Jet Black Victory

[Mike_22W (Mike W)]

@Fab_spinbike_slayer

to compare data between the two systems

For this you need to get the activity file from zwift and a second from your Assiomas.
So I connect my Assiomas to my garmin head unit and record an indoor ride - then I go to garmin and download the activity as a .fit file.

Then use something like DCR Analyzer Tool Overview & Manual or the tool within zwiftpower;

That gives you the option to pick a zwift activity or grab the .fit file from your zwift folders and load it against your garmin (or equivalent).

You need to do a bit of cropping and alignment to get them to line up.

[Lester_Belen (Lester)]

My Victory is getting replaced. Sent them a video of it making the clanging noise when going uphill on gradients around 8% and above with trainer setting at 100%. That and the innacurate power readings. Hopefully the replacement will be much better, and soon. Appreciate JB being on top of this and making it right.

[Mitch_Godfrey (Mitch)]

This is precisely what mine is doing against assioma duos. I also have a Wahoo Kickr bike that I tested using the same assioma duos. Those two match nearly perfectly. I went from the 4.9 JBV firmware to the 4.12 and it seemed to work for a day. Then it went back to being off the same amount as I had with the 4.9 firmware. Which is that the JBV is about 10% lower than the Assiomas.
If it was 10% higher I would be good with that!

[Textuality (Richard)]

Might have to test that on mine.

[Alex_Beebe (Alex)]

Feeling slightly better about power accuracy after last night’s Rolling Highlands TTT with repeated punchy efforts. I was able to quickly grab a temp reading with a laser thermometer between pulls and saw ~58ºC (136ºF). Admittedly, I’ve been pushing mostly Z1/Z2 watts on most of my rides since delivery due to the Festive 500 and sicknesses and haven’t really put the trainer through its paces until recently. Hopefully the bearings have settled into place for good.

I still think there’s a high sensitivity to axial shock during shipping that is messing with bearing preloading and air gap distance between the coils and flywheel. This would explain the belt alignment issues, power inaccuracy, and rumbling.

[Textuality (Richard)]

Have you made any changes to yours, or do you think it just “settled”?

[Alex_Beebe (Alex)]

I did add a small copper washer between the flywheel pulley and the inner bearing that I’ve since removed. Several posts up, I put some pictures showing how my belt was riding on the outside of my pulley and already wearing down through the black paint to shiny metal. I removed the washer after ~200km because it didn’t seem to help things as much as I had hoped, and I worry about galvanic corrosion with the copper.

[Kevin_Barkema_Iowa (Kevin)]

Is there a summary of software updates that JetBlack is working on?

Is there one in the works for WIFI issues and one in the works for power inaccuracy?

Not exactly sure what issues have been “officially acknowledge” besides the obvious mechanical issues (wobble, grinding, belt off).

[Fab_spinbike_slayer (Fabian)]

Does more than me have problems getting 4.12? I try to update via IOS and 4.9 is the latest version available…

[Paul_Southworth (Paul)]

JetBlack Victory and autocalibration issue

In order to acquire 4.12 I put in a support ticket to JetBlack and they emailed instructions on how to download that version. If you just “update firmware”, you’ll only get 3.9. Here’s a copy of the email that they sent me: > Please follow the below instructions. After uploading, please warm the…

[U_K_Bloke (U.)]

I don’t have the vibration problem, but my Victory is definitely reporting low Watts, probably 10-20% though I have not measured against another power meter. I think there is a general firmware problem with temperature compensation and auto-calibration. In my typical Zwift ride/race there really isn’t much opportunity to auto-calibrate because you have to pedal pretty much all the time. There is also something strange with the virtual shifting.

As others have mentioned the gap between gear ratios seems too large, and probably not what Zwift intends. When shifting up it is hard to get on top of the new gear, it’s like riding in mud. This is evident when coming off a steep incline onto an immediate flat (like elevator pitches in Zwift London). I am losing huge amounts of power/time here relative to other riders, it’s almost comical how hard it is to get going again. I feel like I am fighting against the firmware algorithm (using latest 4.12). My power/fitness/vo2max numbers in my fitness apps tell me that I am losing performance every day I ride because all the power stats are wonky. I tell myself that’s not accurate, and that I am probably getting stronger.

I am enjoying the trainer otherwise, and confident that JB will get on top of the firmware issues given a little time. I suspect that they added a feature or changed an algorithm, and that it has had unintended consequences. Maybe the auto-calibration is way too aggressive … in reality. temperature drift is reasonably slow and auto-calibration should adapt over an appropriate timescale, not large changes instantaneously.

Mark.

[Textuality (Richard)]

I’m going to choose to believe that my unit is also reading low, despite it reading higher than my cranks and my previous trainer.

