Dave's 80:1 RC mid drive kit build log

[StinkyGoalieGuy]

Most definitely I will. Cutting production parts this week, I'll work a quick video in there.

You are killing us with anticipation. Can you just make a quick smartphone video held next to the drive so we can see how much quieter it is now? Just put it up on a stand, no need to actually ride the bike.

 

[recumpence]

The issue with this system is not the system. I think Dave has done a wonderful job with it. The limitation of this system is the pedal drive components. A 3220 driving the cranks (a relatively high KV 3220) would be truly insane. It would be difficult to ride and VERY hard on the final drive components of the bike. This is the same as with any system driving through the normal derailer chain system. However, if you keep the power reasonable, the reliability is fine and the system runs wonderfully.

I think the 3210 is an absolutey perfect motor for this type of system. I pull 4kw from my wife's bike with a 3210 and that is a single ratio left side drive. Heck, a 3205 would work on Dave's drive too. The thermal mass of the 3210 is a better match than the 3205 though.

Matt

 

[tangentdave]

Can you just make a quick smartphone video

I learned my lesson on the first video, it needs to be accurate. I'll hurry, trying to do everything at once.

I'm pretty stoked about Grin's Phaserunner. I have the same developer chipsets from ASi and Texas Instruments waiting to be configured for the drive, but now I can have the same thing already productized. These chips are pretty much the state of the art for electric motor control. Their control algorithms are more precise and provide more and better power than a sensored motor. Field weakening, field strengthening, FOC- we're gonna passing the motos pretty soon here.


-dave

 

[jk1]

Tangent uses 80:1 between the motor and crank.

Nathan you're right about the kV tradeoffs in electric motors. Let me revise my statement and say that for a given current input Tangent will make more torque than cyclone due to the reduction. Spinning the crankset at 250RPM seems to be a negative for most riders from what I gather, so to get the same torque (which is acceleration) from the cyclone, more current is required.

Twice the gear reduction to the cranks would cause double the torque at the cranks but at the expense of half the top speed of the bike. And the expense of the gear system having twice the torque loads, people on the sphere often try and increase the chain speed to allow more power to go through weak bicycle components not the other way around.

Also one factor people are ignoring is that you are assuming both motors have the same type of torque curves, this is not the case as they are very different motor designs. The RC motor is good light and high power but the downside is the torque curve is at the higher end of the RPM curve compared to the cyclone motor which is an IPM cored design. The Cyclone Iron core helps amplify the magnet force at low RPM, so the cyclone motor is very efficient at LOW RPM and maybe has twice the torque at LOW RPM vs an Astro, hence you cannot say it will have twice the USABLE torque as the torque curves are not the same.

It was discussed here how cyclone motor has high efficiency compared to Astro at low RPM: https://endless-sphere.com/forums/viewtopic.php?f=30&t=66278

Low RPM torque is what you want to take off or change gears on an ebike. Whereas for an RC plane having the peak torque at HIGH rpm suits having high power at high speeds not at takeoff. In other words the design does not particularly suit an ebike. Also being so light in the motor thermal mass I bet if you did a lot of stop and starts on steep road or heavy bike/rider you would overheat and blow up both motor and ESC very easily. Again its using a motor for what it was not designed for Light Plane with not many stop starts vs heavy EBIKE with lots of stop starts at high load.

 

[tangentdave]

Don't forget the 80:1 reduction taking the Astro into it's efficiency range just about immediately, and it stays there for the rest of the rev range. The cyclone efficiency falls off in the RPMs where you ride it most.

 

[adriftatsea]

Low RPM torque is what you want to take off or change gears on an ebike. Whereas for an RC plane having the peak torque at HIGH rpm suits having high power at high speeds not at takeoff.

I'm not sure I could handle 'more torque'. I can't use anything less than 3rd gear to take off and I'm quickly up to speed. The high reduction means the motor is spinning even before I start to move ... this eliminates the jerkiness on takeoff and the worry of dumping lots of power into a stalled motor. I use mine for stop/go commuting and haven't had any issues with heat.

I too cross-shopped the cyclone, but passed when comparing the quality of parts/machining. Again, short of Tangent and LR, it seems no kit manufacturers actually 'get it' and ride their own kits daily. And without being able to test drive all the new kits, this becomes the best way of proving of quality/parts/whatever.

 

[Ralith]

It's been said plenty before, but I'll repeat that this kit has so much torque that you'll easily do a wheelie from a dead standstill if you aren't in a high enough gear. I also do start-stop commuting on the kit, including starting from nothing in top (38:11) gear on hills up to 20%, and the only thing I've ever ever strained is the controller's current safeties. No temperature issues whatsoever. The 3210 definitely heats up a bit when you run it over 2kW, but it's rated for 200°C and I can tell you I've never gotten anywhere near that. Good engineering goes a long way, and this kit is just swimming in it.

