel is slipping when the trainer adjust to higher power.

Make sure the roller is tight against the wheel.

Thanks both for the suggestions.

@Gerrie_Delport While I have experienced some other problems that I suspect may be tire slipping (waiting for the new tire I order to arrived before troubleshooting those issues), I don’t think that’s the case here. When the tire slips do happen, the 3 second avg. would drop to ~20-60 watts. Even if I completely stop pedaling, it takes a good amount of time for the power to drop to zero as the inertia from the flywheel is converted to power. So instantaneous drops to 0 watts seem much more likely to be a firmware/software issue, somewhere between the strain gauge in the trainer and Zwift.

@Ben_Brawn That type of corrective behavior would make sense, but would be difficult to confirm without another power meter. Also, I don’t think those spikes were likely to have been so erroneously high to require halving the avg. power output by averaging in zeros, as those spikes usually represent me digging in or standing as the resistance kicks in. Subjectively it does feel like I am at least doubling my previous power output when the resistance kicks in, whereas with the zeros averaged in, it would imply little to no change or even a decrease.

I’m guessing it probably makes the most sense for me to reach out to Kinetic support. @Ben_Brawn what software are you using to zoom in on the spikes? I’m a fairly regular user of statistical/graphing software in my day job, so I was thinking of looking into some of the R packages that can import .fit data so I can more clearly illustrate the issues I’m having.

I looked you up on strava and clicked view analysis, then highlighted the section of the graph I wanted to look at (because zwift doesn’t let you select a part to look at). If you are using strava for anything other than zwift you may want to increase privacy settings.

I am having the same issue, using wahoo cadence and speed sensor and everytime the watts get close to 400 it drops to 0 for a few seconds. It is quite anoying when doing trainings. Did you find a fix for this? Dont think is your trainer since I am not using the same.

Which trainer do you have?

I have a traveltrac, but then I put the sensors on a stationary bike and have same issue. Last thing I am going to try is to use them with the wahoo app and see if it is a sensor issue or if it is the zwift app.