So this evening I did my first indoor ride for seven months. It was also my first time on a smart trainer (Saris H3) so there’s a bit of a learning curve going on as well.
Erg mode just didn’t seem to be the experience I expected. I’m not sure if it’s the trainer and SW working against one another.
Here’s the workout: https://www.trainerroad.com/app/career/bobw/rides/121481999-bluebell
The resistance did alter correctly for each interval but I wasn’t held to it, if that makes sense. My cadence was reasonably constant but even a drop of 1 RPM would lead to a massive drop in power, maybe 10%, with consequent rebound when I increase cadence. The power profile is as erratic as if I was in resistance mode.
I had Erg mode enabled in the app, Powermatch was on (latest mode). I played around with sensitivity/responsiveness but that didn’t seem to make much difference other than change overrun and the like.
There may be an app for the trainer (it’s my wife’s so I’ve not messed around with it) with some options.
Is all this normal for Erg mode?
looks pretty normal to my experience - when you see the flat lines in others views it usually a Whaoo that has smoothing - its actually a false perspective. ERG gets you close but doesn’t lock you in.
I find with ERG that is really responsive to cadence. The TR screen doesn’t show the the expected cadence range, but you will notice if you go too high or too low you find the ERG doesn’t respond very well . Too high and the ERG gets really soft, and too low and the ERG will do the spiral of death.
This depends greatly on your expectations, and what you did in the workout.
There is a common misbelief that ERG applies “perfect power” and holds you to very precise power targets. To a degree, it can, but it is not the magic control that some think it is.
ERG is a system of your input (pedaling cadence), coupled with a defined power target, and an app that works to set the resistance level on the trainer to obtain that power target. In a perfect world, we apply smooth cadence and force, along with a rapid recognition of the app and response of the trainer in this “dance” that is solving the power equation… that is ERG in a nutshell.
This gets to my primary rule about ERG… “Cadence is King… Ignore power data.”
If you are watching power, and altering your input in an effort to “fix the ERG problem”, you are causing the problem… not solving it.
As mentioned above, this works based upon your own cadence and force input to the system. With that info, the system (app + trainer) work to hit the defined power target. If you alter your cadence, you force the system to change.
Of course, we are not machines, and we can’t produce “perfect cadence” or power, and the fact that delay in the system recognition of the need to change, and the delay in applying a resistance chance will lead to inevitable fluctuation of power output. This is not that bad of a problem, and it’s the reason I recommend people stop watching power when they swap to ERG, at least initially.
With our history of working in Resistance mode, we naturally want to “fix” the power target with our input. That is necessary for RES, but harms the ERG process. So, stop watching power data, focus on holding your cadence smooth and predictably in all times other than when you want to make a deliberate change in cadence, and you should get better results.
Per your workout, here is a clip of the last set:
- Notice that you show a clear and steady ramp up in your cadence. Based upon your comments, it seems you were “speeding up” your cadence to “fix” the dropping power you saw and interpreted as a problem (at least that is my take on the evidence shown).
In case you haven’t seen it, this is what I consider to be required watching for new ERG users. That and doing your best to ignore power (let the system do the work), and be steady can make the whole thing work as intended.
Edit to add, here is my workout done on a Wahoo Kickr 17, in ERG with power from the Kickr (ERG mode smoothing off).
- Note that my cadence is “flat” and even with that, the power fluctuates. This is NORMAL and to be expected. Again, “perfect power in ERG” is a myth that needs to die.
Yes, it probably was user error! I wasn’t expecting “perfect” response but the magnitude of the swings did surprise me. As @mcneese.chad notes there was definitely a bit of chasing the target power.
I think there’s an effect of the trainer’s flywheel as well so I’ve got several new things to deal with at once.
Maybe not the best workout to do as my first in Erg mode but autumn has arrived with a vengeance and the roads around here are now covered in leaf slime which is like riding on ice. The next workout is Warscheneck which is much more steady state. I might cover the power figures on the screen so I can just focus on cadence.
Good deal. Just a couple of random things worth stating, if you haven’t covered them.
Always retest your FTP when making “big” equipment chances. This includes when you swap between power measuring devices (power meters, trainers, virtual power, etc.) as well as just plain switching trainers. Even if you have the same power meter and use TR PowerMatch, the difference in flywheel can be enough to alter the results of an FTP test. So, if you haven’t already done so, make sure to retest when you can.
You mention flywheel, and the Saris Hammer series are awesome for how large and fast it spins. But this great feature can have negative consequences with some workouts. Specifically noting that it will hold that speed well into a recovery interval in some cases, especially for harder efforts and if you are using a taller gear in ERG (like 50t up front).
