Torque simulation controllers do not control torque

[avandalen]

Torque simulation controllers do not control torque​

In recent times, there has been a surge in the marketing of motor controllers as torque simulation controllers. This led me to assume that these controllers control the motor torque rather than its speed. However, this turned out to be false, I discovered that these controllers merely regulate the motor voltage, just like regular motor controllers.
See my full article here:

Torque simulation

Here, I share technical solutions that arised during the development of my solar-bike.

www.avdweb.nl

 

[amberwolf]

Thanks for the article; yes this has been discussed here over the years to correct various people's misconceptions because of their incorrect marketing terms.

 

[stancecoke]

However, this turned out to be false

Interesting finding, but I think the control strategy is much simpler. The Kunteng normally just controls the battery current to the value, that is assigned to the recent assist level, with a cadence dependent ramp from standstill. The mechanical output power characteristic you are showing, is simply caused by the effiency vs. speed behaviour of any BLDC motor...

The throttle behaviour can be set in the menu of the display

 

[Comrade]

mechanical output power characteristic you are showing, is simply caused by the effiency vs. speed behaviour of any BLDC motor

And that's why they are even called torque simulation controller, not torque controller.

 

[Blacklite]

However, this turned out to be false, I discovered that these controllers merely regulate the motor voltage, just like regular motor controllers.
See my full article here:

Torque simulation

Here, I share technical solutions that arised during the development of my solar-bike.

www.avdweb.nl

Nowhere does your article show that these controllers are just controlling voltage.
You are showing a relationship between cadence and power.
Do you have further data showing the control paradigm is voltage? I would assume it is actually current instead, and torque is directly proportional to current.

 

[AHicks]

There's speed based PAS, and there's power based PAS. KT's "imitation torque control" is chinglish for power based PAS. It's that simple.....

 

[richj8990]

There's speed based PAS, and there's power based PAS. KT's "imitation torque control" is chinglish for power based PAS. It's that simple.....

It affects throttle response too. That's why I'm going to turn it off later after I put the battery back on for this new build in the garage.

Instead of these replies reminding everyone that imitation torque is not torque, maybe someone (mua) should describe what really happens out on the road or trail with imitation torque. Like, you know, real-world application experience?

I like speed control because you know what you are getting, and the power application is consistent. With imitation torque, the controller senses that the speed is slowing down even though the power is constant, therefore the controller quickly ramps up the power to compensate for the slower speed. So if you go from a flat surface to an incline, imitation torque will ramp up the power to compensate for the incline. Sounds nice in theory, and yes it can be nice in reality. Sometimes it ramps up the power too much and the wheel will spin. The thing I find annoying about it though (and this may or may not be helped by reprogramming the controller) is that when you are on a level surface or going slightly downhill...not enough where you can gravity coast, but enough that you'd want to max out your gears and have just a little bit of e-power to cruise at 20-25 mph, suddenly the throttle is unresponsive. Because it's doing the opposite of what it does on an incline: it's cutting back (all or almost all) of the power because it senses additional speed w/o any extra wattage. I don't like that. Some imitation torque functions are smoother than others, but speed control is consistent and I like that better.

 

[AHicks]

It affects throttle response too. That's why I'm going to turn it off later after I put the battery back on for this new build in the garage.

Instead of these replies reminding everyone that imitation torque is not torque, maybe someone (mua) should describe what really happens out on the road or trail with imitation torque. Like, you know, real-world application experience?

I like speed control because you know what you are getting, and the power application is consistent. With imitation torque, the controller senses that the speed is slowing down even though the power is constant, therefore the controller quickly ramps up the power to compensate for the slower speed. So if you go from a flat surface to an incline, imitation torque will ramp up the power to compensate for the incline. Sounds nice in theory, and yes it can be nice in reality. Sometimes it ramps up the power too much and the wheel will spin. The thing I find annoying about it though (and this may or may not be helped by reprogramming the controller) is that when you are on a level surface or going slightly downhill...not enough where you can gravity coast, but enough that you'd want to max out your gears and have just a little bit of e-power to cruise at 20-25 mph, suddenly the throttle is unresponsive. Because it's doing the opposite of what it does on an incline: it's cutting back (all or almost all) of the power because it senses additional speed w/o any extra wattage. I don't like that. Some imitation torque functions are smoother than others, but speed control is consistent and I like that better.

