Geared Hub Motor Quality

When talking low speed torque, I believe the MAC 12t is the 800lb gorilla. The GMAC's, being 10t and 8t, are only slightly below that.

There is an outstanding spreadsheet all set up to plug in all sorts of variables to allow answers to some of the questions you are asking here: https://ebikes.ca/tools/simulator.html

This spreadsheet is incredibly accurate and eliminates a lot of guessing. Many, or most of us find you can trust it....

AHicks said:

neptronix said:

Variable regen is a controller feature that works with any motor that can do regen.
The GMAC is not any more/less complex to deal with than a DD.

Click to expand...

This needs to be put in a little better context, don't you think?

For starters, a DD regen will work just fine with a pretty simple/inexpensive controller and an on-off switch. That same controller, set up the same way and powering a GMAC, would have the rear tire locking up when applied.

The GMAC would require the sophistication available with a Phaserunner (and CA3) to control the regen potential from an armature turning 5 times faster than a DD, able to supply up to 90 amps of braking power. That's some serious braking I personally would want to be careful with....

Anyone trying to tell a total rookie any different is doing that rookie a dis-service - suppling bad information that could cost somebody a serious amount of time and money to find out they were fed a line of BS, intentionally, or not.....

Click to expand...

Do you know that this is a risk for sure? the only real difference, electrically, between a DD and a geared motor is the speed which the poles alternate... and also some differences in induction values. The difference in speed itself does not create higher or lower amps.

Also, you're aware that the motor controller controls the level of regen that happens, not the motor itself? therefore, you could program your controller to take in 90 amps or much above that during regen if you wanted to, and both a DD and a MAC would cause your MOSFETs to explode. Or you could set it to a reasonable level and have it not explode.. either way, it is not the motor that determines the amount of power coming in/out in any way.

A GMAC/MAC can handle peak phase amps beyond 300A if unlimited.. if 3 phase controllers did not have current limiting then you would need a giant 36 FET controller..

A GMAC does not require a phaserunner in any way.. and in fact, a phaserunner is not entirely capable of pushing a MAC/GMAC to it's limits. You buy a phaserunner because you want all the nice features, small size, quietness, and efficiency it has. You don't buy it because it's the only controller that can handle a MAC.. which reaches saturation at 4-5kW, as i showed you guys in 2011..

neptronix said:

you could program your controller to take in 90 amps or much above that during regen if you wanted to, and both a DD and a MAC would cause your MOSFETs to explode. Or you could set it to a reasonable level and have it not explode.. either way, it is not the motor that determines the amount of power coming in/out in any way.

Click to expand...

My Infineon clone controller's parameters are EBS "force" (0 - 200) and "limit voltage" (0.3V - 77.6V). In principle I think the electrical unit of "force" is volts, but my guess is that what we're talking about here is stopping power and it's essentially watts; the parameter is scaled against some unknown factor such that 0 is none and 200 is too much. Mine is set to 80, for an ordinary direct drive hub; limit voltage is 77.0V.

Would a GMAC need substantially different configuration parameters?

AngryBob said:

Perhaps someone in California with a hundred or so of them sitting in his warehouse with his logo emblazoned on them, which no one wants? Also interesting that the 2-speed thread has once again recently come back to life, albeit briefly.

Click to expand...

Heh.. :lol:

I remember those 2 speed motors.. i wanted one until i saw Xiongda's actual power graphs.. showing they were capable of less than half the continuous power of what said warehouse owner was claiming.

I notice that those motors have not been for sale on their site for a while.

donn said:

My Infineon clone controller's parameters are EBS "force" (0 - 200) and "limit voltage" (0.3V - 77.6V). In principle I think the electrical unit of "force" is volts, but my guess is that what we're talking about here is stopping power and it's essentially watts; the parameter is scaled against some unknown factor such that 0 is none and 200 is too much. Mine is set to 80, for an ordinary direct drive hub; limit voltage is 77.0V.

Would a GMAC need substantially different configuration parameters?

Click to expand...

I have ran infineon clones for over a decade. Mine do not have an EBS 'force' parameter. There was a parameter that would change regenerative braking from 0, 10, or 20 battery amps in my older ones.. I would guess that your parameter is a limit of phase amps.

The limit voltage parameter always had to do with the voltage limit of the battery. It was a high voltage cutoff to prevent batteries from being overcharged. Once you hit the threshold, regen would stop working at the specified voltage.

GMAC isn't electrically any different than the MACs that have been around for over a decade. The only difference to a controller is that it has a fixed clutch and is therefore capable of regen.

neptronix said:

The limit voltage parameter always had to do with the voltage limit of the battery. It was a high voltage cutoff to prevent batteries from being overcharged. Once you hit the threshold, regen would stop working at the specified voltage.

