Direct-drive 3220 setup with a 3-Speed Hub...

[grwsaltspring]

Hi Gary;

I must be a bit slow with how this all works together.

Would it be possible to post some pictures of the intermediate steps.

Kind of a slot A fits into slot B kind of thing for the assembly challenged amongst us.

Cheers Greg

 

[kfong]

Nice work Gary, thanks for the pictures; it was clear once I could visualize it. From my calculations. If I try your setup on a standard 26in wheel. My top speed should be 35mph if I run at 45volts. This seems like a good setup. If I go with 36v, I should be able to keep up with the motor on short sprints. Does this sound about right to you? Do you think the 3220 motor can handle a direct drive setup on a 26in wheel without overheating and without having to add some serious heatsinks? I know this motor has some serious power behind it, but I’m thinking that power is best when running close to no load speeds. For trail riding when the power is needed at low RPM’s. I’m thinking at least a 1st stage gearbox will be needed to keep the motor cool. Having a smaller tire is certainly easier on the motor. I would still like to stick with a 26” tire to keep the frame geometry intact. Interested in your inputs on this matter.

 

[GGoodrum]

Oh, I think the 3220 definitely has more than enough power for a 26" bike. Earlier, I got the bike back together, and went for a 3 mile test ride. Wow, what a difference not having the bike chain move, makes on performance. It runs smooth as silk now, and it seems to have much quicker acceleration. I got the front wheel to pop up, shifting from 2nd to 3rd. I tried goosing it in first, from a very slow "crawl", but I couldn't actually open it up all the way because it almost flipped me off the back. Frankly, this is a bit too much for me. Today I hit 116A and 5500W peaks, and after 15 minutes, or so, of tearing around our neighborhood, nothing got hot at all. The controller was 96F and the motor was 108F.

I also did a couple of speed runs, to see how close my spreadsheet predictions were, and near as I can tell, they are spot-on. 1st gear topped out at 20 mph, 2nd got up to 28 and I stopped at 35 mph in 3rd, as it was just to scary to go any faster. The spreadsheet says it should top out at 37 mph in 3rd, so I was close. Using a 26" wheel, the spreadsheet says the top speeds should be 27 mph, 36 mph and 48 mph, in 1st, 2nd and 3rd, respectively, assuming you use a 3-speed hub. This is also assuming you use the same 11t/94t gearing, and a 7-turn 3220, which has a kV of 97. I think on a 26" bike, this setup is much more suitable. I just can't see using 1st gear on this bike, going up any hill, as I'd be too afraid I'd end up on my butt.

I'll tell you the setup I'm dying to try, and that is this similar direct-drive configuration, but on a singlespeed beach cruiser. Instead of the 3-speed hub, however, I'd do delta-wye switching. I've got some 60A latching relays coming, that I will use with a circuit Richard did that will ensure these six SPST relays all switch properly, and quickly. Looking at the numbers, I think I'd probably use an 8-turn 3220, which has a kV in its normal "wye" mode of 84. In delta configuration, the kV goes up to 145. In wye mode, the top speed would be about 30 mph, and in delta it is about 53 mph. Actually, 50 is probably the max, as the wind drag really comes into play once you get above about 45. Anyway, this would be a great all-around setup, with gobs of low-end torque, and a top speed that would let any commuter keep up with traffic.

I took some more pics before it got dark, but they are still in my phone. I'll post them in a bit. Next up, I want to redo the setup in my other folding bike, which currently has a 3210 10-turn (kV: 113...) and a Nexus-3 hub, with an "extra" reduction stage. It has a #25 chain 1st stage reduction with 80t/24t and a regular bike chain 2nd stage reduction to the hub with 34t/12t. That's a total reduction of 9.4:1, which I can match with the direct drive using the same 94t big sprocket and a 10t motor sprocket.

-- Gary

 

[liveforphysics]

Excited for pics!

 

[GGoodrum]

Finally got them into my computer:

View attachment 4
View attachment 3
View attachment 2
View attachment 1



The added freewheel is a 17t, which makes the pedals a lot more useful. I can actually use them up to about 20 mph. With the old setup, the smallest I could use was a 22t.

-- Gary

 

[liveforphysics]

Finally got them into my computer:
-- Gary

Gary taps gallery on phone, taps photo, taps share photo, email to gary, done. At least 20 seconds of work, I'm glad you had time to get that done for us Gary. Just teasin you!

