I would just like to say that it would be nice if ERG resistance cut out under 50 RPMs or so to prevent spirals of death. Having to wait for the workout to pause before I can start pedaling again is counter-productive.
Zwift does this, although the implementation is lacking IMO.
Cadence data in ANT+ only needs to be changed every second. Add to that time for the trainer to request a power target & it’s not enough.
Some trainers measure cadence themselves; they could do it in firmware quickly. It is annoying.
The protocols for both cadence & FEC (&BT equivalent) could also both be improved to having more frequency. That would allow software to do it to an improved degree.
Yes, TR sends the resistance change instruction 2 seconds before the start (and end) of an interval. It’s done to compensate for the inherent delay in most smart trainers, and hopes to align the power output more closely with each h interval block.
It can catch people off guard if they have a fast changing smart trainer (Kickr or Neo are good examples) and you need to be ready to apply power a bit early, but usually isn’t the cause of issues like you describe.
Currently, the best way to handle a Clinch or time when you have slowed your cadence too much in ERG:
Stop pedaling (if not stopped already from a Clinch).
Back pedal for about 5 seconds. (This signals to the app to release the resistance unit).
Resume pedaling with a quick and forceful effort. The resistance will be low, but there is a very small “bump” of resistance that you must overcome to get to a higher cadence like 90 rpm.
Continue pedaling with the desired cadence (resistance will be low and easy) and the app will raise the resistance to meet the current power target.
Finish the interval and workout as normal.
You mention an issue in a recovery interval. Are you having issues with the trainer not dropping resistance low enough in those easier efforts? If so, that can be a problem with some trainers if their resistance floor is too high. Which trainer do you have?
That may relate to the smoothing setting you use in the TR app. But it only alters the digital display you see above the power target value. So it should not really alter what TR uses behind the scenes, AFAIK.
Reviving this thread, as I would really appreciate an easier way to come back from a death spiral in erg mode.
I’ve tried switching to resistance mode, to manually force the trainer to back off, which lets me spin back up again. But when I switch back to erg, the trainer will clamp back down on the previous resistance and I’m forced to grind it out. This is even when I’ve made it to a recovery valley with low target watts - when I switch back to erg it will initially go back to the resistance I was in when I left erg.
Seems an odd behaviour to me. Unless it’s trying to punish me for my lack of concentration??
That’s the thing - it doesn’t back off no matter how long I wait or backpedal. I would have thought it might be an issue with my trainer (elite direto), except it will back off when I switch to resistance mode. So the app is obviously able to tell it to back off if it wants to.
I can keep pedalling in resistance mode 10, 20 minutes or whatever - but once I switch back to erg, it slams on the brakes.
It’s like the app is saying “oh, you’re going back to erg mode now? Let’s jam you back into the resistance you left erg mode at, no matter what the cadence or target wattage currently is.”
Zwift does this in workout/erg mode. Although its not perfect its a great feature. Go significantly below the interval target and Zwift automatically shifts out of erg mode to resistance mode then automatically shifts back to erg mode when your power stabilizes somewhere close to the prescribed zone. Its not perfect but I get way fewer lockups than I get on TR and its quicker to get out of them.
Zwift’s workout/erg mode also is more forgiving than TR’s in that its lets you wander a bit from the target and is more gentle in forcing you back which both provides for a more pleasant and realistic experience (i.e. its smoother and makes you do a little mental work on hitting the target) and makes it less prone to death spiral/lock up issues.
I usually do fine on work intervals but run into lock up issues most often on TR on recovery intervals if I back off too much or coast for a bit. Maybe TR could just tone erg mode down a bit on the easy stuff and still keep its hard ass erg mode on the hard stuff
This doesn’t seem right. There is no way of knowing you are backpedalling. It only knows there is cadence without any power. (or if using a trainer like the kickr 2018 for cadence won’t pick up any cadence)
I agree with what STP said and think the behavior should be (though not completely sure what the numbers should be, some tweaking may be needed):
if Cadence <minCadence and actualPower<targetPower*0.5
switch out of erg mode to level mode
give an auditory and visual signal to the user (beep every 5 seconds and the color behind the cadence value switches to red until erg mode switches back on) Or if doing a ramp test, end the test
Wait till Cadence > minCadence * 1.3 (maybe overshoot cadence a bit to make sure user is recovered) and actualPower>targetPower*0.8 and then switch back to erg mode
Correct, it isn’t really about “backpedaling” specifically. It IS about no power input for a period of time (5-10 seconds is commonly recommended). You could simply stop pedaling instead, wait the same amount of time, and get the same results.
The backpedal is an attempt to keep the body working as you deal with the death spiral and stay at least a bit active while waiting for the trainer to reset resistance down.
I highly doubt there’s a difference between apps as far as “wandering a bit from the target” is concerned. The ANT+ FE-C protocol and its BT equivalent, FTMS, provide for a power target being transmitted by the app to the trainer, and the trainer itself is responsible for controlling resistance vs power deviations. The only difference between apps is power-matching functions, which act as a second loop controlled by the app itself, trying to bias the power target to meet a secondary power data source. Unless you are using a power-match function, there is no difference in the erg mode power control loop between apps on a given trainer.
The backpedal will give zero power, but non-zero cadence, thus preventing TR from pausing the workout. Cadence sensors are incapable of telling which way you are spinning, thus backpedaling will provide a valid cadence signal.
If all the app is doing is telling the trainer what wattage to hit in erg mode then the apps should be the same. Zwift doesn’t do that and some us want TR to not do that either. You can detect the start of the spiral of death, stop it and allow the rider to get back to what the target is. All using the existing functionality the trainer provides: https://www.thisisant.com/developer/resources/downloads/#documents_tab
There’s a distinct difference between an app including code to recover from a large divergence in power vs target (e.g. “spiral of death recovery”), and an app that presumably “lets you wander a bit from the target and is more gentle in forcing you back”. The former can be done by changing the mode and/or target power, the latter can’t be done - you can’t change the power control loop unless you get out of the loop or change the command.