Hi,
I use a Justo linked by bt to my pc. I use the Elite bt bridge to link trainer, power, cadence, and hr so a single ‘signal’ is used. I have successfully connected this setup and zwift play controller to an atv which can only take 2 bt connections - 1 for Justo and all it’s data, 1 for zwift play controllers. This is also how I connect to my pc. I don’t know if the Elite bridge would be helpful here - I have lost track of all the connections.
I believe that could bridge the cadence sensor, but I have no experience with the Core sensor so I’m not sure if that would work. I think Robb might have mentioned earlier that it did not but it’s not something I know about.
The way you refer to the Rizer suggests you are not familiar with the device. Since you are helping people this might be useful to you. I believe the same holds for all of the incline devices.
There are 2 separate parts to the Rizer. Steering connects to Zwift via Bluetooth the same as the Steerzo or other steering devices.
The incline part I believe is invisible to Zwift. The Rizer reads an Ant+FEC data stream from any trainer that supports that. For trainers without Ant+ connection is to the trainer via BT. I guess there is a bit more radio noise but the Ant+ is transmitted anyway.
Elite trainers transmit an actual incline. If that is not sent from a different trainer brand then incline is calculated from power, speed, and rider weight which doesn’t work well in my experience. I had a Wahoo v5 and pedaling harder had the effect of the Rizer moving up and coasting on flat ground could make it pitch down. QZ app can be used between Zwift and the Rizer to make a Rizer function properly on other trainer brands.
‐-----------------------------
This last ride (the one last night in this thread) was mostly reliable despite the resistance changes which I mentioned as data. I did things to increase the likelihood of problems to see what would happen. I moved cadence to native BT. I stopped for breaks which has caused loss of VS in past rides.
The ride before that (The one before last night also above) with my regular setup I considered a success. I didn’t know there were any disconnects until the TrainerDX scan. Possibly milder terrain made resistance changes less obvious if they occurred.
What I will do for some future rides:
TrainerDX after each ride.
Connect as normal. Chose a route with climbs like last night that will exaggerate resistance changes. Stop on steep parts. Stop on flat parts. If changes are noted try disconnecting extra devices midride.
Same as last example but start with only power, resistance, Play, and HR. Instead of removing cadence, Click, and steering after a resistance change start without them.
If I still noticed changes in resistance connect only Power, resistance and HR. Use mechanical.
I will post results.
TP-Link will be on my shopping list unless all disconnects stop.
The Rizer is another Bluetooth paired device. I don’t imagine the Rizer connection involves a lot of data but the suggestion is simply to strip it down until it works well (hopefully). Whatever you choose to remove is going in the right direction in terms of testing. Do it any way that works for you. Keep the things you care about the most.
Tonight I connected the way I usually do.
Power = lan
Bt Cadence sensor on crank = bridged on Justo 2 to lan
Resistance = native bt as required for virtual shifting
Steering = native bt
Play = native bt
Click = native bt
CORE (HRM pairs with CORE) = native bt (CORE temperature is not supported by the bridge)
Connection was a bit slow probably because everything was set to BT from last night and needed to reset. Once connected I had no disconnects during the ride. I stopped twice. It felt like the resistance changed a bit when resuming but went back to normal. A short successful ride.
My only frustration tonight was the Elite fan froze at maximum speed. Only the red light was on and the button was unresponsive. I left it for half an hour to see if it would stop but finally had to unplug it. Not the first time.
Another ride another log. I should probably ride outside instead…
Connecting was a bit slow but acceptable.
One BT disconnect when I stopped to go get a waterbottle. Came back to no gear indicator. Started pedalling and after about 20 seconds or so the gear indicator reappeared and resistance returned to normal.
One note: the fan I have operating off of speed (Ant+ from Justo 2) did not start-up until after the gear indicator returned.
Makes me wonder if the problem is the trainer?? Because the PC/Zwift is not involved with that connection at all, right?
One other thing you could try is disconnecting the trainer from your network. I have no theory about why it would help but it is a thing you can do.
Running Zwift on a different device such as a phone could also be tried.
Makes me wonder if the problem is the trainer?? Because the PC/Zwift is not involved with that connection at all, right?
