Shifting gear under load

[lightning]

So, what's the take on changing up (or down) while under power on an ebike.

When accelerating away from standstill l will change up the block under power. ls this acceptable or a no-no?

The drivetrain seems to shift seamlessly when used in this way, but am l doing damage.

Changing down is less smooth and l assume not to be recommended unless it's absolutely necessary (for example, to avoid stalling and falling off)

 

[Gary]

So, what's the take on changing up (or down) while under power on an ebike.

DON'T. Same as with a normal bike. Unless of course you are a world class racer and sprinting pays your wage.

When accelerating away from standstill l will change up the block under power. ls this acceptable or a no-no?

Your bike, your money. If you find it acceptable. who can argue with you?

The drivetrain seems to shift seamlessly when used in this way, but am l doing damage.

Yes.

Changing down is less smooth and l assume not to be recommended unless it's absolutely necessary (for example, to avoid stalling and falling off)

Avoid stalling by planning ahead and chosing appropriate gearing while not under high load. As you have assistance modes to change to if required you can actually avoid changing gear under load far more easily than on a normal bike.
Learn to trackstand and you won't fall off anywhere near as often just from stalling

 

[lightning]

Okay, will do!

l realise that changing down under load is normally due to "wrong gear selection" and try to look or plan ahead so l can avoid it.

But l can't say l always get it right!

 

[steve_sordy]

I don't know how you can bring yourself to shift under load, I just can't do it. Not physically of course, but I am psychologically incapable of shifting under load. The realisation that I could shift up a mode at times of drivetrain stress was a revelation.

 

[lightning]

l have to say, that the seamless acceleration from a standing start to maximum speed is great, but probably at the expense of the cassette.
Having said that, despite this abuse, my cassette lasted 1500 miles and was only replaced due to wear on gears 9,10 and 11

 

[Tim1023]

I don't know how you can bring yourself to shift under load, I just can't do it. Not physically of course, but I am psychologically incapable of shifting under load. The realisation that I could shift up a mode at times of drivetrain stress was a revelation.

I think the term is "mechanical sympathy". It just hurts the soul when I hear my gears graunching like that.

 

[KenX]

Learn to trackstand and you won't fall off anywhere near as often just from stalling

Try track standing with a Yamaha motor ?

 

[Gary]

Completed it mate.

When I was 5

 

[steve_sordy]

I think the term is "mechanical sympathy". It just hurts the soul when I hear my gears graunching like that.

"Mechanical sympathy" is a phrase I have been using since my apprenticeship days 1969 -74. So "hurts the soul" resonates with me.

 

[lightning]

My gears don't "graunch" if l change up under load, they change seamlessly a bit like clutchless upshifting on a motorcycle.

lt's a testament to the design of the chain and cassette, that it can actually work so well when stressed in this way.

 

[salko]

l have to say, that the seamless acceleration from a standing start to maximum speed is great, but probably at the expense of the cassette.
Having said that, despite this abuse, my cassette lasted 1500 miles and was only replaced due to wear on gears 9,10 and 11

1500 miles is very low for a cassette. Mine 12spd chain should survive up to around 1500 miles until 0,5% wear and I normally replace 2-3 chains per cassette. I also have this thing - the Mechanical sympathy.

 

[lightning]

Well, l have just replaced the chain and cassette.
l will avoid gear changing under power, where possible, and see if l can improve on the 750 mile chain /1500 mile cassette life l got from the last one.

Although it was only 10 and 11 that gave trouble on my cassette, probably because those are the gears l mainly use when riding on the road.

 

[Gary]

Even if you use all sprockets evenly the smaller sprockets will still wear out first as the chain passes over each tooth many times more than a larger sprocket at the same cadence plus the reduced tooth count has less area to spread load (your 10t sprocket only ever has 5 teeth in contact with the chain at any time.). Because of this your chain will also slip under load far sooner on the smaller sprockets due to chain wear.
Running a 10t sprocket on an Ebike is a really really stupid idea

 

[lightning]

l never use the 10. l might change the front chainring for a larger one, although it's already a 24 l think

 

[Gary]

Although it was only 10 and 11 that gave trouble on my cassette, probably because those are the gears l mainly use when riding on the road.

l never use the 10.

Lapse in medication today?

?

