how the mission control assist settings work (creo SL, maybe all the SL bikes)

[etgo]

Thanks Guys..

 

[paul2.0]

there has been a lot of confusion about this, rightfully so since specialized's language is not clear. the current language around the SL motor on the creo states "the smooth 240 watt SL 1.1. motor doubles your effort..." and then states "rider amplification: 2x you."

the mission control documentation says "The Support level defines the proportion of motor assist against your own pedaling input," and about peak power, says "This feature refers to the power output by the motor for each Support Mode. ... If Peak Power is set at less than 100%, you'll limit the amount of power that the motor delivers and this will create an artificial ceiling that cannot be compensated for with leg power."

i originally believed (many posts here and elsewhere state!) that the support level was a simple percentage of your output - rider does 300w, support is at 50%, motor does 150w. as pointed out by @Stefan Mikes, not so. in fact, it's 50% of the maximum "2x" amplification factor for the motor. if the rider was doing 100w, and the support level was set to 100%, the motor would add 100% of 2X the rider power (aka 2x), which would mean a motor OUTPUT of 200w. the motor is not 100% efficient, so the input would be around 240w. if the rider was doing 200w, and the support level was set to 40%, the motor would add 40% of the 2x of the rider power (aka .8x), which would mean a motor output of 160w, and a motor input of approximately 200w.

the maximum motor output is 240w, and the maximum motor input (the figure displayed by mission control) is 300w. specialized has stated that for the SL motors, the peak and sustained power are basically the same.

i experimented today with a wide range of mission control settings to confirm this, and it's pretty much correct, although there's one big caveat - the motor power is surely in large part determined by the torque sensor, and we really have no way of knowing torque without backing out cadence and power, introducing a lot of error into the equation. we're making our comparison based on the creo's firmware having already converted the torque * cadence value into a power estimate, and there definitely seems to be a bit of lag, a bit of smoothing, and probably all kinds of other mathematical wizardry going on. the data i got (1 second sampling of rider and motor power) is a bit spiky, and there are some clear anomalies introduced by rapid changes of cadence, speed, etc. nonetheless, it pretty much works the way i've described above.

this chart shows motor power in blue, rider power in red, and a smoothed "support percentage" value in orange. the mission control setting is above. for this experiment, i set the max at 100% so the only determinant here is how much work the rider is doing, up to the motor's maximum power. at a mission control value of 10% assist, the actual motor output (assuming the 80% efficiency we mention above!) is 20.46%. it's fairly smooth, but at this low setting not particularly reactive to big increases in rider power, as you can see where i pedaled up to 300w and the assist... stayed the same. at the mission control value of 20% assist, the average support level is 43%. i pedal with 250w, motor outputs 108w. you can see things are drifting up a bit - at 25% assist, the average actual output is almost 55%, and at 30% it's 65.5%. this pretty clearly proves that the mission control "support" value is not the percentage of 100% rider power, it's actually the percentage of 200% rider power (2x you... but really, it's 3x you ) incidentally, this is all confirmed by a specialized rep directly at emtbforums - https://www.emtbforums.com/communit...r-in-mission-control-app-v2-0.4038/post-56053

View attachment 98133
having pretty much found the assist percentage behaving as expected, i did another experiment, the other way around. 100% assist (so it's going to always try and be "2x me" plus the original me) but stepping the max support from 10 to 90 in 10% increments. this behaves exactly as expected, with some spikes in the data and the big caveat that it seems around 10% ahead of itself - at 40% max power, it's actually at 50%, at 90, it's pretty much at 100, etc.

gold line is percent of maximum possible motor output (240w, 300w battery draw), thick blue line is battery draw in watts. dashed red line is rider power, not very relevant here since we're at 100% assist level and i'm basically always pedaling hard enough to make the determining factor the "max" power.

View attachment 98134

there is one final thing i need to explore a bit. when i set eco to 10% assist and 10% max, i got ... zero motor output regardless of speed or rider power. not sure what that's about. will have to experiment a bit more.