The only noticeable issue the HV80 has with starts is a bit of roughness due to a less-than-state-of-the-art low-RPM control algorithm. Despite that, the torque's fine, and it smooths out as soon as you're moving at all.

 

[jk1]

Again, short of Tangent and LR, it seems no kit manufacturers actually 'get it' and ride their own kits daily.

LR doesn't seem to personally actually ride or test his kits or design but relies on failures and his customers feedback to do that. Tangent not sure how many KMs they have personally tested it or also rely on your customers feedback? I found on the AFT kit they say they have done 10-15,000 km of their own testing.

This is a good point to make, that the product should be full tested not just up and down the street or 500km but to the real life of say 5000 to 10000km before it is sold ,otherwise the customers are the guinea pigs and the parts repair and fail rate is unknown. Eg on other thread the BBS02 750w is cheap to buy but very costly to run since you need to rebuild the gearbox and replace the plastic gears quite often if you do lots of Km's.

 

[macribs]

I'm pretty stoked about Grin's Phaserunner. I have the same developer chipsets from ASi and Texas Instruments waiting to be configured for the drive, but now I can have the same thing already productized. These chips are pretty much the state of the art for electric motor control. Their control algorithms are more precise and provide more and better power than a sensored motor. Field weakening, field strengthening, FOC- we're gonna passing the motos pretty soon here.


-dave

Are you saying that you will also make a controller that fits your drive, what kind of FET will you be using and what kind of voltage and current limits?

Have you tested your drive with the Phaserunner?

 

[tangentdave]

I plan on incorporating a purpose built ESC into the Tangent kit. ASi are nice folks with nice tech and I planned on starting with the BAC500 just like Grin did. Grin's $300 asking price is exactly what I imagined and they put it in an awesome looking package with convenient access to the heatsink and improved a few internal components. What else could I add? I haven't personally had the chance to ride the ASi controllers yet, but district9prawn is running Grin's BAC500 (basically a Phaserunner) down under on a Tangent kit.

It's not so much better FETs, it's better everything. Did you guys know this controller pings the motor between phase pulses on startup to determine the rotor position using the variation in winding inductance caused by the magnets? These controllers know way better than 120deg hall signals where the rotor is and how to manipulate the B field to make the most of it. Combustion engines are so 20th century.


-dave

 

[macribs]

So start from standstill will be smooth like a hub with that ESC? I mean if the position of rotor is known there should be no stutter or rattling at take off?
Will it work for take off from stand still if parked in a slope? Or is pedal take off needed as with regular RC ESC?

 

[brumbrum]

Apologies for my naivety but is there no way that this motor can for instance have two composite gears and one steel, or something similar to quieten it down. I would bite Dave's hand off for this kit if it could be applied to using on the road or more fittingly in the woods over here in the UK without it turning heads. I can see that the creator is doing his damned hardest to dampen down the straight cut/dog box type gear whine, but could there be a more disposable 'race and replace' softer gear option?

 

[Ralith]

So start from standstill will be smooth like a hub with that ESC? I mean if the position of rotor is known there should be no stutter or rattling at take off?
Will it work for take off from stand still if parked in a slope? Or is pedal take off needed as with regular RC ESC?

It starts from a standstill just fine up a 20% grade with the RC ESC. The reduction gearing mostly absorbs any roughness from the RC-style startup. I expect it'll be even better with a Phaserunner or similar, but it works great even as-is.

 

[dirkdiggler]

Eg on other thread the BBS02 750w is cheap to buy but very costly to run since you need to rebuild the gearbox and replace the plastic gears quite often if you do lots of Km's.

I've followed that Bafang thread since the get go. That statement is pure FUD. Guys have ridden the Bafang for three years or so now and there are very few issues with them. Even if they did - the plastic gears are not "costly". I have put a ton of abuse on mine and feel at this point even buying a full replacement unit wouldn't even make me upset. I've gotten my moneys worth already. The Tangent Dave kit is much more expensive and you'll likely end up with the same issues with parts failing. That doesn't mean the kit is junk. Parts wear out, its just a fact of life.

 

[tangentdave]

Apologies for my naivety but is there no way that this motor can for instance have two composite gears and one steel, or something similar to quieten it down.

What are you trying to do, ride a motorcycle down the sidewalk? The drive is not loud, I'll show you in the next quick video.

The gears are gonna stay chromoly steel for higher efficiencies, longer life and they're seriously tough. This style of gearing uses compressive stress to cause rotation so we want a very hard material.

