I've had my Kickr for over 2 years and I really don't see that. I've been sick lately, but I did an experiment this morning at an easy 120 watts for 15 minutes. First 5 minutes was warming up till I got cadence to 90 rpm. Then each minute I went from big chain ring down through 5 gears, then small chain ring and up 5 gears. Took about 10 seconds to get my cadence back to 90 as the shift would make the initial power different, but once I was back to 90 rpm, the effort felt exactly the same. Only different was the sound of the kickr with flywheel speed. Big chainring, small gear would have the flywheel spinning pretty good and nice loud whinning. Small chainring and big gear was so quiet I could hear the chain noise.

So through 10 different combinations of chain ring/rear gear, I honestly didn't feel any perceived effort difference. My heart rate stayed relatively constant too. I've done similar experiments when I first got my Kickr till I found the 'best' combination that resulted in a very quiet Kickr. I only ride in ERG mode so IMO the flywheel speed doesn't seem to make any difference to me. I usually ride with the small chain ring and about the 3rd gear because that keeps the chain lined up the best.

Karl,

I wouldn't expect to see any difference at a very easy sub-threshold effort either. Try it again at 220 watts and report back.

Hugh

---

Genetics load the gun, lifestyle pulls the trigger.

The guy you quoted said "at 250W..."

I could tell little difference up to ~150W. Over 200 and it was night and day.

I don't notice an effort difference in the various flywheel speeds either, only road feel differences.

For my kickr, it clearly has more trouble maintaining ERG resistance levels at high flywheel speeds and thus the effort can be much more spiky-variable.

• Low wattage efforts at high flywheel speeds are highly variable (130w at 35kph kickr speed)
• High power efforts with rapid changes in resistance take longer to stabilize. (300w-340w-300w-340w)

Low flywheel speed feel worse from an road-feel perspective but my kickr is able to maintain the resistance level much more stably.

ST should really do a tech article about the physics of smart trainers so people can understand the physics behind control systems of smart trainers.

I'd love it if the article included tips for avoiding the cadence death spiral. Once I'm spent and my cadence drops below 80 I've got 20 seconds or less to total implosion. For a "smart" trainer it sure doesn't have any mercy.

Haha...220 is my old FTP. I'll try again once I'm feeling better.

That said, when I was in much better shape and pushing well over my FTP at times, I did similar experiments with the gears and still didn't notice that much difference. Heck, even when I'm going 250 watts, I don't even notice what gear I'm in and just have my head down and grinding away. I do have to point out, I only weight 128 lbs and so my power to weight ratio is in the 'decent' range. I'm also pretty damn old.

I'm with you and hence the tone of my posts. I have never felt one iota of difference when using erg mode from one gear to the next regardless of wattage, once the trainer micro-adjust to the new cadence/gearing/whatever. My FTP hovers between 260-300 depending on my training and level of give-a-shitness, and much of my training was at threshold and above. I bought a KICKR as soon as they were available in early 2013 and used it for nearly three years before getting the Tacx Neo Smart in December 2015. As mentioned before, no flywheel on the Neo so I wouldn't expect the same feel or inertia effect on that but I never noticed it with the KICKR either which has a flywheel. I just didn't. I only changed gears to avoid wearing down the same gear.

Additionally, my n=1 was always backed up by articles from TrainerRoad, Zwift, Wahoo, and so on. I'm not saying there isn't an inertia effect from the flywheel as many of you pointed out or corrected me on. I've just never experienced it with either or the smart trainers I've owned.

---

Favorite Gear: Dimond | Cadex | Desoto Sport | Hoka One One

Thought I'd post an update after one week at it.

Observations:


1. It's all about the back half


Historically my power output generally drops off as fatigue sets in. On a dumb trainer I might hit the TR blue lines on the front half of a ride but would usually be below them on the back half. Riding in Erg mode that's not really an option. Either you hit the power numbers or you come to a halt. This changed the way I ride completely. You almost have to sandbag the front half to nail back half targets. I think this is going to change the way I train. I'm curious to see what happens when I bump up to a new FTP.


2. The cadence death spiral


Holy smokes is it sensitive to cadence. I can tell when my cadence is dropping because the trainer adds more resistance to try and get the power up. Even dropping cadence from 87 to 85 there is a noticeable change in resistance. At that point you have two options. The best option is to try and get your cadence back up. But doing so means sending your power well above the current target, usually by 20-30W. Early on in the week I didn't have the ability to do that. By the end of the week it was getting better. The other option is to slowly watch your cadence drop to 0 as the resistance continually rises (aka the cadence death spiral). I found myself in this trap once. And it sucked.


3. Not all smart trainers are created equal


As I found out here different trainers (wheel on vs direct drive) have different responses. In my case the gear I'm in matters. My early rides were in the easiest gear. As power targets went up it felt like I was riding uphill. According to my speed sensor I was putting out ~9mph on +200W. I could see how this could be beneficial training for climbing. As the week went on I switched to big chain ring / higher gears and cadence / speed / power output felt much more normal. This does not seem to be the case from others with other setups; just my n=1.




Remaining Questions:


1. Dual FTP?


I'm considering turning off my Vectors in TR and doing a new FTP test. If I do I'll clock my test intervals on my Garmin so that I can get FTP (trainer only) vs FTP (Vectors). Initial observations are that my Vectors run +20W over the trainer. Doing a test this way will do a few things. First, it will give me a legit FTP for TR so that my sessions are more meaningful. At the same time I'll have the FTP for my Vectors that will hopefully translate to a closer outdoor number and get me more meaningful TSS numbers in TrainingPeaks. I'm considering doing this just to test the feel of simple Erg mode vs powermatch Erg mode. I'd love to hear from anyone that's done anything similar or recommendations for / against.


