Looking through the logs, it appears that Zwift does not read the sensors for 3 or 4 seconds, whilst calculating the position of riders. Not even a Rx fail message. No logging of the sensors.
Further analysis of the logs after experiencing the scenario on the “10 min ramps’ workout and on the Innsbruck sprint afterwards! When Zwift is calculating the positions of riders in close proximity, it stops reading the power meter for 4 seconds. This cuts the average power calculated, dependant on the gradient and speed. I do not know if this is peculiar to Garmin Vector 3, or any setup with power meters. I had the power meter on Ant+ channel 2 with the trainer on channel 1. It makes workouts very hard when erg is set to 210 watts, pedalling at constant cadence and the power display is fluctuating between 0W and 128W. Once the one or two cyclists have eventually gone past, or fallen away, I find I am actually pushing 240-250W. It becomes tough to follow the workout accurately. Next I shall try swapping the channels, see if I can get the power meter on channel 1, I wonder if this then causes erratic trainer response instead? I remember in the past I had a lot of problems that the trainer sometimes did not change resistance on interval workouts.
Hi Rob,
you may have already done this but, have you a Garmin head unit that can read the Watts your Vector 3’s are putting out along side the power that Zwift is recording.
To be able to view the two power outputs at the same time may be of help locating the problem.
Also what trainer are you using out of interest? Is its resistance controlled by Zwift?
I don’t have a set of Garmin Vector 3’s but I have read that they have not been the
most reliable of units. If you have a access to another type of power meter then it would be good if you could double check your Vector 3’s output against a different make.
Not sure if any of that will help. Best of luck.
Cheers
Hi Troy, thank you for the suggestion with using Garmin head unit to record the sensors in parallel to Zwift. This could help to gather more evidence for Zwift, if I need it. You are right that the Vector 3 had a troubled introduction. This was largely due to a bad connection with the battery holders in use. On Zwift it showed up as the sensors disconnecting. The battery housings were replaced under warranty, the new style have a much wider electrical contact area. The firmware updates also helped, 3.60 seems quite stable. The last problem I had with the Vectors was self-inflicted. I did not notice the batteries were low and they shutdown midway rough a ride,!
I am almost certain that I have identified what is causing the current problem. When Zwift calculates the positions of drafting riders, it appears to stop reading the power meter parameters for 4 seconds. This is very clear in the logs. The question is why is this not more widespread, or is it just that anyone who reports it as a problem is told that it is their sensors at fault? Normally the Zwift log shows a reading for the power meter and trainer every second. When I experience the loss of indicated power, the logs only show the trainer. If this was due to a sensor error, it would show Rx Fault, but the log has nothing. After 4 seconds, it calculates the average power (over 3 seconds) and uses that result. I have raised this with Zwift support, the 2nd line looked at the logs and pushed it to 3rd line. I hope that they can find what combination of settings and circumstances causes this and develop a fix. In the meantime I still do not understand why no one else appears to have reported this.
Cheers
It see like you are the only one with this problem, and that make it even harder to simulate.
Hi Rob,
I have not experienced the problems you are describing of total power drops.
I have used two different setups one with a power meter the 2nd with smart trainer.
The only thing I have noticed is that during a workout my Garmin power reading is much more stable than the Zwift reading. I think I have 3 second smoothing on both.
This may also be true during normal riding/racing but I have never checked to see.
Your second paragraph is out of my depth I’m afraid. I have never looked at my Zwift logs etc. This is an interesting finding though. Maybe some of the other Zwifters will be
able to give their findings on this point.
Sorry I can’t be of more help.
Hi Troy, This is the detail from the logs. In the log below, the average power drops to 0 at 10:56:14. There is a power reading at 10:56:04, 05, 06, 07, 08, 09 and 10. This is in a workout with the trainer (KICKR) in ERG mode, I believe at 210W at this stage. I was riding more or less constant cadence. At 10:56:11 it starts calculating rider positions (under 4m). At 10:56:11, 12, 13 and 14 it is communicating with the trainer. The Rx Fail at 10:56:14 (Channel 1) is the trainer. At 10:56:14 the power meter is read again, clearly the average power for 3 seconds without a value is 0.00. The power meter is read every second again from 10:56:15 and the average power rises to the expected value. The worst instance was at the last segment of the sprint, when I was aiming for PR until I was passed by a rider with my average power showing 0W.
