There's a new sheriff in town (new pics on p5)

[John in CR]

Mighty Volt,

Excellent. Run the factory controller in stock form before modding it for higher power, so you can get a baseline performance. Be sure to include actively ventilating it in your mods. That should enable you up to fun levels, though not eye popping with those fets. 60V nominal if the caps are up to it, and double the current by coppering half the shunt length should be fun. I can't wait to see how you use it.

After blowing SuperV's big controller I was forced to do an inventory of blown controllers and find one I thought I could repair myself to put on my cargo bike, freeing up that 24fet4100 brick to install on SuperV. Heat damage, other than blown fets accounted for almost half of the failures, and 3 out of 5 of the modded factory controllers. Mostly it was melted phase wires that either shorted to each other or to caps. From now on I will ventilate any controller I take to higher power, including my one remaining 24fet that I'm going to dial up to 150A battery side in gradual increments.

Blown controllers has been the achilles heel of just about everyone's high power pursuits, but the only ventilated controller I've blown was my last remaining factory controller that I was pushing past triple the factory settings, and that occurred on the first hill I encountered after forgetting to turn the fan on. Ventilation works well, and we've been pushing a 4310 based controller at 80A for months as proof, though we really should get in there and beef up the primary wires and traces to minimize unnecessary heat, since the air blowing out of that controller is quite warm.

I successfully repaired one of my factory controllers, YAY, so I'm moving it to my daily rider which isn't fun anymore thanks to the Sheriff. It's will still get 5kw via the new and improved ventilated stock controllers. The 24fet will get dual blowers, doubled up wiring, and thickened traces. It's been dependable at 100A, so with mods, 150A should be no problemo.

I'll do a post soon about ventilated controllers.

John

[The Mighty Volt]

Mighty Volt,

Excellent. Run the factory controller in stock form before modding it for higher power, so you can get a baseline performance. Be sure to include actively ventilating it in your mods. That should enable you up to fun levels, though not eye popping with those fets. 60V nominal if the caps are up to it, and double the current by coppering half the shunt length should be fun. I can't wait to see how you use it.

After blowing SuperV's big controller I was forced to do an inventory of blown controllers and find one I thought I could repair myself to put on my cargo bike, freeing up that 24fet4100 brick to install on SuperV. Heat damage, other than blown fets accounted for almost half of the failures, and 3 out of 5 of the modded factory controllers. Mostly it was melted phase wires that either shorted to each other or to caps. From now on I will ventilate any controller I take to higher power, including my one remaining 24fet that I'm going to dial up to 150A battery side in gradual increments.

Blown controllers has been the achilles heel of just about everyone's high power pursuits, but the only ventilated controller I've blown was my last remaining factory controller that I was pushing past triple the factory settings, and that occurred on the first hill I encountered after forgetting to turn the fan on. Ventilation works well, and we've been pushing a 4310 based controller at 80A for months as proof, though we really should get in there and beef up the primary wires and traces to minimize unnecessary heat, since the air blowing out of that controller is quite warm.

I successfully repaired one of my factory controllers, YAY, so I'm moving it to my daily rider which isn't fun anymore thanks to the Sheriff. It's will still get 5kw via the new and improved ventilated stock controllers. The 24fet will get dual blowers, doubled up wiring, and thickened traces. It's been dependable at 100A, so with mods, 150A should be no problemo.

I'll do a post soon about ventilated controllers.

John

Thanks John.

Because the controller came with no throttle, I hooked up a 24v Ananda controller to it and immediately encountered some grinding action. The motor was drawing 15A under no load.

I test the motor with Lyens controller and there appeared to be no phase or hall issues.

Could it be that the motor will only work with the stock controller?

Unfortunately the stock controller has only a few components I can recognize, 3 phase wires, 2 power wires and the halls/throttles.

Can I put a big enough tyre on this motor to make sure it comes out at around the same circumference as a 20" BMX front wheel?