Thanks for the feedback as I navigate new waters.

I've been doing a bit of digging regarding the topic. The British time trialists seem to take how the bike on trainer reacts to pedaling way more seriously than the average Slowtwitcher. The whole thread is an interesting read but here's a single interesting post. This post is regarding riding in slope mode but I think you can still apply it to think about erg mode.

http://www.timetriallingforum.co.uk/...php?showtopic=103637

Posted 26 October 2015 - 01:14 PM
Clayton has generously provided me with a file from the Wahoo Kickr to analyse. This is the result:



Again, it doesn't match the simulation parameters that he specified. For example, he specified a mass of 74kg, but it has only provided an inertial effect equivalent to that of a 14kg mass. There is also something a little odd going on in terms of Crr, with the best fit occurring with a Crr of zero. Even then, there is no perfect overall fit between the model and the actual speed data. Matching the speed at ~300W leaves it a little low at ~100W. The only way to resolve this is to reduce Crr, but even with Crr=0, it doesn't quite match. I can make it match by taking Crr negative. Crr -0.015 and CdA 0.308 makes the speed match at both powers, but a negative Crr is kind of silly.

I think this strange effect is probably down to the Kickr fundamentally just being a magnetic trainer with some automatic control of resistance, so it doesn't naturally have a power curve that matches outdoors like a wind or fluid trainer does. My feeling is that the inertial effect is most likely just that which is provided naturally by the geared flywheel, and there is most likely no high frequency adjustment of resistance within the pedal stroke to simulate inertia."


Here's a post from JV Coaching Alpha Company that addresses the Kickr in erg mode.

"
KICKR Training Notes:
Often a discrepancy can be found between the power values of the KICKR and your power meter. This is attributed to the differences in inertia of a cyclist on the road, versus the inertia of the KICKR's flywheel. Cycling on the road is a high inertia environment, demonstrated by the degree to which the cyclist can coast when not pedalling, versus the rapid decline in speed of a fluid or magnetic wind trainer when pedalling stops. Due to the KICKR's heavy flywheel, it carries a lot more inertia, which allows it to have the 'road-like' feel that is appealing to cyclists. But, depending on the speed of the flywheel, it's inertia changes markedly.
Considering the ERG mode of the KICKR is generally the most used programme for a cyclist when undertaking strength work, use of the small chain ring will generate a lower level of inertia requiring the muscles to work more consistently throughout the entire pedal stroke, especially through the 10 o'clock to 2 o'clock position where muscle strength is lowest. Often an 8-10% difference in wattage can be seen between the KICKR's power output and that of the cyclist's power meter due to this difference in inertia, and also that the cyclist's power is being read at different points in the drive chain (i.e. crank arm vs. spider vs. rear hub/KICKR). Use of the large chain ring results in the KICKR flywheel having a higher rotational inertia, more in line with that of the cyclist riding on the road. In this situation we usually see the difference in power readings much smaller or the same."




Hugh

---

Genetics load the gun, lifestyle pulls the trigger.

I think the question really is when using day the small chain ring you are pushing the same watts but at a lower wheel speed. I have a computrainer and maintaining day 250 watts at 90 Roma is way harder in the small chain ring versus the larger chain ring because of the wheel speed and inertia.

Great thread as I was thinking about this when it comes to thinking about a direct drive trainer purchase like the Neo. I'm still a little confused by the "slope vs. erg" mode and maybe that factors into my question. So can it simulate hill climbing basically running at like 80 cadence but in the small chainring or should I care? With my current cheap magnetic trainer, I always had to use the big chain ring for climbing intervals and it just felt "off" which could be all mental and not real.

Thoughts?

With the NEO when in slope mode you can adjust the slope to a positive value which will increase the work needed to ride a specific speed as well as decrease the virtual inertia of the system i.e. make it like riding up hill. For simulating riding on the flats in aero position setting the Neo to a -.5* slope makes for a very realistic inertial feel. In Erg mode the folks at Tacx chose a lower than realistic virtual inertia so that the unit would react quicker to changes in power settings. With the Neo in erg mode the virtual inertia is constant whatever gearing you choose. With a Kickr in Erg mode the inertia is greater with a higher gearing due to the increased flywheel speed.

Hugh

---

Genetics load the gun, lifestyle pulls the trigger.

Just wanted to post a quick update. I finished my four week build and then disconnected the Vectors from TR. Did a new FTP test and tracked my test intervals on my watch. Bushido FTP was 202 (via TR), Vector FTP was 223 (via watch). Today I started a new build (8 week 40k TT) in just Erg mode w/ no powermatch. It was probably the smoothest feeling ride since I moved to a smart trainer. Trainer power was extremely responsive to cadence pick ups, much more so than when in powermatch. It felt like on my powermatch rides the trainer brake was constantly adjusting; it was probably doing the same thing in Erg but was much less noticeable. In powermatch a 2-3 bump in rpm always seemed to produce an initial +30W bump; over the course of an hour those just hurt. In Erg the feedback loops feel MUCH quicker, to where that the same rpm adjustments produce a 5-10W initial shift that settles back down to the prescribed power almost immediately.

I think this is going to be my goto setup going forward. I swallowed my pride and watched my TR FTP drop by 20W but the NP from my Vectors this morning was right in line with where I've been in the past. Hopefully this info is useful to someone else starting out on this journey.