 

[KenX]

Even if you use all sprockets evenly the smaller sprockets will still wear out first as the chain passes over each tooth many times more than a larger sprocket at the same cadence plus the reduced tooth count has less area to spread load (your 10t sprocket only ever has 5 teeth in contact with the chain at any time.). Because of this your chain will also slip under load far sooner on the smaller sprockets due to chain wear.
Running a 10t sprocket on an Ebike is a really really stupid idea

I try to avoid running the 11t on my 11 speed if poss for this reason........

 

[Gary]

You don't have to avoid it.
Just choose appropriate gearing for the riding you do.
Hint: larger chainring and closer ratio cassette if you find yourself in the high gears all the time.

 

[lightning]

Lapse in medication today?

?

Ah yes, l meant l never use the 10 TOOTH sprocket, gear no. 12 !!

Although "never"' is not really true either, it's occasionally useful on road descending.

 

[maker]

Bosch advise easing back on power when changing gear. Read when chain is momentarily off alignment as it transitions cogs the power of the motor can break the chain

 

[jbv]

depends. Shimano 12 sp loves shifting under power.

 

[Zimmerframe]

depends. Shimano 12 sp loves shifting under power.

It does ? I wince every time I've not eased off enough or miss timed .. I was riding home the other day on a flat battery and it was sooooo smooth ..

 

[steve_sordy]

depends. Shimano 12 sp loves shifting under power.

Mine doesn't. It is the noisiest shifting on a bike I've had for many years. It's a proper clang when dropping down from big gears to smaller ones. I can do it quietly but I have to be extra specially sympathetic, which even the best of us can't always manage.

 

[jbv]

It does ? I wince every time I've not eased off enough or miss timed .. I was riding home the other day on a flat battery and it was sooooo smooth ..

yes, it's a design feature. personally i quite like it and make use of it, but many people don't really value it because they are so self trained to never shift under load, as it's just not a good idea. when i got my latest bike it came with Sh 12sp and i tried it while sitting and spinning and it was great, then once i got caught out and had to downshift while standing up full gas and no problem. on my previous 11 sp it might have sheared or bent teeth on the cassette. do a google search and you will see it mentioned in all the reviews. the eeb i've got on order comes with SLX 12 and i'm stoked. if only the bike were to ever arrive, 10 months and waiting....

 

[salko]

Just because it shifts silently and nicely under load it doesn't mean it will not eat up the drivetrain way prematurely compared to if you shift with mechanical sympathy ... Keep in mind that manufacturers are still the ones which want to sell you as many parts as possible ...

 

[Bearing Man]

I have to agree with Gary on this, "Your bike, your money. If you find it acceptable. who can argue with you?"

But, lets take a look at what's happening when you shift under load on an ebike:
First you have what ever power you can generate and an extra 350 to 400% coming from the motor. Bear in mind you are still running a standard pushbike sprocket, cassette and derailleur, even if you have an "E" specific chain.

Apart from the obvious chain stretch caused by high side loads and shock damage, the main damage we see is:
Bent and snapped teeth on the rear cassette
Cracked cassette freewheel ratchets
Hammered and or destroyed ratchet pawls

We also see possible issues due to the shock loading through the motor:
Sprag bearing failure
Belt failure
Excessive backlash on some gearing
Plastic gears stripping

I tend to stick with the mechanical sympathy method, but... "Your bike, your money."

 

[Waynemarlow]

As much as I would want to agree that changing under load is wrong, if you think it through on an engineering basis in regards the cassette ( the expensive part ), as long as there are multiple teeth engaged then it’s almost irrelevant if you are in the process of changing gear. The loading will be spread evenly over X number of engaged teeth regardless if for example they are spread over some on gear 7 and some on gear 8. Good quality mechs will have ramps and the likes to correctly engage the teeth and other than a very very minimal extra side contact, the direct loading will be well within limits.

The chain on the other hand ( the cheap component ) has to take all the side movement and angle correction of the ramps on the mech and will be working not in the intended manner of a point to point in perfect alignment chain, hence I change my chain at 1/2 worn and get 3 chains to every cassette. It’s noticeable also in the mud of the U.K. winters just how much quicker the chains wear, 2 in 3 months of winter to just one for the other 9 months. Let the chains go beyond 1/2 and you are into changing a cassette for every chain ( we’ve all had the chain skipping over the teeth sound on a new chain fitment ).