Click to expand...

Really? I'm reasonably certain mine is set to 77V, which as you may imagine is a little high for my battery. I was thinking that might be relative to the input voltage from the motor, instead.

The "xpd" software I'm getting this from uses the designation KH6xx (Infineon 4). In the absence of any information to the contrary, if I hooked up a GMAC, I'd set the EBS "force" parameter to the same 80 out of 200 and expect to the get the same result as my direct drive.

donn said:

neptronix said:

The limit voltage parameter always had to do with the voltage limit of the battery. It was a high voltage cutoff to prevent batteries from being overcharged. Once you hit the threshold, regen would stop working at the specified voltage.

Click to expand...

Really? I'm reasonably certain mine is set to 77V, which as you may imagine is a little high for my battery. I was thinking that might be relative to the input voltage from the motor, instead.

The "xpd" software I'm getting this from uses the designation KH6xx (Infineon 4). In the absence of any information to the contrary, if I hooked up a GMAC, I'd set the EBS "force" parameter to the same 80 out of 200 and expect to the get the same result as my direct drive.

Click to expand...

Honestly i am not aware of the specifics of how the infineon v4 works ( i only have older ones and use the oldschool em3ev supplied programming software which i've done some hex editing on to expand the range of values ).. i can just tell you that in the old version, the voltage limit was a high voltage cutoff meant to prevent the battery from charging to too high of a voltage.

If you lived on top of a mountain, fully charged your battery, then regeneratively braked down the mountain for miles 'n miles.. then you'd need that high voltage cutoff as a safety measure to save your battery from being overcharged.

Your BMS has a high voltage cutoff though.. so you likely don't need this extra form of protection. But ideally it should be set to the fully charged voltage of your battery.

AHicks said:

Surprised nobody has mentioned the MAC and GMAC geared hubs? They are just about the strongest available.

Click to expand...

Outside of ES dot com "Cutler" MAC is little known.
Cutler MAC

neptronix said:

AHicks said:

neptronix said:

Variable regen is a controller feature that works with any motor that can do regen.
The GMAC is not any more/less complex to deal with than a DD.

Click to expand...

This needs to be put in a little better context, don't you think?

For starters, a DD regen will work just fine with a pretty simple/inexpensive controller and an on-off switch. That same controller, set up the same way and powering a GMAC, would have the rear tire locking up when applied.

The GMAC would require the sophistication available with a Phaserunner (and CA3) to control the regen potential from an armature turning 5 times faster than a DD, able to supply up to 90 amps of braking power. That's some serious braking I personally would want to be careful with....

Anyone trying to tell a total rookie any different is doing that rookie a dis-service - suppling bad information that could cost somebody a serious amount of time and money to find out they were fed a line of BS, intentionally, or not.....

Click to expand...

Do you know that this is a risk for sure? the only real difference, electrically, between a DD and a geared motor is the speed which the poles alternate... and also some differences in induction values. The difference in speed itself does not create higher or lower amps.

Also, you're aware that the motor controller controls the level of regen that happens, not the motor itself? therefore, you could program your controller to take in 90 amps or much above that during regen if you wanted to, and both a DD and a MAC would cause your MOSFETs to explode. Or you could set it to a reasonable level and have it not explode.. either way, it is not the motor that determines the amount of power coming in/out in any way.

A GMAC/MAC can handle peak phase amps beyond 300A if unlimited.. if 3 phase controllers did not have current limiting then you would need a giant 36 FET controller..

A GMAC does not require a phaserunner in any way.. and in fact, a phaserunner is not entirely capable of pushing a MAC/GMAC to it's limits. You buy a phaserunner because you want all the nice features, small size, quietness, and efficiency it has. You don't buy it because it's the only controller that can handle a MAC.. which reaches saturation at 4-5kW, as i showed you guys in 2011..

Click to expand...

There is no way on God's green earth that a DD hub can provide the braking power (regen) potential of a GMAC (all else being equal). There's SO much more potential available from the GMAC you need to deal with it in an entirely different manner. Makes no difference what controller you are using. Bottom line, you DON'T simply turn a GMAC going down the road at 20mph "on" like you can a DD.

Then there's the potential GMAC braking/regen available at MUCH lower speeds than would be practical with a DD (due to the fact the motor is turning 5x times faster)- that's still effective at speeds below 10mph - also not possible with a DD.

tomjasz said:

AHicks said:

Surprised nobody has mentioned the MAC and GMAC geared hubs? They are just about the strongest available.

Click to expand...

Outside of ES dot com "Cutler" MAC is little known.
Cutler MAC

Click to expand...

Thanks Thomas. As a relative newbie, I didn't know that (among other things!).