That setup looks so damn slick Gary! That was exactly what I wanted to do for my build, but I ended up pussing out on going through the hassle of mounting double freewheels on the rear hub. You're a good man for not folding to comprimise like I did. Now, this setup no longer requires the front crank set to freewheel, but does still require the pedaling to drag the motor chain, which you have freewheeling at the motor drive sprocket (right?).

Does this mean the ultimate ultimate setup would be a tripple freewheeling rear hub adapter? True isolation of both drive sources from each other. Miles? got your awesome CAD skills handy? Maybe before having 50 of the double freewheel adapters made, you should at least considder a tripple There is space for it to fit onto the splines of a normal 7-9spd rear hub IMO.

 

[Miles]

Does this mean the ultimate ultimate setup would be a tripple freewheeling rear hub adapter? True isolation of both drive sources from each other. Miles? got your awesome CAD skills handy? Maybe before having 50 of the double freewheel adapters made, you should at least considder a tripple There is space for it to fit onto the splines of a normal 7-9spd rear hub IMO.

Yes, this is possible, but not easy, with a standard 9 spd cassette. I have been giving it some thought... :wink:

 

[liveforphysics]

I've used the #25 chain, on the 1st stage of my other setup, but I had problems with keeping it tensioned properly. Compared to the #35 stuff, it is downright puny.

Size 25 chain is very wimpy. Like bicycle chain wimpy, and often of a lot lesser materials and manufactureing quality than bicycle chain.

However, with a single stage setup, for any desired torque value at the rear wheel, and any motor torque value to begin with, the strain on the chain will be much less than the loading on a dual stage drive. So, if you were trying to get away with running #25 chain on something, a single stage drive would be the best chance it's got at lasting.


The other option is #219 chain, which is also used on racing go karts. This is what D is using between Matt's drive and his front crank. This size is a bit smaller pitch (.308" vs .375"...), but every bit as robust as the #35 chain. There's several sources for big sprockets for these, also aluminum, but the problem is I could not find motor sprockets in this pitch that didn't have tapered bores, and/'or keyways, much less something with a 3/8" bore. I actually ordered and received some big sprockets and matching #219 chain, but I never solved the motor sprocket issue, so I switched over to #35 chain. Either way, I'm definitely glad I went this route. Originally, it was just to get a large enough ratio to make the direct-drive work, but I think the heavy-duty nature of this size is a very big plus, especially given your demonstrated tendency to break things. :lol: For your buddy, Luke, I think only tank tracks will survive.

Amazingly, I've never once broken #219 chain in my life. It has the advantage of being designed around racing, so it's all made with top quality metals and processes. I went with #35 chain on my personal bike build for the same reasons you did Gary. #219 is a better chain, but selection was limited to kart specific sprockets. if you are going to have custom sprockets made, no reason not to go for #219 chain.


-- Gary

 

[liveforphysics]

Does this mean the ultimate ultimate setup would be a tripple freewheeling rear hub adapter? True isolation of both drive sources from each other. Miles? got your awesome CAD skills handy? Maybe before having 50 of the double freewheel adapters made, you should at least considder a tripple There is space for it to fit onto the splines of a normal 7-9spd rear hub IMO.

Yes, this is possible, but not easy, with a standard 9 spd cassette. I have been giving it some thought... :wink:

Awesome Miles You're the man! And you know, if you could make that, it would be the holy grail for all those guys doing gas weed-eater engine powered bikes, as well as all the RC bike guys, and various other applications.

 

[Miles]

Agreed #219 is pretty much the ideal size for our use. It has the advantage of the Extron composite sprockets, too. There's just the problem of conveniently adaptable drive sprockets.

 

[Miles]

Does this mean the ultimate ultimate setup would be a tripple freewheeling rear hub adapter? True isolation of both drive sources from each other. Miles? got your awesome CAD skills handy? Maybe before having 50 of the double freewheel adapters made, you should at least considder a tripple There is space for it to fit onto the splines of a normal 7-9spd rear hub IMO.

Yes, this is possible, but not easy, with a standard 9 spd cassette. I have been giving it some thought... :wink:

Awesome Miles You're the man!

I'll start a new thread to discuss this........

 

[liveforphysics]

Awesome Miles You're the man!

I'll start a new thread to discuss this........

You don't think it would seem a bit vain to start a new thread just to discuss about you being the man?