Thanks again for your support and encouragement. Ultimately I want to run this baby at 72v.

[John in CR]

Grinding action, just wrong wiring combo I hope. Until you are certain about correct wiring, only try small throttle pulses.

Does the axle turn with grinding, or only cogging resistance?

Don't you have a working ebike to just borrow the throttle from?

How are you doing a run test, clamping axle flats in a vise with some flat stock protecting the axle from the vise's teeth? Be careful, it ain't no puny ebike motor.

Tires, I would think so, but I'm no expert.

[The Mighty Volt]

Grinding action, just wrong wiring combo I hope. Until you are certain about correct wiring, only try small throttle pulses.

Does the axle turn with grinding, or only cogging resistance?

Don't you have a working ebike to just borrow the throttle from?

How are you doing a run test, clamping axle flats in a vise with some flat stock protecting the axle from the vise's teeth? Be careful, it ain't no puny ebike motor.

Tires, I would think so, but I'm no expert.

Hi John, the flat of the axle is to the flat of the vise, the teeth are OK, but I might wrap the axle just in case.

With no controller, just with a part of the axle dropped into a clamp/vise......it spins smoothly and freely by hand, no issues at all, either direction. If I spin it fast enough by hand, I get that reassuring motor "whine" sound.

I do have a working ebike to borrow a throttle from, but I don't use connectors, I always solder my wires together, braid them, solder them, heatshrink them, that way there is no issue with dodgy connections or water getting into them. Salvaging a throttle now would be a PITA, but I could do it, sure.

I might just strip off the Ananda twist-throttle, as long as it is hall-effect, thats fine.

[John in CR]

There are ways to hot wire the throttle wires, but I don't know of an easy one to give you good throttle resolution, and the last thing you want is a barely held motor and hitting it with big chunks of throttle trying to find the correct wiring combo.

I've posted many times the easy way to find the correct wiring, so please don't go the hard route. As long as the phase angle is correct or you have a controller that auto senses the phase angle (your motor tester tells you the phase angle of the motor halls.) Then you prove the motor works within 6 attempts, actually 5 more since you know the current one is incorrect. You'll get a good forward or reverse simply by trying either the other 5 phase combos without changing the halls, or the other 5 hall combos without changing the phases, whichever is easier. Do not change both.

You may not even decide yet which way is forward, since you might want to put the drum brake on the left. I've done it both ways, and the only trick about putting it on the left is that you have to pull the lever the clockwise. At least mine are definitely directional and if you pull the wrong way, yes the brake shoes still engage, but it has very little braking power.

[The Mighty Volt]

GOT IT!!!

It actually snaps the unsecured clamp around on the workbench like it was a bit of wood or foam, not like the cast-iron that it is. Not even the 5305 was doing that at 72v.

Damn it is torquey on just 55v!!!!!!

Like a Lyen, it has a red wire which needs wiring to the controller positive.

My cheapo pedelec Ananda throttle has a massive flat-spot in the first 1/3rd of the twist...then bang.

Incredible wallop and RPM on just 55v.

I got the motor and controller for around $250, all in.

Now as soon as i can I will upload some photos for identification of the remaining controller terminals.

[John in CR]

Like a Lyen, it has a red wire which needs wiring to the controller positive.

After routing it up and back thru a keyswitch within easy reach while riding.

Glad to hear everything works. Now you too can laugh at X5's, but at less than half the price. We expect to see some videos of you tearing it up in Ireland. Showing off the capabilities of a high power ebike is one of the best things we can do to forward the cause, and as long as we're extra courteous to cars and peds there's no downside.

[The Mighty Volt]

COurteous ?

When I am on the trails on an X5305 BMX, the people on horses take pity on me, and offer to let me pass.

I accept their invitation, and promptly disappear up a hill at a rate of knots.

Have to say that there are controllers, and there are controllers.