AHicks said:

There is no way on God's green earth that a DD hub can provide the braking power (regen) potential of a GMAC (all else being equal). There's SO much more potential available from the GMAC you need to deal with it in an entirely different manner. Makes no difference what controller you are using. Bottom line, you DON'T simply turn a GMAC going down the road at 20mph "on" like you can a DD.

Then there's the potential GMAC braking/regen available at MUCH lower speeds than would be practical with a DD (due to the fact the motor is turning 5x times faster)- that's still effective at speeds below 10mph - also not possible with a DD.

Click to expand...

As someone who has shoved 4kw into a MAC and 8kw into a mid-sized DD and also ran VERY STRONG regen as the exclusive rear brake, i have to say... nope.

There's dozens DD motors out there that can provide much larger regen power than a MAC ( and also vastly greater power moving forwards - both peak and continuous ). Yet the thing is, on an upright frame, you'd never really be able to use the full capabilities of the braking of a MAC or larger motor because of physics.

I could never use the full regen power of my leafmotor because i would have been thrown forward off the bike the second i hit it.. so what did i do.. tune it down to a fraction of what it was capable of.

The MAC has a lot of low end torque because it effectively has a very high pole count, but again, it's up to the controller to determine how much of that power potential is used, whether we're talking regen or forward power. You could use a bargain basement controller on a GMAC and set the regen low and you will get....... a low amount of regen.

A MAC stator may spin 5 times faster but got it's an extremely small mass of copper/iron/magnet. Around half of it's weight is gears, bearings, shells, etc. So it's not the big monster motor you think it is

( but it's still a damn good motor! )

The bigger the diameter of the hub the more torque you get: Leverage

The bigger the diameter, pole count, rpm, width, copper fill, lamination thickness, and RPM, you mean.

( but again, this just changes power potential. The controller determines how much of this power potential is used both in regen and forward power. )

Certainly, the more (potential) regen the motor offers

the more critical the sophisticatiin of the controller becomes

Just like a smooth "ramp up" helps the ability to tap megaWatts while preventing gears from stripping

a tweakable "reverse" rampup increases the regen power you can capture not just for recharging but also

for effective braking without pad wear

**without** throwing the rider over the handlebars.

Right

Correct.. if you want high power regen then a very tuneable controller, especially if it does variable regen, is important.
If you don't, then the wimpy settings on most controllers and the on/off action is fine.

OK, I get it now. We are playing on vastly different levels. My level is more along the lines of production bikes using commonly seen production components (including batteries).

When you start talking megawatt powered Leaf (DD) motors many or most will never even see, let alone ride, you might want to make yourself a little clearer - providing the context I was talking about in my first reply......

No experience there, nor interest really..... -Al

There's a megawatt Leaf motor?

What "I could never use the full regen power of my leafmotor" means is he's where practically everyone with a direct drive hub is - we use only a fraction of the potential, because we don't want braking that's anywhere close to locking up the wheel.

We'd be doing the same with a GMAC. But at this point it seems to be academic - every single GMAC owner has apparently opted for a Baserunner or Phaserunner controller, likely for the "electronic freewheeling" feature as much as anything else, so it's immaterial whether a common Chinese controller could easily be configured for one. If there's a special setting for fixed regen appropriate to the GMAC as a geared hub, no one knows what it is. It's a point of dispute between several people who don't have a GMAC and can only guess.

Well I for one would need a far more powerful controller than those, peak 90 phase amps, cont. 50A is hardly making full use of the motor eh?

I'm looking for not only FOC with proportional regen,

but the motor's power to really only be limited by the heat rising too quickly on long steep hills.

I assume Nucular 24F is one decent choice.

Sevcon would likely also be great, so long as a tech / dealer were available for programming / tuning help.

Bigger ASI BAC? Same programming issues?

Everyone dumps on Kelly, but they at least apparently give good support?

Sabvoton at this point seems to just be a controller category, with many different manufacturers. Was it ever a single company? Where do I fid out about the "zombiess variant"?

APT and Mobipus still exist?

Maybe PowerVelocity?

You'd need more mass to help shed the heat in the motor, dd's like the mxus 3kw 45H have the mass, alum stator to shed heat. Statorade helps as well. There aint much mass or metal in geared hubs. The windings would smoke before the plastic gears melt, be interesting to do a search on melted gears, or the power level required to start chipping the gears. Be simple enough, but at 3am time to snooze.

Again, only talking geared hub motors here.

I am of course not looking to **actually** pull the full output to destruction "no matter what".

A temp sensor would allow derating to prevent heat buildup.

Just not willing to be limited by an underpowered controller.

by john61ct » Nov 08 2020 7:28pm

Certainly, the more (potential) regen the motor offers

the more critical the sophisticatiin of the controller becomes

Just like a smooth "ramp up" helps the ability to tap megaWatts while preventing gears from stripping

a tweakable "reverse" rampup increases the regen power you can capture not just for recharging but also

for effective braking without pad wear

**without** throwing the rider over the handlebars.

Right

Click to expand...