I used to be of the opinion that backing off was right, but the motors we all use do have some run on after we stop pedalling to change and my betting is that we do not back off long enough to change, than the run on ( think how long it takes to completely power down once you stop pedalling ). Now most motors if re engaged whilst the pedals are being rotated, will ramp up faster and with more power to quicken the process of matching the cadence and thus getting smooth power to the wheels. We notice this as the nice power surge as we start to pedal. That rapid power surge is probably putting more power / loading through the gear set than a constant steady power, whilst changing the gears.

So without scientific proof I have to ask, are we doing more damage by the motor kicking in with a fast ramp up to match the cadence, every time we back off momentarily to change gear ? Bearing Mans appraisal of seeing shock loading damage such as sprag clutches, belts gears etc may well support this.

 

[Gary]

No.

When you back off pedalling you should then be timing your gear shift to correspond with the motor over run so that the chainring is still spinning but pulling the chain at a greatly reduced load. Ie. Almost no load. Same as a smooth shift with no motor.

If you ride at a decent cadence (unlike a shit load of Ebikers) the motor cutting and pick up will also be far less severe.

Different motor (and indeed modes) operate differently so that also needs taken into consideration.

 

[Gary]

Beginning your thoughts with

if you think it through on an engineering basis

I actually can't believe you'd think higher load when changing gear won't increase wear to cassette sprockets.
Basic physics and engineering principles apply..

 

[Waynemarlow]

Beginning your thoughts with
I actually can't believe you'd think higher load when changing gear won't increase wear to cassette sprockets.
Basic physics and engineering principles apply..

I think with that statement you are agreeing with my hypothysis, without actually understanding why. I am stating that potentially if you back off to change, you will increase the wear rate of the gear teeth on Ebikes, unlike analogue bikes.

Before I start to explain more, what do you consider " load " to be ? Is that sideways load, individual point load on the teeth or a combination of both ? I think you need to detail what sort of load you are thinking of when you make such an all encompassing statement like that.

One would have to consider also the spread of load over a number of teeth as load in my book is simply a force applied to a single point ( in our case pressure on the teeths surface ), if that pressure is spread over 7 teeth rather than 4 then obviously the " load " will be of different values. Oh so complicated.

I think you are missing my point which I am stating that if you back off the human pedaling load momentarily ( cadence or the torque sensor reading begins to lower ), the engine controller will begin to shut down its power and probably the speed of the motor. You the human controller then begin to rotate the pedals again at X cadence and with some force Y on the pedals. The motor is still in its shut down phase ( inertia and magnetic forces can't be instantly controlled ) slowing rotation. At this point the sensors will report that the cadence has increased and will demand that the motor speeds up to match the cadence and inputs Y amount of current to match the requirements of the torque sensor.

Now for those who have worked in the electronics field will know, algorythms are used to control rotation and power in motors. Now dependant on firmware, the motor will apply too much voltage to first ramp up the speed as fast as it can to match the cadence, then apply too much power to find the level of the torque sensor requirements. The motor firmware then shuts off and reapplys both voltage and power ( amps ) in short bursts until an equilibrium occurs which matchs the riders requirements of pedal cadence and speed. Thats just the way its done. For those who want a bit more knowledge heres a primer Field Oriented Control (FOC) - A Deep Dive

Its this initial burst of power and voltage that I think is wearing the gears more than if a cyclist simply changes gear in a smooth manner ( the motor power won't change ). Now I have to admit a bit of extra information. I sometimes log my EBikes power and there are definate spikes in that recording which are way beyond the actual requirements of constant pedalling. Due to the very slow logging of data I have ( 1 second average at best rate ) its almost impossible to actually say that coincides with changing gear, but the question has been asked, are those spikes coinciding with the backoff I do when changing gear and when the motor repowers back up as I begin to repedal ( motor extra power + human power ) ? If so then the motor spikes will cause greater wear ( increased point loading, thats the little surge you can feel just after reapplying pedal power ) than say an analogue bike when backing off to change.

 

[Gary]

No. Mate. I'm not agreeing with you at all. I'm suggesting you educate yourself in basic engineering principles before talking absolute bobbins