Between the wife and I, we currently have 6 bikes, and I ride every day. 4 of these bikes are set up with KT controllers, and all have the "imitation torque control" enabled. The other 2 bikes use Bafang mid drives, one of which is UART based (an Ultra), where they give you the choice or power or speed based. Based on my experience with KT (8 bikes dating back to 2017), and after trying it both ways, THAT bike is set to speed based.

None of my KT based bikes, dating back to 2017, will add uncommanded power in ANY mode, let alone when "imitation torque control" is enabled. The KT's do have cruise control, but even that will shut down/cancel if it slows more than a couple of miles an hour.

"Sometimes it ramps up the power too much and the wheel will spin."

This is getting pretty creative based on my experience. One of my bikes is powered by a 1000w MAC 12t, one of the most powerful geared hubs available. That bike has wheelied, but I've NEVER experienced wheel spin.....

"The thing I find annoying about it though (and this may or may not be helped by reprogramming the controller) is that when you are on a level surface or going slightly downhill...not enough where you can gravity coast, but enough that you'd want to max out your gears and have just a little bit of e-power to cruise at 20-25 mph, suddenly the throttle is unresponsive. Because it's doing the opposite of what it does on an incline: it's cutting back (all or almost all) of the power because it senses additional speed w/o any extra wattage.

With the speed limit set to 99 mph, down hill runs on any of my KT bikes will be WAY faster than MOST will ever want to go. Even on the flat and level, the MAC 12t powered bike will run well over 30mph. Down hill, I once saw something close to 40, but I'm too old to enjoy anything like that any more.

Point being, they do not shut the motor down.

If your experience contradicts any of my results, I'd suggest you might want to try some different parameter settings..... like I did when figuring all that out......

 

[richj8990]

The mid-drive LCD5 display will not show watts for the mid-drive, but for the hub drive, the LCD when under imitation torque would sometimes go all the way down to 11 watts when going downhill. When climbing a small incline, it would then ramp up to 600+ watts. All in the same PAS gear, usually 2 or 3. That's not exactly consistent. It would be one thing if the standard flat-surface watts were set at 250; downhill went down to 50-100 and uphill went up to 400-500. But it's way more extreme than that. The SW900 does a somewhat better job at imitation torque; it still ramps up watts almost too much when going up hill but the downhill watts stay 'there' at 50-100, at least useable.

I'll try imitation torque again throttle only and see how it does with the mid-drive. Really missing PAS up those switchbacks. The throttle simply is not smooth enough for those exact uphill sections where you need the bike's power to behave a certain way or it will not work. I'll live with it for now. Will try PAS again with speed control after rearranging the drivetrain. Drivetrain really, really needs upgrading. You would not believe how messed up the kit was.

 

[AHicks]

What mid drive do you have? I'm not familiar with any of them that are compatible with a KT controller. Same story with the SW900 display? I'm not familiar with an SW900 that will work with a KT controller. KT controllers require a KT display......

Torque sensing bikes WILL add/subtract power depending on load and PAS level. PAS level sort of sets the upper limit available through the torque sensing. Lower PAS level = lower upper limit. Any setting alows the system to return to 0 when coasting.

A KT controller equipped bike is not torque sensing! They do not make a controller with the ability to read/have an input from a torque sensor.....

I'm left wondering if you have a KT controller? Model number on your controller is?