10:56:04] ANT : [PowerMeter] CrankPower: EC: 149 WT: 149 Period: 30111 AccumTorque: 45281
[10:56:04] ANT : [PowerMeter] Vang = 9.90 radians/sec Tavg: 19.75 nm Pavg: 195.49 watts
[10:56:04] FE-C 17 received
[10:56:05] FE-C 26 received
[10:56:05] ANT : [PowerMeter] CrankPower: EC: 150 WT: 150 Period: 31418 AccumTorque: 45941
[10:56:05] ANT : [PowerMeter] Vang = 9.85 radians/sec Tavg: 20.62 nm Pavg: 203.06 watts
[10:56:05] FE-C 224 received
[10:56:05] ANT : Transfer Completed.
[10:56:06] ANT : [PowerMeter] CrankPower: EC: 151 WT: 151 Period: 32719 AccumTorque: 46607
[10:56:06] ANT : [PowerMeter] Vang = 9.89 radians/sec Tavg: 20.81 nm Pavg: 205.85 watts
[10:56:06] FE-C 26 received
[10:56:06] FE-C 224 received
[10:56:06] ANT : Transfer Completed.
[10:56:06] ANT : [PowerMeter] CrankPower: EC: 152 WT: 152 Period: 34020 AccumTorque: 47277
[10:56:06] ANT : [PowerMeter] Vang = 9.89 radians/sec Tavg: 20.94 nm Pavg: 207.08 watts
[10:56:07] FE-C 26 received
[10:56:07] ANT : [PowerMeter] CrankPower: EC: 153 WT: 153 Period: 35318 AccumTorque: 47943
[10:56:07] ANT : [PowerMeter] Vang = 9.91 radians/sec Tavg: 20.81 nm Pavg: 206.32 watts
[10:56:07] FE-C 17 received
[10:56:07] ANT : Transfer Completed.
[10:56:07] NETWORK:Resetting bike pos delta move > 4000 cm (35170484224) - rider name=xxx
[10:56:07] NETWORK:Resetting bike pos delta move > 4000 cm (1198250368) - rider name=xxx
[10:56:07] FE-C 224 received
[10:56:08] ANT : [PowerMeter] CrankPower: EC: 154 WT: 154 Period: 36618 AccumTorque: 48599
[10:56:08] ANT : [PowerMeter] Vang = 9.90 radians/sec Tavg: 20.50 nm Pavg: 202.91 watts
[10:56:08] FPS 27, 259280, 46973, 462456
[10:56:08] FE-C 26 received
[10:56:08] ANT : [PowerMeter] CrankPower: EC: 155 WT: 155 Period: 37916 AccumTorque: 49273
[10:56:08] ANT : [PowerMeter] Vang = 9.91 radians/sec Tavg: 21.06 nm Pavg: 208.80 watts
[10:56:08] FE-C 224 received
[10:56:08] ANT : Transfer Completed.
[10:56:09] Calibration data [1] [187] [0] [7] [118] [86] [170] [153]
[10:56:09] NEW Calibration offset = 43673 stdDeviation = 0.00
[10:56:09] FE-C 26 received
[10:56:09] ANT : Rx Fail on channel 3
[10:56:09] Unknown power data [249] [255] [151] [2] [88] [154] [2] [1]
[10:56:09] FE-C 224 received
[10:56:09] ANT : Transfer Completed.
[10:56:09] ANT : [PowerMeter] CrankPower: EC: 157 WT: 157 Period: 40529 AccumTorque: 50579
[10:56:09] ANT : [PowerMeter] Vang = 9.85 radians/sec Tavg: 20.41 nm Pavg: 200.98 watts
[10:56:09] Starting critical power curve search job (37927778) mTimeStamp = 2613.58594 m_lastSearchTime = 2613.58594 calculating=false
[10:56:09] FE-C 17 received
[10:56:10] ANT : [PowerMeter] CrankPower: EC: 158 WT: 158 Period: 41828 AccumTorque: 51249
[10:56:10] ANT : [PowerMeter] Vang = 9.91 radians/sec Tavg: 20.94 nm Pavg: 207.40 watts
[10:56:10] FE-C 26 received
[10:56:10] ANT : Transfer Completed.
[10:56:10] FE-C 224 received
[10:56:11] ANT : dID 783406 MFG 1 Model 2787
[10:56:11] NETWORK:Resetting bike pos delta move > 4000 cm (764992896) - rider name=xxx
[10:56:11] FE-C 26 received
[10:56:11] ANT : Transfer Completed.
[10:56:11] FE-C 224 received
[10:56:11] FPS 27, 258823, 47047, 463984
[10:56:12] FE-C 17 received
[10:56:12] FE-C 26 received
[10:56:12] ANT : Transfer Completed.