Trainer? Todays tests were inconclusive. A different trainer but there are still disconnects. On the climb I made more rest stops then there were disconnects. Resistance was lan although that still uses Bluetooth packets.
I am not skilled in interpreting the results but perhaps doing some Wireshark recording would identify a pattern. No promises but I will see.
I don’t know if the trainer requires some sort of data from Zwift before sending out the Ant+ stream. Tonight most of the ride I was going too slow for the minimum 8kmh fan start. I forgot to switch the Wahoo fan from manual to Speed at the top.
Short first ride and then AlpDuZwift which for me was a 3 hour trek.
I did stop and change to climbing gearing and then stop to adjust trainer difficulty. No idea if those changes cause resets or how that is handled.
@AlanZ is there an obvious reason for the disconnects in the following ride?
Ah you have a second trainer, that is a great test.
I’m back to thinking you should substitute something (anything) for the PC as a test.
I also note that your PC performed significantly better on the Alpe ride even though the level of crowding was about the same and it’s in the same game world. Presumably you have a 60Hz display and on the Alpe ride you were at 60 FPS much of the time.
Ah you have a second trainer, that is a great test. I do now.
I’m back to thinking you should substitute something (anything) for the PC as a test.
I will try my s10FE tablet. Before I bought this laptop I used a 10" Samsung S8 tablet but I have handed it down to a nephew. I don’t remember losing the resistance for the first couple months using VS and then believe it got worse which to me implied software updates. I bought the Justo 2 August 2025. Aside from the trainers the timeline of which devices I was using has faded. I do remember using the A10 laptop a few times but cannot remember if that was just for the bigger screen or troubleshooting.
I also note that your PC performed significantly better on the Alpe ride even though the level of crowding was about the same and it’s in the same game world. Presumably you have a 60Hz display and on the Alpe ride you were at 60 FPS much of the time.
Interesting. I missed that. I wonder if it was something different with the trainer or if the room was cooler last night. Maybe there has been thermal throttling? I looked at temps when I accidentally opened the monitor program by hitting the touch bar a couple rides ago and gpu and cpu were hitting 90C. That monitor pop-up didn’t show frequencies though.
My to do list now includes riding in a more congested Watopia area, on the Neo, with a fan blowing on the laptop. Then the tablet. I am also curious about the built in BT4 now that the resistance is over lan.
Another hmm, I have race mode turned on. I don’t know how much extra load 10Hz sampling puts on the cpu. In the world of GHz processors I didn’t think it would make any difference.
So many variables.
I also do not know that, but there are several discussions here on the forum that mention improvement after disabling it.
I did look at your TrainerDX reports while asking the question whether dropouts appear to be associated with frame rate dips, which can be a way to spot an exhausted CPU (or throttled CPU) and I did not arrive at the conclusion that they are related, but it would not be at all shocking that an ancient PC has cooked thermal paste, or is packed with dust, and needs to be serviced. You had dropouts in the ride where the laptop seemed to be performing about as well as one might expect given its specs.
Zwift running on Samsung s10FE. Tacx Neo3m.
The good
No disconnects.
I am impressed how well the Elite Rizer is working with the Neo. Perhaps the Neo does send out gradient in the Ant stream because Garmin also has an incline simulator. The only comment here is the Neo is higher than the Justo2 so setting the zero point uses up part of the maximum incline. A spacer under the Rizer might correct this but would need to be very stable. Once the Alpine arrives I will compare.
The less good
Some odd behavior like coasting to a stop on negative 4% grade. This was particularily noticeable because the Neo 3m virtual flywheel kept spinning. When I stopped at the finish the virtual flywheel kept spinning until I exited the game. This did not happen on PC. Will do another ride with the tablet to confirm behavior was not a one-off.
There were resistance variations. I believe they were triggered by rolling to a stop for some seconds when I was playing with the Wahoo Headwind fan.