Thanks so much for the really excellent post! I learned a ton from both you, @mschwett, and also @Stefan Mikes! I have a 2021 Specialized Turbo Levo SL with a 1.1 motor (https://support.specialized.com/turbo/en/turbo-super-light-system-sl/sl-motor). I'm hoping someone in this thread can help me with a problem I'm having in Turbo mode.
The specs on that official page do reflect the 2xYOU marketing language, but they also verify what Stefan said that the actual "rider amplification" number is 180% of you or 1.8xYOU. That also gets multiplied by the SUPPORT% in the presets, so the real formula is MotorPower = 1.8*RiderPower*SUPPORT%. To verify this for myself, I exported the FIT file from one of my rides into Excel. The screenshot below shows some graph from the data. The top one shows the SUPPORT% in green (Eco=30%,Trail=35%,Turbo=40%), 240W*PEAK% in light blue (Eco=80%, Trail=70% by mistake, Turbo=80%), the MotorPower in dark blue, and MotorExpected=1.8*RiderPower*SUPPORT% in purple. You can see that, in general, the dark blue MotorPower and purple MotorExpected match up fairly well except when you see the green SUPPORT% number go to their highest value of 40% in Turbo mode. Then all bets are off and the motor seems to behave as if SUPPORT/PEAK%=100%/100% even though I verified in the FIT file that the values are 40%/80% in Turbo mode. To make this even more clear, in the bottom graph I used a moving average (averages 10 samples) to smooth out RiderPower and MotorPower and then plotted SupportActual = MotorPower/(1.8*RiderPower) in red as well as the programmed SUPPORT% in blue. I excluded all data points where Cadence<70, or Speed>20mph, or RiderPower<10W so that's why you see the red line drop to zero periodically, but you can still see the red SupportActual matches the programmed blue SUPPORT% fairly well except when it goes into Turbo mode in which case it just ignores the 40% support setting and behaves as if it's programmed to 100%.
The bike has the latest firmware update (updated Feb 21, 2023) and I guess I'm curious whether anyone has seen this crazy behavior where in Turbo mode it doesn't matter what you program SUPPORT/PEAK% to and the bike just ignores it and uses 100/100%? I've verified this on several sets of ride data both before and after this last firmware update. Any thoughts?

 

[Stefan Mikes]

Too cryptic for me too understand Paul

Take into account the MOTOR POWER is electrical, that is, Mechanical Motor Power = Electrical MOTOR POWER * 0.792. In the Electrical Motor Power terms it is "2x you". In the Mechanical Power Terms, it is "1.8x you". Does it change your reasoning?

 

[paul2.0]

Too cryptic for me too understand Paul

Take into account the MOTOR POWER is electrical, that is, Mechanical Motor Power = Electrical MOTOR POWER * 0.792. In the Electrical Motor Power terms it is "2x you". In the Mechanical Power Terms, it is "1.8x you". Does it change your reasoning?

Sorry for the cryptic nature of the post LOL. Yeah, I was aware of the motor power being electrical power drawn from battery and that the efficiency to convert to mechanical output power is about 79%. The only difference that would make is some small change on the order of 20% regardless of how you account for it. What I'm talking about is that in Turbo mode, the motor power drawn from the battery goes to on the order of 250W-280W even though supposedly SUPPORT=40% and RiderPower is on the order of 125W. For a RiderPower=125W with SUPPORT=40%, I would expect on the order of MotorExpected=1.8*125W*40%=90W not 250W! Even if you want me to throw in the efficiency conversion factor it only knocks the MotorPower down from 250W to 0.79*250=198W, which is no where close to the 90W it should be. You can really just focus on the bottom graph and look for places where the red and blue deviate, which is in Turbo mode.

 

[Stefan Mikes]

Your reasoning is correct. The power draw from the battery should be some 90/0.792 = 114 W.
However, there is something like motor power curve that depends on the manufacturer's motor tuning. That is, the motor delivers more assistance at low speed. I would take the "1.8x" or "2x" with a grain of salt.


Now, let me look to it from the practical angle. I was on a 50 km ride yesterday with my Vado SL. I set the assistance to 50% and the max motor power to 50%. (The speed limiter was at 25 km/h). The e-bike used 173 Wh. I was pedalling for 2.7125 hours. The average electrical power was thus 63.77 W. Strava tells me (based on the Vado SL data) that my average leg power was 79 W.

Now: 0.5 (assistance) * 79 W * 2.0 (electrical max assistance) = 79 W (theory). Actual electrical assistance was around 64 W though? Yes, that was the Speed Limiter that made the e-bike consume less energy! Unrestricted, the average consumption is close to theory.

 

[paul2.0]

Your reasoning is correct. The power draw from the battery should be some 90/0.792 = 114 W.
However, there is something like motor power curve that depends on the manufacturer's motor tuning. That is, the motor delivers more assistance at low speed. I would take the "1.8x" or "2x" with a grain of salt.