/flexes mighty legs

[MarkyMarkZwifts (Mark)]

Since updating the firmware to 4.12 I haven’t had any major issues until today, during my FRR stage race. The wifi connection to the trainer dropped mid-race and when I reconnected it I was unable to use virtual shifting. Needless to say I came last in the race.

A couple of times now, when I have started Zwift, the virtual gears haven’t worked - there’s been no virtual gear shown on-screen and it’s hard to pedal. I’m not sure if this is a Zwift issue or a trainer issue. Anyone else experienced this? I’m thinking it may be since the latest Zwift update.

Regarding power, I have been dual-recording most of my races (when I remember to start the recording) and the power numbers seem pretty accurate.

[Yannick_Martinez]

I wanted to upgrade to beta version 4.12 but the update was not successful. I did several competitions today with version 4.9, and the power difference in favor of the jetblack victory compared to my assioma duo varies between 0.6% and 1.4% on each power data between 5 seconds and 5’
edit: I always calibrate the home trainer after 8’ 10’, I leave the pedaling for about 10 seconds for it to calibrate if that helps

[Alex_Beebe (Alex)]

I’ve had some time to think about this and test different things, and I’ve come to a similar conclusion. Please prepare yourself for armchair engineering…

Resistance on the trainer comes from two sources:

1. Rolling resistance from the belt, bearings, freewheel, etc.
2. Braking resistance from the electromagnetic coils and flywheel (eddy current braking)

Rolling resistance changes as the bearings and belt heat up, and spindowns currently occur with no resistance as far as I’m aware, so temperature compensation then is only accounting for rolling resistance.

Braking resistance is more complicated and is a function of voltage applied to the coils (magnetic field strength), design of the coils (number of turns, etc.), electrical conductivity of the flywheel (probably affected slightly by temperature changes), rotational velocity of the flywheel (measured with optical sensor and target, so controlled), and air gap distance between the coils and flywheel (largely controlled by manufacturing process, but could change due to thermal expansion). Since the spindown is done with no braking resistance, there is no calibration or compensation for changes. Voltage applied to the coils and design of the coils should never really change and the rotation velocity is accurately tracked, so there’s no need for calibration for those parameters. Electrical conductivity and air gap distance probably change slightly as the unit heats up, and this is why watts drift up during use (air gap at the very least should increase as the unit heats and the shaft/spacer expand and push the flywheel away from the coils).

Worth noting is that braking resistance is inversely proportional to air gap distance, and the distance on mine is already 1 mm or less (you can see the squares from the coils if you look through the gap between the flywheel and the plastic shroud). What this means is that a very slight change in gap distance can have a pronounced effect on braking resistance.

Wild assumptions ahead, but if gap distance is 1 mm, then a 0.01 mm change in gap distance yields a 1% change in power. For reference, a sheet of paper is ~0.1 mm thick (10% change in calculated braking resistance vs actual braking resistance). You can imagine how the slightest of movement, either due to thermal expansion or movement due to shipping practices could affect the accuracy of the unit. I’m especially concerned that shipping during the stressful holiday season could open the opportunity for mishandling of the unit which could cause subtle changes in the air gap distance between factory calibration and delivery. In other words, the flywheel and shaft basically act like a slide hammer, so if the package was dropped on its side (not an uncommon practice for FedEx), then the fly gap distance might be “adjusted”.

Since power is calculated as the sum of rolling resistance (linear increase with flywheel velocity) and calculated braking resistance (voltage applied, coil design, linear increase with flywheel velocity, and an assumed, fixed air gap distance and flywheel electrical conductivity), you can see where the error could come from. Especially important is the shipping “adjustment”.

So, what’s the fix? No amount of spindowns as currently implemented will allow for calibration of the braking resistance. What we need is an option for a “factory” or “advanced” spindown that Wahoo uses that does indeed calibrate using braking resistance. Otherwise, you’re putting the fate of accuracy into the hands of FedEx et al. Now, you might get lucky and have just the right “shift” occur to put things closer to factory air gap. I’ve seen other report this happening, and mine seems to have turned a page after ~800 km and some really hard (high heat) efforts.

[Fab_spinbike_slayer (Fabian)]

“I don’t think this is too bad. Favero has a ±1% accuracy, and JBV comes in at 2%. So theoretically, if one reads a bit high and the other a bit low, the difference could be up to 3%. For example, at 300 watts, JBV might show 294, while Favero shows 303—giving a difference of 9 watts, or about 3%. Even with this, they’re still within what they claim. I would be happy with your JBV, or am I missing something here?”

[Fab_spinbike_slayer (Fabian)]

@Alex_Beebe

I’ve had some time to think about this and test different things, and I’ve come to a similar conclusion. Please prepare yourself for armchair engineering…

Resistance on the trainer comes from two sources:

1. Rolling resistance from the belt, bearings, freewheel, etc.
2. Braking resistance from the electromagnetic coils and flywheel (eddy current braking)

Rolling resistance changes as the bearings and belt heat up, and spindowns currently occur with no resistance as far as I’m aware, so temperature compensation then is only accounting for rolling resistance.