LOL :lol:

 

[Miles]

:lol: :lol: :lol: :lol: :lol: :lol: :lol:

http://www.endless-sphere.com/forums/viewtopic.php?f=28&t=12449&p=185209#p185209

 

[kfong]

Thanks Gary, that confirms that it makes good solution. Does your machinist have a source for the tap for the threads, or did he just use a lathe. Trying to get a second freewheel and 9 speed cassette is going to need some close fitting custom parts.

 

[GGoodrum]

Thanks Gary, that confirms that it makes good solution. Does your machinist have a source for the tap for the threads, or did he just use a lathe. Trying to get a second freewheel and 9 speed cassette is going to need some close fitting custom parts.

I think he just used a lathe, or one of his CNC machines. He had to order a special bit, though.

Regarding the triple FW idea, I agree that would be optimum, but no way it would fit with an internal hub. I barely got this to work by cutting off the four removal tabs. That might not be necessary with the Nexus-3, but I don't know yet, until I get my Mariner apart today. With the motor off, pedaling does turn the motor and it doesn't take much for the controller to come alive and start beeping. It is pretty easy to just give enough throttle to unload the motor, and the pedaling is easy.

<off to Miles' new thread...>

 

[drewjet]

Gary, can you start from a dead stop, no peddling? If so do you have to ease into it, or can you gun it?

 

[methods]

Nice work Gary, thanks for the pictures; it was clear once I could visualize it. From my calculations. If I try your setup on a standard 26in wheel. My top speed should be 35mph if I run at 45volts. This seems like a good setup. If I go with 36v, I should be able to keep up with the motor on short sprints. Does this sound about right to you? Do you think the 3220 motor can handle a direct drive setup on a 26in wheel without overheating and without having to add some serious heatsinks? I know this motor has some serious power behind it, but I’m thinking that power is best when running close to no load speeds. For trail riding when the power is needed at low RPM’s. I’m thinking at least a 1st stage gearbox will be needed to keep the motor cool. Having a smaller tire is certainly easier on the motor. I would still like to stick with a 26” tire to keep the frame geometry intact. Interested in your inputs on this matter.

I am surprised that nobody caught this.
I have bad news for you friend - it is not quite that easy

The reason everyone wants to run motors like the astro is because they achieve very high efficiency - in this case >90% at 7,500RPM
What sucks is figuring out a way to harness such a high rpm on a slow wheel (hence all the huge gears and fancy drives)
If you lower the speed of the motor by running a lower voltage (and most likely a higher current to reach similar power levels) you will be running the motor in a region of sub optimal efficiency.

I cant quote a number - but it would be more like 70% - not the really high numbers that allow people to "run hard and not get hot". So if you tried to buzz around on your bike at a few KW at those RPM, not only would you loose a lot of power to eff, but that lost power would become heat - and eventually - your limiting factor.

When choosing the motor you have to gear it such that you keep the motor spinning as fast as possible - or at least close to the targeted top speed of 7500rpm.

The 3220 comes in flavors suitable to 36V like the 2turn or 3 turn (at something like 150A ) but then your top speed would end up being the same top speed Gary would have if he ran 26" wheels - which would be quite fast - and you could end up "lugging" the motor around in an inefficient zone at normal riding speeds.


-methods

 

[MitchJi]

Hi Gary,

I barely got this to work by cutting off the four removal tabs. That might not be necessary with the Nexus-3, but I don't know yet, until I get my Mariner apart today.

Do you know if it will work with a Nexus 8?

 

[GGoodrum]

I'm pretty sure the Nexus-8 will work, but I will check. There's a Nexus-8 on my wife's Townie.

 

[GGoodrum]

Here's another video I just did, with my iPhone zip-tied to the front reflected:
[youtube]IWVJUtFP-bs[/youtube]


The shaking you see is pretty much how I see things riding it. I had problem with the shifter adjustment for the hub, which caused it to want shift on it's own. I've since fixed that. At the end, going up the driveway, the frontend popped up, and I came off the back and hit the big sprocket with my shoe.

I topped out at 21, 28 and about 34, although it seemed to still have a bit more left. I won't ride this above 30 anymore. Too scary for me!

It is still a bit noisier than I'd like, so I'm going swap out the 11t motor pinion for a 13t this week. That should quiet it done some more. It sounds pretty cool when it shifts, though.