Now, let me look to it from the practical angle. I was on a 50 km ride yesterday with my Vado SL. I set the assistance to 50% and the max motor power to 50%. (The speed limiter was at 25 km/h). The e-bike used 173 Wh. I was pedalling for 2.7125 hours. The average electrical power was thus 63.77 W. Strava tells me (based on the Vado SL data) that my average leg power was 79 W.

Now: 0.5 (assistance) * 79 W * 2.0 (electrical max assistance) = 79 W (theory). Actual electrical assistance was around 64 W though? Yes, that was the Speed Limiter that made the e-bike consume less energy! Unrestricted, the average consumption is close to theory.

Yeah, I totally agree on the speed limiter. That's why I mention in my original post that I actually excluded all the data points where speed>20mph because the limiter kicks in at that point. I also excluded all data points where cadence was <70 because the SL1.1 motor below 70rpm cadence the power begins to drop off and it's quite flat and pretty above 70rpm. I also exclude data for RiderPower <10W because dividing by small numbers leads to large error sensitivity.

What I can't figure out is why no matter what I set support to in Turbo mode, it just ignores it and roughly speaking tries to double my rider power LOL. I do have a whole separate spreadsheet where I've done similar calculations on total energy consumed and average rider power and average motor power and I agree that if you don't account for "coasting" and motor power vs cadence and speed limiting that things don't make sense, but like I say I exclude all those cases and just look at the "ideal" portions of the ride where the 1.8xSUPPORT% or 2xSUPPORT% (as you prefer) should apply. Works fine in Eco mode and Trail mode, but not in Turbo where it just ignores the SUPPORT% setting. I'm not really sure where to turn to next. Explaining this data to a rider care specialist seems hopeless. I think I read you're a software engineer so these types of discussions make sense to you. I myself am an electrical engineer so I can also clearly see something is wrong in Turbo mode, but I'm not sure that it's going to make sense to anyone in Specialized RIder Care support. I wouldn't even know how to begin to explain it to them.

 

[Stefan Mikes]

Works fine in Eco mode and Trail mode, but not in Turbo where it just ignores the SUPPORT% setting.

That is very strange! Set 20% assistance but also 20% Max Motor Power for Turbo and tell me

I have never changed the settings on SL Turbo but I did if for a full-power Vado and it worked. Fancy my frustration. Turbo set to 70/70%, me without the smartphone, an emergency situation, and I had to pedal fast on the reduced Turbo when I needed 100% of assistance...

 

[Calcoaster]

Paul, I’m glad to see your post here because I have similar experience with my Creo that I have confirmed many times with both the Blevo app and the specialized rider app. My eco and sport mode do not use the 1.8X factor, it’s more like a 1.2X multiplication factor which makes the mechanical amplification roughly equal to the support setting. But turbo mode gives me almost full power no matter what I set it to.

I used Blevo to log this, with eco set to 35 support, sport 36 and turbo 37, all with the same max power setting of 50. (Blevo won’t allow the same support settings so I added 1 to each). Eco and sport give me roughly 40-45% motor power but turbo at least doubles my power. I tested this by holding my rider power steady at various levels (100, 140 and 200 watts), in each mode. Then I used the Blevo app to look at the instantaneous rider/motor power levels.

I reported this to rider care and they said to get latest updates and have a motor calibration. I’ve done that twice now and nothing has changed.

My motor works fine and I like riding with the lower level of support. I could just change my eco mode support level, for instance, from 35 to 60 or so to give me the same motor power as other creos get at 35, but I like to work harder. On my Creo, 35 support is like about 22 on others that conform to the 1.8X condition. And turbo basically gives me 300 w (electrical) no matter what it’s set to as long as I’m pedaling at 120 W or more.

So on my bike, turbo mode does NOT behave like eco or sport mode. If you get any clarification please let me know. I rarely use turbo mode and my unique eco and sport settings are actually more intuitive than the standard 1.8X algorithm. They’re basically 1X mechanical with a little more added to give the electrical power needed for that. And Paul, like you, I have an electronics background too so I’m sure my analysis is correct, at least for my bike.

 

[Stefan Mikes]

@Calcoaster, thank you so much for your input! When two confident people say the same, then there is no doubt you are onto something!

 

[paul2.0]

Paul, I’m glad to see your post here because I have similar experience with my Creo that I have confirmed many times with both the Blevo app and the specialized rider app. My eco and sport mode do not use the 1.8X factor, it’s more like a 1.2X multiplication factor which makes the mechanical amplification roughly equal to the support setting. But turbo mode gives me almost full power no matter what I set it to.