[10:56:13] ANT : dID 783406 SW Revision 9216 SerialNum -330370002
[10:56:13] FE-C 224 received
[10:56:13] FE-C 26 received
[10:56:13] ANT : Transfer Completed.
[10:56:14] ANT : Rx Fail on channel 1
[10:56:14] ANT : [PowerMeter] CrankPower: EC: 164 WT: 164 Period: 49693 AccumTorque: 51249
[10:56:14] ANT : [PowerMeter] Vang = 9.82 radians/sec Tavg: 0.00 nm Pavg: 0.00 watts
[10:56:14] NETWORK:Resetting bike pos delta move > 4000 cm (191514976) - rider name=xxx
[10:56:14] FE-C 26 received
[10:56:14] FE-C 17 received
[10:56:14] ANT : Transfer Completed.
[10:56:15] ANT : dID 783406 Battery Level 4
[10:56:15] FE-C 224 received
[10:56:15] ANT : [PowerMeter] CrankPower: EC: 166 WT: 166 Period: 52313 AccumTorque: 51935
[10:56:15] ANT : [PowerMeter] Vang = 9.82 radians/sec Tavg: 10.72 nm Pavg: 105.29 watts
[10:56:15] FPS 27, 258449, 47112, 465394
[10:56:15] FE-C 26 received
[10:56:15] ANT : Transfer Completed.
[10:56:16] FE-C 224 received
[10:56:16] ANT : [PowerMeter] CrankPower: EC: 167 WT: 167 Period: 53633 AccumTorque: 52655
[10:56:16] ANT : [PowerMeter] Vang = 9.75 radians/sec Tavg: 22.50 nm Pavg: 219.33 watts
[10:56:16] Calibration data [1] [18] [3] [255] [255] [255] [255] [255]
[10:56:16] FE-C 26 received
[10:56:16] ANT : Transfer Completed.
[10:56:17] Calibration data [1] [187] [0] [7] [242] [89] [170] [153]
[10:56:17] FE-C 224 received
[10:56:17] ANT : [PowerMeter] CrankPower: EC: 168 WT: 168 Period: 54948 AccumTorque: 53393
[10:56:17] ANT : [PowerMeter] Vang = 9.79 radians/sec Tavg: 23.06 nm Pavg: 225.67 watts
[10:56:17] FE-C 17 received
[10:56:17] Calibration data [1] [187] [0] [7] [158] [88] [170] [153]
[10:56:17] ANT : Transfer Completed.
[10:56:18] Calibration data [1] [187] [0] [7] [222] [87] [170] [153]
[10:56:18] FE-C 26 received
[10:56:18] ANT : [PowerMeter] CrankPower: EC: 170 WT: 170 Period: 57591 AccumTorque: 54993
[10:56:18] ANT : [PowerMeter] Vang = 9.74 radians/sec Tavg: 25.00 nm Pavg: 243.43 watts
[10:56:18] FE-C 224 received
Does this only happen in ERG mode?
What trainer do you have?
What are your PC specifications?
It has happened on every ride since June 1st, workouts, free rides, races including ToW stages, L"Etape, Vatternrundan. I thought it only occurred during races, but that is not the case. It is possible that it might have happened before then, but classified as general Zwift dropouts.
Trainer is KICKR 2017 v3.4.68 with Climb. Running macOS 10.14.6 on Intel Core i7 8th gen (6 core) with 16GB ram (Mac Mini8,1), with Intel UHD Graphics 630, 137GB free disk. It is direct ethernet connection, Zwiftalizer shows no network issues. Ant+ is Tacx. Judging from the logs, this is a multithreading issue in the app. I have tried BTLE as well as ANT+ but makes no difference.
And do you pair your trainer as FE-C? Very strange. 
I would start by simplifying things.
-
Kickr to Zwift w/out Vectors, w/out the Climb. Kickr paired via BLE. Kickr is both Power and Controllable Trainer.
-
Same set up as above but using ANT+ FE-C. I assume when you say “ANT+ is Tacx” you mean the Tacx USB antennae. I have no experience w/their dongle or antennae, but you might even consider testing w/another USB dongle (e.g., Garmin) for comparison.
-
Again for each of the above but w/the Vectors as power and Kickr as Controllable Trainer.
-
Add the climb back in for each of the above scenarios if the problem didn’t already present itself.
Lastly, perhaps a typo, but you are “Running macOS 10.14.6 …” Isn’t that a beta?
1 Like