-- Gary

 

[oatnet]

. At the end, going up the driveway, the frontend popped up, and I came off the back and hit the big sprocket with my shoe.

The dramatic ending <grin>

It sounds much quieter. Maybe I'll mess with one of these after my BMC V3 experiment.

-JD

 

[kfong]

Thanks Methods, kinda thought it was too good to be true. That's why my inquiry on the no load speeds. Looks like a single stage on his setup would still be needed to keep the motor from overheating.

Nice work Gary, thanks for the pictures; it was clear once I could visualize it. From my calculations. If I try your setup on a standard 26in wheel. My top speed should be 35mph if I run at 45volts. This seems like a good setup. If I go with 36v, I should be able to keep up with the motor on short sprints. Does this sound about right to you? Do you think the 3220 motor can handle a direct drive setup on a 26in wheel without overheating and without having to add some serious heatsinks? I know this motor has some serious power behind it, but I’m thinking that power is best when running close to no load speeds. For trail riding when the power is needed at low RPM’s. I’m thinking at least a 1st stage gearbox will be needed to keep the motor cool. Having a smaller tire is certainly easier on the motor. I would still like to stick with a 26” tire to keep the frame geometry intact. Interested in your inputs on this matter.

I am surprised that nobody caught this.
I have bad news for you friend - it is not quite that easy

The reason everyone wants to run motors like the astro is because they achieve very high efficiency - in this case >90% at 7,500RPM
What sucks is figuring out a way to harness such a high rpm on a slow wheel (hence all the huge gears and fancy drives)
If you lower the speed of the motor by running a lower voltage (and most likely a higher current to reach similar power levels) you will be running the motor in a region of sub optimal efficiency.

I cant quote a number - but it would be more like 70% - not the really high numbers that allow people to "run hard and not get hot". So if you tried to buzz around on your bike at a few KW at those RPM, not only would you loose a lot of power to eff, but that lost power would become heat - and eventually - your limiting factor.

When choosing the motor you have to gear it such that you keep the motor spinning as fast as possible - or at least close to the targeted top speed of 7500rpm.

The 3220 comes in flavors suitable to 36V like the 2turn or 3 turn (at something like 150A ) but then your top speed would end up being the same top speed Gary would have if he ran 26" wheels - which would be quite fast - and you could end up "lugging" the motor around in an inefficient zone at normal riding speeds.


-methods

 

[John in CR]

Gary,

Yes, the shifting is cool. How much noise does the hub itself make?

I found a place locally that has a few old stock Nexus3's. I say old stock because the price was almost silly at under $30, so I picked one up for my current build, and will go buy all they have if it works out. I just love places here that don't adjust their inventory prices with the steadily increasing exchange rate against the dollar, because when you find something they established a price for a few years ago, it's possible to pick up dirt cheap. Typically prices are high across the board due to high import duties + 13% sales tax, which includes shipping costs in the tax base for online purchases. The surprise bargains are rare but pleasant to find.

John

 

[liveforphysics]

John- Those are awesome prices on Nexus 3's! I would buy all of those in a second.


Gary- The sound of your bike shifting is so cool! I MUST have a geared hub on my bike! That is just too cool! Looks like your bike pulls pretty hard! When you let Astro Bob ride your bike, you gotta give him full power, and just let him flip over if he can't control the throttle. If you mounted your battery off the front of the bike, it would look very goofy, but it would do a lot to keep the front tire down. Great job! Awesome bike! I bet it's nuts to do 34mph on a tiny wheel folding bike with a 2ft long handle bar riser! Not exactly something designed for speed , but that just makes it that much more exciting .

 

[GGoodrum]

John, I too would definitely grab all of those Nexus-3 hubs. That is at least half-price. I love them, and I think they might actually be a little "tighter" than the S-A hub I'm using. There seems to be a bit of "catch up", or slop in starting out with the SRF-3, that I didn't notice with the Nexus-3 on my other bike. You can hear it when I start out. It sounds like letting a clutch out. Anyway, I'll check the Nexus on the other bike when I get it back together. I'm in the process of removing the jackshaft and will set this one up the same way, except that it has a 3210-12t (kV: 113...), which is a much better fit for a bike this size, in my opinion. In any case, it seems to shift just as smoothly.

I think all the noise is basically coming from the #35 chaindrive. The hub itself is pretty quiet.