-dave

 

[brumbrum]

.[/quote]

What are you trying to do, ride a motorcycle down the sidewalk?
-dave[/quote]

No, i am Just trying to break the law without being caught. :wink: That would require something that has a certain amount quietness as to not raise eyebrows. A 250w and 15mph limit is not much fun, and yes i absurdly break these limits every time i pull a throttle on my current rides with hub motors, but what i know for sure is that metal gears are loud. This is not a criticism, i love what you have done and i have money to spend, but the 'potential' noise/sound level is holding me back. Your full kit ticks ALL the boxes for what i need/want on my commencal down hill mtb which currently has a five year old tired and heavy magic pie on with not enough battery or space on it to get good current and range from(currently 67v and 10ah lipo, pulling 40amps continuously on hilly trails, it would pull more if my CA was not holding back to preserve the pack). Your kit would be a perfect solution whereby i would have enough battery to take it to its limits and get good range. Dave, I want one! But i have to feel confident and comfortable with the sound level of the gears. I could run a 50v 20ah pack and get a 60amp current, it would be great.

Andrew.

 

[Lurkin]

i love what you have done and i have money to spend,

The kit is what it is and they are staying steel for the time being. Why not buy one, try it, and if its too loud either develop an alternative material gear or sell it?

 

[jk1]

Tangent will make double the torque of a cyclone.

Even if the torque curves where the same which they are not as the cyclone has higher torque at LOW RPM.

Say it was double torque at the cranks that doesn't mean their is double the torque at the wheels. For example the tangent kit only has one bike chainring whereas the Cyclone has two, hence if you use the smaller 24T chainwheel on the front of a cyclone you would have double the reduction to the rear wheel vs the tangent with 48T and hence double the torque at the wheel, which is where it counts not at the cranks. But the benefit is the cyclone would have twice the top speed potential and same torque potential due to the two front chain rings, this is assuming both motors have the exact same average power rating.

 

[tangentdave]

Herr Brumbrum, I'll get you your video. I want to show you the actual hardware I'm shipping, not just the prototype in the pics. The chips are flying as fast as I can make 'em. It's quite a mess. Also the new grease seals are only days away from arriving, then gotta make 'em fit, then gotta make sure it holds correctly. That means testing tho

-dave

 

[district9prawn]

Even if the torque curves where the same which they are not as the cyclone has higher torque at LOW RPM.

Compared to a 3210 the cyclone motor should be more powerful. It weighs 2kg compared to ~1kg for the astro. But the power of the cyclone is difficult to make use of if you plan to pedal along with the motor.

 

[Nathan]

2kg compared to ~1kg for the astro

The cyclone weight is more like the 3220 astro and hence power can go to more like the 3220 i.e 5 kw continuous with forced air cooling.

https://endless-sphere.com/forums/viewtopic.php?f=28&t=56666&p=845919&hilit=chupa#p845919

Since when does peddling make any difference with a high power bike above 2.5kw ? if you peddle or not will make no difference to the top speed or efficiency when using 2.5kw of power or more i.e peddle input is such a small percentage it will make no difference to speed or consumption, peddling is most useful on low power bikes or from take off and regardless of the motor-crank ratio all motors and cranks start from zero rpm so you provide input with 40:1 as well as 80:1. just that in the 40:1 case you have halved your top speed.

 

[recumpence]

2kg compared to ~1kg for the astro

The cyclone weight is more like the 3220 astro and hence power can go to more like the 3220 i.e 5 kw continuous with forced air cooling.

https://endless-sphere.com/forums/viewtopic.php?f=28&t=56666&p=845919&hilit=chupa#p845919

Since when does peddling make any difference with a high power bike above 2.5kw ? if you peddle or not will make no difference to the top speed or efficiency when using 2.5kw of power or more i.e peddle input is such a small percentage it will make no difference to speed or consumption, peddling is most useful on low power bikes or from take off and regardless of the motor-crank ratio all motors and cranks start from zero rpm so you provide input with 40:1 as well as 80:1. just that in the 40:1 case you have halved your top speed.

You have no idea what you are talking about. A 3220 is 92% efficient while the cycle is around 80%. That means a 3220 can produce over twice the power of the cyclone per pound of weight. I have run 8kw continuously through a 3220 without any problems (10 kw with fan cooling).

 

[brumbrum]

Question: Has this system actually been tried at over 35amp continuous? Is there any estimate how long it can sustain a higher amp load if it is rated at 4kw or 80amp peak input?

Andrew.

 

[aCeMadMod]

Question: Has this system actually been tried at over 35amp continuous? Is there any estimate how long it can sustain a higher amp load if it is rated at 4kw or 80amp peak input?

Andrew.

how long do you want me to test at 35 amp . it with 3220 motor

 

[Ralith]

I've run it around 35A for minutes at a time with the 3210 and things merely get warm. It's efficient enough (and my chainring is small enough) that it can be difficult to actually push the motor harder than that without a good hill.