 

[richj8990]

What mid drive do you have? I'm not familiar with any of them that are compatible with a KT controller. Same story with the SW900 display? I'm not familiar with an SW900 that will work with a KT controller. KT controllers require a KT display......

Torque sensing bikes WILL add/subtract power depending on load and PAS level. PAS level sort of sets the upper limit available through the torque sensing. Lower PAS level = lower upper limit. Any setting alows the system to return to 0 when coasting.

A KT controller equipped bike is not torque sensing! They do not make a controller with the ability to read/have an input from a torque sensor.....

I'm left wondering if you have a KT controller? Model number on your controller is?

L-Faster / Dome Motor 450W Mid-Drive Build

Finally got everything to work. I wanted to do a test ride up a 12% fire road near my home but decided to instead clear some overgrowth 20 minutes away by riding this offroad about 8 miles total. 50 lbs with front wheel/tire. Will try to trim closer to 45 lbs with different crankset later. Notes...

endless-sphere.com

The vast majority, like 95-99% of these online are sold as rear left-wheel bolt-on's, as in you take off the left brake rotor and bolt this on. Which is very lame, because you just lost your rear braking and now have to pray the chain will stay on between the rear wheel and motor mount with all the bumps that even a pavement bike has. The mid-drive version is the only way to go, although has other issues though (that I'll fix one at a time).

Besides PAS not working, the chain tensioner is very finicky and cheap. The motor chain has to be very lax or it will seize and literally pop right off the 42t chainring. That's another reason why I'm going to try and switch to a 3x Shimano hollowtech knockoff crankset for this, so I can use a real 104 BCD narrow wide chainring for the motor drive. And also a narrow wide for the middle chainring that's driving the rear cassette too. No front derailleur. The kit comes with FOUR chainrings, all steel and no this not a typo: they are all 64 BCD. Even the 42 and 44t ones. I have no clue why they made it like that. No conventional crank spider. Just a center bearing they drilled some 64 BCD holes into lol. That's called off-brand Chinese!!!



Now to answer your questions.

SW900 is on a different bike, a hub drive. I'm just comparing the throttle smoothness. I feel the 'imitation' torque on that controller is smoother for throttling while the KT is 100% for sure smoother for PAS.

The controller I'm using is KT 36/48 ZWSRM-ff01 12 mosfet. They last for a couple of years, couple thousand miles and then die. Most of my bike riding is with a normal acoustic bike, only use e-bike for screwing off or doing chores like trail work or dropping my car off for an oil change.

Neither of my two e-bikes have torque sensing. The KT has the imitation torque option, and I don't know what EBikeling calls it on SW900, but it works relatively better. Remember I'm typing the term imitation torque, not torque sensing

The power is added or subtracted for imitation torque to me based ironically on speed sensing. This terminology is going to get really confusing real fast. They should not call something either speed sensing or torque sensing, especially when there is no 'real' torque sensing. But when I'm going slower under otherwise CONSTANT throttle or PAS power (cannot emphasize the term constant enough), the controller knows it and ramps up power accordingly. That's really speed sensing power adjustment under what was constant power on a level surface. Elevation changes, controller senses speed change due to elevation change, power adjusted accordingly. Then when I'm at the top of the incline, the controller senses more speed and ramps down the power accordingly. Then when I'm back on a level surface the power goes back to constant. This happens with both SW900 and KT. It doesn't matter if SW900 doesn't have a 'term' for that, it still happens.

Here is your (AHicks) quote from a different site: "What it amounts to is a fixed amount of power supplied to the motor, which changes according to the PAS level, and speed no longer has any affect on what's going on (none!)."

Some comments on that sentence above. First, on this topic, for a KT controller everything revolves around the P3 and C4 parameters. Just my extremely humble opinion. See below.