At times while pedaling at a constant 25kmh the Wahoo fan set to Speed input kept starting for a few seconds then stopping then starting, etc. Would be interesting to know what is used for the speed value that is transmitted on Ant+ (and/or Ant-FEC) from the Neo 3m. Because Ant is not secure Garmin was required to add toggles to enable/disable Ant (and Bluetooth secure mode). I have both Ant+ and Ant-FEC enabled during my efforts getting the Rizer and the Wahoo fan to work. I don’t know which type of connection is correct and whether they interfere with each other. The Garmin app implies Ant+ is data from the trainer to devices and FEC is for control of the trainer but I had trouble getting the fan to work until I turned on FEC. Maybe I should have turned off Ant+? I really wish companies would stop putting “simply” in directions that are “simply” out of date. Nice that Wahoo added extra features to the fan but to get the Wahoo app to see Ant devices requires the device running the app to have Ant. The app doesn’t act as a display for connecting the fan internal Ant radio as one might hope, the app needs to find the device first. Fortunately my old backup phone has Ant but the majority of people won’t have a device that supports Ant. The Wahoo instructions to “simply” press the power button for 3 seconds when in Speed mode to pair does not cause the light to behave in the described manner. It did finally seem to connect tonight using the button method but grrrr not smooth at all and then the erratic behavior above.
I am guessing this is what happens when there is not a proper agreed protocol that supports all use cases so hacks and bodges are pasted on? Of course it is the users that just want to ride that are left dealing with this.
S10FE. Neo 3m.
Headwind should automatically operate on the last input selection. Didn’t and I didn’t bother with it.
I think you should direct the fan questions to the manufacturers. You bought expensive accessories and they owe you that support.
The tablet looks like it’s working well.
Streaming apps are not a good way to test a network because they buffer the content to handle brief network interruptions or congestion. That does not simulate how Zwift uses the network.
I experience this same behaviour with mye Neo 2T from time to time. I have read about others also experiencing this. It doesn’t bother me. I have no connection problems.
@Paul_Southworth I was surprised to see the disconnects on TCP. I am on fiber internet and just have Zwift and YouTube on one TV going. I don’t notice it affecting gameplay so it may not matter. Good point about the streaming apps buffering but the game data being real time.
I have been browsing the interwebs regarding the fan. I believe Ant+FE-C is the correct choice so I turned off the regular Ant+ now. I don’t remember the option on the Justo2 but I may have missed it. Online info suggested both Ant+ and Ant+FE-C are both used on the Justo2 as not all devices support FEC. The FEC is bidirectional according to Google. My legs need a rest so it will be a few days before I really test it thoroughly but I did get a connection and fan spin with Ant+ turned off on the Neo.
@Audun_Simonsen_Milto I accept the virtual flywheel behavior may be a limitation of the protocol. If Zwift is not broadcasting the in game speed then the trainer can only rely on a computed value (fancy guess). The virtual flywheel and the Headwind fan therefore won’t be in sync with the game on screen speed.
I am learning the ways the Neo is different than the Justo2.
For example on the Justo using the mechanical gears to allow a faster cadence on hills works fine. I have used 39f -42r on the AlpDeZwift in addition to virtual gear 1. The mechanical mag brake is not quick responding but it is powerful.
This does not work on the Neo for 2 reasons. When I shift to an easier mechanical gear the torque feels the same. I believe the Neo senses the cadence and compensates for the gearing to deliver the “correct” feel. Second, once I was on the 32 or 36t cog I experienced slippage. Looking online the Neo1 and Neo2 are rated 85Nm torque with the marketing blurb that Neo2T and Neo3m are “nearly impossible” to slip. Well they do slip at a torque I can easily achieve with 39f-36r. Helpful advice they offer is to spin at 95rpm lol. Trying to raise my cadence is why I shifted my mechanical to start with. On the Neo the solution is to adjust the in game resistance down. This is just a trade-off for the neat virtual flywheel and road feel features that rely on the Neo’s electronic mag brake.
I was actually more confused why I had come to a stop in game while on a negative 4% grade.
In your 49 minute ride you had a TCP disconnect right at the start. There’s no sign that the 3M is affected by this despite being connected via the LAN. The disconnect could have been related to an Internet problem which would not affect communication on your LAN, or something the trainer handled seamlessly by retransmitting.
In your 2:05 ride, you had the disconnection later and it’s associated with UDP timeouts communicating with Zwift, but not with trainer disconnections, so that’s the same scenario - trainer has no issues or it handles whatever happened, but you get disconnected from Zwift’s servers for some reason, could be a problem anywhere between you and Zwift on the west coast of the US but it wasn’t something awful on your LAN or I would expect the trainer to disconnect as well.