I used Blevo to log this, with eco set to 35 support, sport 36 and turbo 37, all with the same max power setting of 50. (Blevo won’t allow the same support settings so I added 1 to each). Eco and sport give me roughly 40-45% motor power but turbo at least doubles my power. I tested this by holding my rider power steady at various levels (100, 140 and 200 watts), in each mode. Then I used the Blevo app to look at the instantaneous rider/motor power levels.

I reported this to rider care and they said to get latest updates and have a motor calibration. I’ve done that twice now and nothing has changed.

My motor works fine and I like riding with the lower level of support. I could just change my eco mode support level, for instance, from 35 to 60 or so to give me the same motor power as other creos get at 35, but I like to work harder. On my Creo, 35 support is like about 22 on others that conform to the 1.8X condition. And turbo basically gives me 300 w (electrical) no matter what it’s set to as long as I’m pedaling at 120 W or more.

So on my bike, turbo mode does NOT behave like eco or sport mode. If you get any clarification please let me know. I rarely use turbo mode and my unique eco and sport settings are actually more intuitive than the standard 1.8X algorithm. They’re basically 1X mechanical with a little more added to give the electrical power needed for that. And Paul, like you, I have an electronics background too so I’m sure my analysis is correct, at least for my bike.

Well, at least we both know now that we’re not crazy! I’ll keep you posted and please do the same. A bit frustrating to have such basic functionality problems on what for me was an almost $10k bike!

I didn’t mention it in the original post to avoid complicating it even further, but FYI Smart Control distance-mode does not work for me either. You can watch how it sets the Support/Peak% in the FIT file as the ride progresses, but when you graph the effective support, it’s about half of what it claims to be. In other words if it says it’s 30% support at a given moment in smart mode, it will behave the same as it would if you programmed support as 12-ish%.

 

[Calcoaster]

@Calcoaster, thank you so much for your input! When two confident people say the same, then there is no doubt you are onto something!

@Calcoaster, thank you so much for your input! When two confident people say the same, then there is no doubt you are onto something!

Thanks Stefan. And before I got my Creo I had a Vado sl for six months that I had tested the same way after reading your ( and mschwett’s) detailed analysis. That bike did perform like you say (roughly 1.8 X rider input + 20% for the motor electrical energy conversion). I never bothered to test that bike in turbo mode since I assumed it would behave the same way. So it’s easy for me to see my Creo does not follow the same pattern.

My Creo is a 2021 and my Vado sl was a 2020 so I wonder if there is a hardware difference that accounts for it. Or maybe my Creo is unique and just likes 1X in eco and sport and ‘gimme all you got’ in turbo.

Oh, and I once rode a friend’s Creo (2020 model), connected it to my Blevo app, and it behaved like your formula. I imagine most people don’t care so much about these details so they don’t know or report whether theirs boosts at 1X or 1.8X, or if turbo has a mind of its own.

In the meantime, it really doesn’t matter to me, my Creo is a fantastic bike that gives all the support I need, just with a different multiplication factor. Thanks for all your informative posts.

 

[paul2.0]

Thanks Stefan. And before I got my Creo I had a Vado sl for six months that I had tested the same way after reading your ( and mschwett’s) detailed analysis. That bike did perform like you say (roughly 1.8 X rider input + 20% for the motor electrical energy conversion). I never bothered to test that bike in turbo mode since I assumed it would behave the same way. So it’s easy for me to see my Creo does not follow the same pattern.

My Creo is a 2021 and my Vado sl was a 2020 so I wonder if there is a hardware difference that accounts for it. Or maybe my Creo is unique and just likes 1X in eco and sport and ‘gimme all you got’ in turbo.

Oh, and I once rode a friend’s Creo (2020 model), connected it to my Blevo app, and it behaved like your formula. I imagine most people don’t care so much about these details so they don’t know or report whether theirs boosts at 1X or 1.8X, or if turbo has a mind of its own.

In the meantime, it really doesn’t matter to me, my Creo is a fantastic bike that gives all the support I need, just with a different multiplication factor. Thanks for all your informative posts.

It’s potentially concerning that different bikes of the same model might have a different multiplier though I can see from your comments that it could also be a difference in model year since we don’t have data showing side-by-side comparison for same model and model year. Firmware version is another variable. Ugh…

 

[Calcoaster]

Well, at least we both know now that we’re not crazy! I’ll keep you posted and please do the same. A bit frustrating to have such basic functionality problems on what for me was an almost $10k bike!

