Are they correct?
They told me:
My drivetrain is clean/oiled.
If they are right then can’t wait to switch to pedal/crank power and have an immediate 12% boost on TR/Zwift.
My performance on Zwift won’t stay the same.
From my experience 3-4% difference is about right. Strange that Tacx adamant that there is such a large difference.
I have always used a crank PM, which came to the end of its useless life last month. My Neo power is reading lower than the old crank power. Interesting in whether anyone agrees with the 12% drivetrain loss or not? Seems high. Strange that would be Tacx official position!?
Aware that changing power meters does not make you faster in real life.
hm just from memory the numbers I hear thrown around the most is 20 - 30 W loss from a DIRTY drive train. So from that standpoint 12 % is realistic. Obviously you can reduce this quite a bit with a proper maintained drive train. I can’t recall a specific number right now, but I can’t imagine it should be more than maybe 10 W.
What crank based PM are you using?
I’ve never seen 12% loss in any testing of any trainers or power meters. And I’m pretty sure Tacx engineering wouldn’t agree with that either (I know they wouldn’t).
The highest I can ‘see’ is about 3-4% on a dirty chain from pedals to trainer. And that’s fairly rare. Usually it’s around 2-3%.
Isn’t the power meter on a trainer always lower than a power meter on the cracks/pedals because you measure farther away from the force applied or is it just a calibration thing.
Thanks. Exactly, I can’t imagine that anyone (engineer or athlete) would be OK with that, wanted to check I wasn’t losing my mind. I have (cheekily) asked the Tacx support team for their data / evidence that minimising lost power to 4% is impossible. But I would be genuinely be interested in the results, not something I could easily see other than anecdotal forum posts etc.
V2 - thanks agreed would definitely expect some lost watts through the drivetrain, question was more what % others have experienced.
Agree 12% is a lot. I have had 2 Tacx Flux 2 Smart T2980 with a big difference in power. I tested it by doing a ride on the taxc app while also measuring my power with a 4iiii meter, the difference was for sure around 12%. The trainers had also other flaws so they accepted the returns. Expecting a third one in a few days (fingers crossed I guess).
Btw, my drive-train was as good as new, waxed etc…
Yea that sounds like a specific trainer issue if you had to return it (hopefully). Also there may be left/right imbalance with Precision single sided but that can be tested by pedaling with one leg.
I have looked around a bit more. Tests by ceramic speed below show around 2.5%-4% loss in efficiency: Test Data & Reports for CeramicSpeed Products | Read The Tests here
Clearly doesn’t matter a whole amount on TR if you always use the same PM. It does make Zwift races more challenging however if you give crank based PM users a 3% headstart.
Thanks for the input.
"Are they correct?
They told me:
brand new chain = 2-3% loss.
12% loss is realistic"
Well, I’d say you were talking to SUPPORT and not ENGINEERING. 
Losses aren’t going to scale as a percent, anyhow. If you’re FTP is low enough, it could be 12%, sure.
I would say, based on the data I have, that if you saw a 5 to 10 watt loss between your pedal-based power meter & your hub-based power meter that wouldn’t be unusual. I’m sure you can make that worse if you do ridiculous things with your chain, cassette, & derailleur.
I’m running Assioma pedals and an Elite Direto trainer, and my pedals are ~2% lower than the trainer across my whole power range.
Difference between Tacx Neo and Favero Assioma has never been more than 3-5W across power range 150-400W in any of my indoor indoor rides. That’s using both round and osymetric chainrings. Assioma have been my most reliable power meter so far.
“I’m running Assioma pedals and an Elite Direto trainer, and my pedals are ~2% lower than the trainer across my whole power range.”
I’ve seen that a lot! More often, I’ve seen power curves that match pedal/crank based power meters almost exactly to a hub-based power meter. And users typically seem very pleased with that! Ha!
In reality, pedals should always be a little higher than hub-based because of drive train losses.
In the end it doesn’t matter. If you become serious about racing on zwift and are good, you’ll need 2 power sources anyway as well as a confirmed weight (basically like a passport)
In my experiences they scale closer to a percentage than an actual value. I have P1’s and a GS hub. Cleaned and lubed chain I see around 2% loss. A chain that needs cleaned and has lost all lube I’ve seen close to 10% loss. And those losses have been measured over different intensities which is why I’m saying it’s not a fixed wattage number.
Brennus - agreed on support vs engineering point. Still surprised that support would be so definitive (and wrong) on this.
Agreed it doesn’t matter. No real TR impact assuming the drivetrain is well maintained and FTP is retested fairly regularly / each time a PM is changed.
@cwiggum, you are absolutely correct. I am wrong.
I just had this argument privately. I’ll concede that power losses vary linearly with rider output. You just gotta really zoom in on the y-axis to see it. So on a good chain the difference between a 200W rider’s drive train loss vs a 300W rider’s drive train loss is maybe a watt. But it does vary linearly.
Just to give everybody an idea, the difference between a new chain/new cogs and a worn chain/worn cogs at 250W is about 4W. Or so sez Jason.
Using Squirt to lube the chain vs WD40 to lube the chain is about 2.5W (at 250W rider output).
No, they are not correct, but if they can get you to accept this then they can close the ticket without actually fixing anything. (OK, maybe 12% could be realistic if the whole drivetrain is covered in sticky mud, but I think it would be something like that, or everything covered in rust to be in that range.)
I had a friend who had a neo2 that was 10-15% off, and it took him months of back and forth to get them to do anything about it. He ended up doing power comparisons with 3 different power meters showing that it was way off from all of them, then they finally relented and admitted it was their problem.
Tough if you don’t have multiple power meters to compare against…
Here’s a suggestion for @Nate_Pearson for when Trainerroad could have interns in the office again - have an intern test all the trainers you have against a crank based + pedal power meters, and publish the results. A crank + 2 pedals gives you 3 independent readings to compare against the trainer. This would be a very interesting accuracy survey of a population of trainers.