I like Mega-Watts with smooth ramp-up, ride that edge of destruction and out of control but keep the pressure at the maximum.

The geared hub motors having positive mention here are probably good enough. I'd rather ride good enough than not ride best. Rather ride $ than not ride $$.

My only significant experience is with a MAC 12 turn rear geared hub. It works very well, never a problem in 720 miles. It's powered by a 52v, 20ah battery through a 12 FET controller set to a 30 amp limit. There is a Cycle Analyst V3 in the system and it turns out to be very, very useful. The previous ebike was ridden 6 years without a dashboard, that worked OK but the information gathered by the CA V3 makes life a lot more worry free.

Watt ratings are not very accurate or useful, watts produced by a hub motor depend on battery voltage and the amps the controller sends to the motor. The CA V3 records all the data for each ride, this is then photographed with an iphone and the data cleared when the recharged battery is installed. The data is entered on a excel spread sheet to calculate useful things.

Back to watts, every ride will have a max watts and an average watts. My max watts ever was 1778, the average max watts for 23 rides totaling 720 miles is 1420 watts. Average watts over that 720 miles is only 130 watts. Avg ride speed was 11.9 mph, but riding is often in the 14 to 16 mph range. Top speed with pedaling has been 26 mph, easy top speed is 23 mph. This performance is similar to an Aotema direct drive, sensorless, front hub motor that I used for 6 years. Both the Aotema and MAC are brushless, that is they are three phase AC motors. Note: max amps ever were 32.5 amps, if the controller was reset to 40 amps max then the top speed might be higher, but there would be more chance of breaking the rear dropouts or swing arm.

I'd rather have the rear GMAC rear hub motor because it uses round axles, instead of slotted axles, and has a really good torque arm arrangement. Problem is it does not have the freewheeling clutch that all other geared hub motors have. Well, that and it uses a specific controller and the kit cost lots more than my regular MAC motor and 12 FET controller from EM3ev. If I was willing to spend more, I'd have the GMAC, matching controller and CA V3.

Regen? My MAC doesn't have it because it has a clutch, my direct drive didn't have it either. Range without regen: using the 52v, 20ah, 14s, 8p battery, theoretical range from full charge to minimum charge is around 115 miles. I never fully charge the battery, usually it is charged to about 4.13 volts per cell. Average final volts are 3.80 volts per cell, average ride distance is 31.3 miles. The MAC, controller, CA V3, and battery can easily do a 50 mile ride. My crotch can't easily do 50 miles.

Hope this information helps. It's a good idea to sort of know what to expect.

donn said:

If there's a special setting for fixed regen appropriate to the GMAC as a geared hub, no one knows what it is.

Click to expand...

In case it's useful, I had already gone to the extreme of contacting Grin on this question, and just now received their thoughtful reply:

Grin said:

We're not familiar with this controller, but I'd recommend not using more than 100A of phase current with a GMAC in either regen or traction due to torque arm limits.

Click to expand...

I asked in terms of my controller, a common presumably Chinese Infineon type. I can't say that this information really adds much to my ability to correctly configure regen for a GMAC the first time, but at least there's no hint here that the GMAC "uses a specific controller" or that there's any likely disaster awaiting someone who configures it as for common direct drive regen. So that's what I'd do, and if it turns out the regen needed to be dialed back, then I'd do that.

I bet limiting the GMAC motor current for traction to 100A is only necessary with an unsophisticated non-FOC controller.

A Nucular or other FOC should let you tweak the speed ramping curve, control the actual torque produced.

Anyone know if the GMAC can get up to say 120 N.m? Other geared hubs that can?

Ideally same for the variable regen, but just keeping that under 100A should be fine, no one should be relying on this for emergency panic stops, in my vase just for drag braking down into the valleys with a big load.

john61ct said:

I bet limiting the GMAC motor current for traction to 100A is only necessary with an unsophisticated non-FOC controller.

A Nucular or other FOC should let you tweak the speed ramping curve, control the actual torque produced.

Anyone know if the GMAC can get up to say 120 N.m? Other geared hubs that can?

Click to expand...

I ran over 200 phase amps in a MAC with a 12FET infineon clone ( EB3 ) about 10 years ago. Changing the battery to phase amp ratio on those clones allowed you to control the torque curve.

A MAC/GMAC can produce well over 120nm of torque. The real limit of torque is the gears inside the motor and also with the GMAC, i imagine the torque arm is not designed to handle the full grunt of what a MAC stator can dish out ( which you cannot push for very long anyway.. )