I didn’t mention it in the original post to avoid complicating it even further, but FYI Smart Control distance-mode does not work for me either. You can watch how it sets the Support/Peak% in the FIT file as the ride progresses, but when you graph the effective support, it’s about half of what it claims to be. In other words if it says it’s 30% support at a given moment in smart mode, it will behave the same as it would if you programmed support as 12-ish%.

Yeah, I’ll keep you posted if I ever find a solution or explanation. I’ve never used Smart Control so I don’t know if mine would act the same.

I don’t get the feeling that the rider care techs or my lbs techs have any idea what I’m talking about. I even sent rider care a Blevo csv file that shows second by second rider and motor power numbers (demonstrating the 1X and not the 1.8X behavior, and the full-on turbo contribution). The response was, “everything looks good on our end but we show you’re due for an update and motor calibration.” So I got that again and no change. I guess if I get support that I can set to whatever I want and just use turbo for instant full power, I’m ok with that. It still gives me an incredibly smooth power contribution, just with different settings than others use.

 

[mschwett]

@paul2.0 firstly, i would pretty much guarantee your LBS and rider care have no idea what you’re talking about

when i did the test way back in which i gradually increased assistance by 10% steps to determine what the true ratios were (on my bike they were slightly more than 2x, as noted in the initial post) i believe i did so in the middle “preset.” at the time i asked around if the presets actually changed anything else besides the two percentages, and was told no. in other words, simply by using different values you could turn eco into turbo or vice versa. i’m actually not so sure about this anymore, and i distinctly recall that low settings on turbo sometimes produced no assist at all.

i’d guess that either 1) your custom setting for turbo isn’t sticking or 2) turbo actually has some preset behavior which overrides the custom % to make sure people feel like it’s “turbo” enough.

also, i assume in your description you meant that eco was effectively 60, not 80?

would you mind sharing your xls?

 

[paul2.0]

@paul2.0 firstly, i would pretty much guarantee your LBS and rider care have no idea what you’re talking about

when i did the test way back in which i gradually increased assistance by 10% steps to determine what the true ratios were (on my bike they were slightly more than 2x, as noted in the initial post) i believe i did so in the middle “preset.” at the time i asked around if the presets actually changed anything else besides the two percentages, and was told no. in other words, simply by using different values you could turn eco into turbo or vice versa. i’m actually not so sure about this anymore, and i distinctly recall that low settings on turbo sometimes produced no assist at all.

i’d guess that either 1) your custom setting for turbo isn’t sticking or 2) turbo actually has some preset behavior which overrides the custom % to make sure people feel like it’s “turbo” enough.

also, i assume in your description you meant that eco was effectively 60, not 80?

would you mind sharing your xls?

Thanks for the comments! Interesting thought that Turbo might override presets, which is what @Calcoaster and I seem to be experiencing — though if that were the case one would hope they would not let you program it in app lol. As far as not sticking, that occurred to me too but I can only say that the correct values appear in the FIT file and I’m assuming they are coming from the TCU and not the app. That could be wrong but the creator of BLEvo once claimed that neither BLEvo nor MC (which is what I use) store the values in the app itself. I’m happy to share the Excel sheet. I actually tried to attach it to the post but it wouldn’t take it cause wrong file type. Maybe you can tell me how to share it. As for the Eco mode Peak% being 80, that is actually what it was for that test. The Trail Peak% was only 70% by mistake. All three modes were intended to have 80% peak in that test. FIT file values match what’s in my post.

 

[Stefan Mikes]

So you guys have observed peculiar behaviour of the Turbo mode on some Creos. Technically, an interesting thing. Practically? With the low power SL motor and my weak legs, I never thought I should tune the Turbo mode. If I go for so-called "coffee ride" together with my gravel cycling club, I choose the Vado SL. I need 60, 80, and 100% SL presets to match my companions of both sexes on a such relaxed ride! The 60/60% is what I actually need to match "beginner girls" there The SL Turbo mode is less than 50% of the assistance of my Vado 6.0!

The things are really different if we go on a proper club ride. These boys and girls (often in their 40s, and sometimes 50s) are speed lunatics! I use 40, 55 and 100% presets on my Vado 6.0 there. As I own no Mastermind e-bike (and cannot use the MicroTune), I've fitted the O-Synce remote connected to BLEvo. Whenever I can see the group disappearing at the distance, I vigorously push the Up button there until I have got at the group speed. Or, when I see I could ride into the group, I press the Down button on the O-Synce to slow down to the actual group speed.

We did 104.5 km just under 4 hours. No other e-bike than my full power Vado would be appropriate for me there