Braking resistance is more complicated and is a function of voltage applied to the coils (magnetic field strength), design of the coils (number of turns, etc.), electrical conductivity of the flywheel (probably affected slightly by temperature changes), rotational velocity of the flywheel (measured with optical sensor and target, so controlled), and air gap distance between the coils and flywheel (largely controlled by manufacturing process, but could change due to thermal expansion). Since the spindown is done with no braking resistance, there is no calibration or compensation for changes. Voltage applied to the coils and design of the coils should never really change and the rotation velocity is accurately tracked, so there’s no need for calibration for those parameters. Electrical conductivity and air gap distance probably change slightly as the unit heats up, and this is why watts drift up during use (air gap at the very least should increase as the unit heats and the shaft/spacer expand and push the flywheel away from the coils).

Worth noting is that braking resistance is inversely proportional to air gap distance, and the distance on mine is already 1 mm or less (you can see the squares from the coils if you look through the gap between the flywheel and the plastic shroud). What this means is that a very slight change in gap distance can have a pronounced effect on braking resistance.

Wild assumptions ahead, but if gap distance is 1 mm, then a 0.01 mm change in gap distance yields a 1% change in power. For reference, a sheet of paper is ~0.1 mm thick (10% change in calculated braking resistance vs actual braking resistance). You can imagine how the slightest of movement, either due to thermal expansion or movement due to shipping practices could affect the accuracy of the unit. I’m especially concerned that shipping during the stressful holiday season could open the opportunity for mishandling of the unit which could cause subtle changes in the air gap distance between factory calibration and delivery. In other words, the flywheel and shaft basically act like a slide hammer, so if the package was dropped on its side (not an uncommon practice for FedEx), then the fly gap distance might be “adjusted”.

Since power is calculated as the sum of rolling resistance (linear increase with flywheel velocity) and calculated braking resistance (voltage applied, coil design, linear increase with flywheel velocity, and an assumed, fixed air gap distance and flywheel electrical conductivity), you can see where the error could come from. Especially important is the shipping “adjustment”.

So, what’s the fix? No amount of spindowns as currently implemented will allow for calibration of the braking resistance. What we need is an option for a “factory” or “advanced” spindown that Wahoo uses that does indeed calibrate using braking resistance. Otherwise, you’re putting the fate of accuracy into the hands of FedEx et al. Now, you might get lucky and have just the right “shift” occur to put things closer to factory air gap. I’ve seen other report this happening, and mine seems to have turned a page after ~800 km and some really hard (high heat) efforts.

Way over my head…
I had to use ChatGPT to explain this for me… like a grandfather explains to his little grandson…

Imagine you’re riding a bicycle, and when you pedal, there’s a bit of resistance pushing back—kind of like riding through deep sand. That’s what we call rolling resistance. It’s not a big deal, and it happens as things heat up while you ride.

Now, there’s something else called braking resistance. This is like when you squeeze your bike’s brakes to slow down. On your trainer, this resistance comes from magnets inside. The magnets pull on the flywheel as you pedal, making it harder to turn. The stronger the magnets pull, the more resistance you feel. But here’s the tricky part—if the magnets and the flywheel aren’t perfectly aligned, it can mess up how much resistance you feel. When the trainer gets warm or gets jostled during shipping, that gap between the magnets and the flywheel can change slightly, making the resistance readings a little off.

This means your power readings might not be 100% accurate, because sometimes that gap changes just enough to mess with the numbers. The best fix would be for the trainer to have a way to double-check and fix this alignment right at the factory—so it’s perfectly set when you get it. Until then, little things like moving it around during shipping can cause small changes, and that’s why you might see slight differences in what your trainer shows compared to your actual effort.

If its so sensitive, that makes it difficult to move around - for example if you have it stored one place and bicycle somewhere else… of if you have to wrap it up if you move…

[Mark_Gallagher (Mark)]

Agree, but it helps show that the Victory is a good trainer that’s just going through some birthday pains for some.

I’m not sad enough to do dual recording so I don’t know if I’m being robbed of watts, but my Victory experience has been largely pain free like Yannick’s.

[Wannie (JOHN)]

@Yannick_Martinez

I wanted to upgrade to beta version 4.12 but the update was not successful. I did several competitions today with version 4.9,

A few references to these version numbers. 4.9 is a higher number than 4.12, so this seems backwards. Is 4.12 a rollback that some have found works?

[Mark_Gallagher (Mark)]

Sorry but that’s not how software version numbers work. Version 12 of software release 4 is a later version than version 9.

I don’t know the full version numbering that JetBlack use, but for the company I work for it can have 5 parts, so it can be something like 4.9.0.18.6. But the last three parts are often only used internally within the company.

If I remember our system, it’s ‘major version.minor version.feature pack.service pack.build number’.