P3 parameters affect the power assist control setting. When the value is set to 0, the throttle is dependent on the PAS Gear Ratio. (what they call speed control) This means when the PAS gear ratio is 1, the throttle will provide the least amount of power, whereas when it is 5, the throttle will provide the most power. If the P3 parameter is set to 1, the throttle will provide the maximum power regardless on the PAS gear ratio chosen. (what they call imitation torque)

So for 0 the throttle power is linked to the PAS 1-5 on the display. If I'm doing throttle only, I'll instead set P3 to 1, because throttle at least on a KT DOES NOT give the same range of power as when real PAS or pedal assist is active. It will fade. Even in PAS 4 there may not be enough power to get up the incline again with throttle only. In PAS 2-4 with real pedal assist there IS enough power to climb hills all day long, but not with throttle only. That's why I need to somehow find a way to get PAS back if using a KT controller. Throttle only has consistency/sensitivity problems.

C4 settings are for the handlebar functions. The setting range is 0-4. This is accompanied by a table below, defining each value. (all of this below is non-zero startup only, where you don't need to pedal for the throttle to initiate power)
0 Zero startup handlebar Non-zero startup
1 Zero startup, throttle speed limited to 6km/h
2 Zero startup, throttle speed limited to rider's specification
3 Zero startup, zero gear effectively
4 Throttle gears is distinguished according to the display meter

I don't understand how they determined the speed limit for 2, but for me, if I'm throttling I would choose either 0 or 4 above, and then P3 as "1". "4" for C4 should be the same as P3 being set to "0".

P3 "0" and C4 "4" --- Both meter out five levels of throttle power through the display up and down button. As mentioned above, I don't like the throttle being tied to 5 levels of power because it's not the same level of consistent power compared with real pedal assist. It's just not, I can tell the difference. So I set it to max throttle power and thumb-adjust accordingly. If I can get PAS to work, remember without a freewheel, then I'd use a lower level of PAS with gentler surfaces and then ramp up any extra power I need through the throttle. Can't use PAS 4 or 5 if there is no freewheel because I don't know if I can stop the pedals from turning if there is 500+W turning fixed cranks/pedals.

So in my case I'd choose P3 "1" and C4 "0".
Not P3 "0" and C4 "4".






Here's the future of this motor for me personally:
1. Live with the crappy crankset because it can freewheel, and just throttle it.

2. Keep said mediocre crankset and switch to SW900 so the throttle is relatively smoother, but no PAS again. Because the EBikeling PAS is horrible IMO.

3. Even if I keep the crankset, eventually the rings will wear out, and then what am I gonna do? Buy 64 to 104 BCD adapters? This crankset is already wobbling after 20 miles. Some of the bolt holes have been manually drilled into a center bearing on the axle, how long are they going to last like that? Sooner or later I need a real crankset. The issue with that is that I'll lose the freewheeling and will have fixed chainrings and thus moving pedals whenever power is applied. The advantage is ironically enough that I'll be able to try a hollowtech-modified 2-piece PAS system. Sounds like an oxymoron, because hollowtech cranksets essentially have zero space between the external bearings and the crank pedals. In reality, the screw-on bearing cups are 10-15mm deep. So you can use a 3mm DIY motor bracket between the frame and external bearing, and on the right a 1-2mm PAS sensor next to it, then screw the drive side cap on firmly, everything should be OK. After all, people do that with shim spacers that come with the hollowtech bottom brackets to line up the pedals on each side of the frame. The only real issue is: can I shift like I normally do with a non-ebike, as in briefly stop pedaling, when in this case e-power is still turning the pedals!?! As in get the pedals to stop turning for 1/2 second while I shift? That will take some getting used to...

 

[AHicks]

So what I am seeing is you have a pretty unconventional install, one that I am not familiar with at all, even with you explanation. I'm not even going to try and wrap my head around all you have going on. I'll just leave it with a best of luck!