The other Chad beat me to it but most smart trainers hold power more consistently at lower speeds. Your power graph looks normal to me, but make sure you’re in the small ring up front and pick a gear in the back that gives you a straight chain line to minimize chain and gear wear.
Yours on Saris H3:
One of mine on Kickr 2017 direct drive with PowerMatch and left-only Stages PM
You can see I was putting down a little too much power at the start of most intervals.
and another one with dual sided Stages PM, a couple months later and with better muscle control:
Yours look normal and fine to me. Possibly smoother than you could do outside.
I’m currently working with an FTP value slightly less than one suggested by intervals.icu (288W vs 292W), felt about right. I’m actually just starting three weeks of taper to an ‘A’ event at the start of November so not too worried about my FTP being out of whack - the event is nothing like a TR workout - 300km with 7500m ascent on MTB in Mid-Wales.
Yes there was quite a bit of overrun when I dropped to the recovery intervals and it took maybe 15 seconds until I could actually apply pressure to the pedals.
@rkoswald I was in small ring (34T) and the 16T out back.
@bbarrera - I had powermatch with left only 4iiii PM.
Again, looks normal.
Do you think there are some magical gains from unnaturally smooth power charts? If there was evidence for that, the smart trainer companies would have been screaming it from the mountain tops. Don’t overthink it. Big ring vs small ring - use what feels natural.
No, I knew that some (didn’t know it was Wahoo specific) trainers artificially smoothed the output, I just wasn’t expecting so much hysteresis.
Another learning curve
Your brain will adapt and come to think of it as normal, and FWIW, this is how I felt about my first experience with Erg:
After two years with erg I decided to go back to ‘fluid trainer mode’ and make inside as close to outside as possible This winter I’ll turn on Zwift or RGT and use sim mode as I believe that was even better (at making inside as close to outside as possible).
I have the same trainer and my power doesn’t fluctuate as much as you’re describing with slight variations in cadence using erg.
I train almost exclusively in ERG mode - Using a Kickr Core - My power is almost exact but i do have it on 3 sec smoothing. My workout looks like this.
One problem i experience is when i come off an interval into a rest interval the trainer tends to overspin with the sudden drop in power - its like peddling in air, no tension on the trainer - TR app thinks i’ve stopped and the program pauses, once the flywheel has slowed down and resistance returns to the peddle the program starts again.
Is this normal ? Or is there something i should be doing ?
This is normal. I have the same experience. Especially going from something 125-130% to 50% you sometimes pedal into air. I always use the big ring in front. Middle in de rear. Switching to the small ring in the front should help this a bit but i didn’t find that much of a difference.
note: for me it’s almost never long enough for the app to pause the training.
As Chad said, ignore power. You have to trust the trainer is doing that for you. Just focus on a smooth consistent pedal cadence. It takes a few rides to get used to it.
That’s a consequence of the large flywheel - the inertia means the rotation speed takes some time to decay, it’s the “road feel” that higher end trainers use as a selling point. As I noted above, the H3 takes about 15 seconds to “ease up” to the point that I can actually apply meaningful pressure, in the interim I just keep pedalling - the cadence sensor in the 4iiii will ensure that TR doesn’t pause the workout.
Your graph shows that you still have the details “Enable Power Smoothing in ERG” on for your trainer.
I suggest turning it off to show more accurate power data, not the overly fake smoothing Wahoo seems to think we need.
Yes I would agree, but it looks so much better with smoothing on
But i mainly focus on keeping my cadence as smooth as possible
Do the different trainers have differing hysteresis/lag in how they respond to slight variations in input power/cadence? In the video you linked to earlier (I’ve seen it before BTW) it seems like his trainer responds within a second or two which is what I was expecting whereas my (well my wife’s) H3 is more like ten seconds and then seems like it overshoots. This may be a setting on the trainer itself, dunno at the moment.
In the first video at around the 6:40 mark he varies his cadence quite markedly (between 80 & 120rpm) but there’s little to no change in the displayed rider power (<5W). That’s what I was expecting but it would appear that what’s being displayed is smoothed.
In fact, you make a good point here - when i first got my Kickr it never transmitted cadence, I used my garmin cadence sensor, then after a firmware upgrade this feature suddenly appeared and i took the sensor off - think ill put it back on, then the program won’t or should not pause if i’m soft pedaling waiting for the flywheel to slowdown
Or am i missing something here ?