 

[richj8990]

Torque simulation controllers do not control torque​

In recent times, there has been a surge in the marketing of motor controllers as torque simulation controllers. This led me to assume that these controllers control the motor torque rather than its speed. However, this turned out to be false, I discovered that these controllers merely regulate the motor voltage, just like regular motor controllers.
See my full article here:

Torque simulation

Here, I share technical solutions that arised during the development of my solar-bike.

www.avdweb.nl

From a different website "On the other hand, Speed is also inversely proportional to Torque. Torque is directly proportional to the current and Voltage is directly proportional to the Speed."

If an electric motor is trying to increase the torque, then if the above statement is true, the amps are raised and not the volts. Imitation torque raises the power (watts) when the motor encounters an incline. That means either the volts are raised or the amps. The speed doesn't really increase up a hill with imitation torque compared with the speed the motor (bike) had when it was on a flat surface before the incline. If anything it decreases. Which means volts either stay constant or decrease a moderate amount, and amps are raised up a lot to raise the overall wattage for the incline. The reason people overheat their controllers or god forbid their motors is because of peak amps when climbing hills at high wattage for long periods. That is consistent with amps raising torque not volts.

In the article "Further examination revealed that the torque simulation is achieved by controlling the motor voltage through a combination of the throttle voltage and the cadence, which is the rate at which a cyclist pedals."

The I don't know about the other controller, but the KT controller has an imitation torque option through throttle only. Even their PAS system is not really cadence-based. It only measures if the PAS magnets are picking up a signal or not. You can pedal at 30 rpm or 120 rpm and the watt output is going to be the same, on a flat surface, incline, or decline. It will vary a little but nothing like the imitation torque option with the throttle. It's the same thing with the cruise control option, it will behave the same way: watts up on inclines, watts down on even a 1-2% descent. Voltage relatively constant compared with amps. No one should ever burn out their controller on a 1% decline! Because amps are down, voltage constant.

KT48ZWSRM-HY02L - characteristics?

Hi all, I'm looking at buying a kit (rear hub) that includes this controller: KT48ZWSRM-HY02L But trying to find out more about this controller. The seller claims it is a sine-wave controller but I noticed on some other online resources that ZWSRM are square wave, while SVPRD are sine wave...

www.pedelecs.co.uk

KT make both types and use imitation torque or better known as current control for power though in the settings you can change this to Speed control if preferred.

Two options at least on an LCD 3/5: imitation torque or speed control. One or the other. Again, imitation torque is controlled by current.



From Quora: A DC source with a heavy variable load will cause a fluctuating voltage. At no load the voltage will be higher and under full load the voltage will be slightly lower. This is due to the internal resistance of the source supply. The voltage lost due more current flow through the internal resistance of the supply reduces the available voltage. Unregulated power supplies will show some voltage fluctuation from no load to full load. A well designed regulated supply will allow enough head room so the output voltage does not vary. (as in a controller)

Full load = when going uphill, slightly lower voltage, much greater amperage = more watts to compensate for the incline.

 

[stancecoke]

A KT controller equipped bike is not torque sensing! They do not make a controller with the ability to read/have an input from a torque sensor.....

that's wrong. BMS Battery had a bundle of a (poor) torquesensor and a Kunteng controller (branded as "S06ST") for many years:

B. B. Torque Sensor System

B. B. Torque Sensor System

bmsbattery.com

But who wants to use a Kunteng in these days? It's totally outdated, I can't understand why it's still that popular

regards
stancecoke

 

[AHicks]

that's wrong. BMS Battery had a bundle of a (poor) torquesensor and a Kunteng controller (branded as "S06ST") for many years:

B. B. Torque Sensor System

B. B. Torque Sensor System

bmsbattery.com

But who wants to use a Kunteng in these days? It's totally outdated, I can't understand why it's still that popular

regards
stancecoke

Wasn't talking about a history lesson. Maybe a I should have said KT does not have a controller available TODAY capable of accepting torque sensor input. If they did have one, personally I wouldn't mind giving it a try.

You're entitled to an opinion regarding them, but I think they're tough to beat for most applications. Agree to disagree if necessary.