At 55v, and just jerking the throttle, the controller does seem to be a bit warm

I'm not sure the stock controller is worth a damn, and I would LURVE to see this motor at 72v on 4110 Fet's

What sort of motors are these anyways? What sort of windings do they have, do they burn out controllers at high-Amps or are they kind like a 5305??

[Harold in CR]

$250.00 all in, EH ??? That include shipping and other Govt fees ??? Is yours the 72V model ??

Looks like I need to sell some more wood.

[The Mighty Volt]

$250.00 all in, EH ??? That include shipping and other Govt fees ??? Is yours the 72V model ??

Looks like I need to sell some more wood.

No Government fees, as it was England to Ireland.

Shipping was a cheapo $25 2 day DHL special.

Needless to say they dropped it and now there is a flat-spot on the rim which we need to get out, we think we can, without too much effort.

The Hub was advertised as a 48v 1500w model, so it's not the super powerful version like JohninCr's, but at 55v it is demonstrating some pretty sick power.

I got the motor and controller off a guy who seems to have some left-overs, replacements or something and is selling them off.

Price does seem to be a bit low to be true but I got mine and hooked it up and away I went, eventually.

My brother is the fabrication king, he will have a look at it tomorrow and decide where we go. with it.

[The Mighty Volt]

Not to thread crap, but unless you got some hard data, I will call this bullshit for now. No doubt it looks like.a clean build, but other than that, I don't see what is special.

mvly = f*cking retard.

This is like a hubmotor for a car fitted to a bicycle. This should make for the most powerful hubmotor bicycle that has ever roamed the streets.

For perspective, a 9C has 169.6cm^2 of active flux area.

Johns new motor has 423.9cm^2 of active flux area, and that area is on a 35% longer moment arm.


In the worst case it should do the output of about about 3.3 of the 9C motors in that single motor, and since a substantial amount of the resistive heat loss in a 9C is from the end-turns, and you've got 3.3x 9C motors worth of active magnetic material with roughly 1.35x the increase in end-turn loss over a single 9c, I'm predicting it will be more like 4-5 9c's worth of motor potential in that back wheel.

And just a single 9C can rip pretty damn hard when driven right!


This thing should fly!

Howdy Physics.

I was wondering.....I have a 48v 1500w rated one of these which John recommended i pick up for the price being asked. I would like to try and open it up with one of those gear pullers, what should I expect to find.....copper-wise, stator-wise, wire-wise etc. I noticed immediately that it has phase-wires which are as big as the stock X5 if not a little bigger.

What gives here.......if the 9C's and the X5's are so inferior, why are they selling? Is it a simple matter of convenience, and those motors fulfilling the daily needs of the vast majority of riders?

I should have thought that this only explained part of the issue.....by nature, the E-bike and related communities were hacker-dominated, modder-dominated communities,, where only the best, or latest, or most-over-volted stuff would do, and where generic,stock stuff was discarded.

Yet it seems that JohnInCr is the only guy rocking a Moped build, and you were the only guy with a Lynch motor....both of which are the flagship motors of their class {HUB/CHain}

[The Mighty Volt]

Sent an email request for more info off last night to a supplier in China. They got back to me......they offer a range of Moped Motors. The one i have is for sale at $190 new, so I got mine cheap alright, especially saved on the postage. They also offer a 20Kg, 13" aluminium bodied hub, rated for 8000w. Max torque is described as 200Nm and efficiency is claimed as being _> 85% with RPM's varying from 1000 to 1800 depending on voltage {72/84/96}. That's the baddest one they have.

Price? $585 FOB.

[John in CR]

Imagine if you push that 44 pounder to it's limits. My 1500W motors handle 10kw peaks without issue as long as you avoid getting bogged down on hills. If you're willing to accept the weight, the big issue with these scooter motors is the controllers.

[The Mighty Volt]

Imagine if you push that 44 pounder to it's limits. My 1500W motors handle 10kw peaks without issue as long as you avoid getting bogged down on hills. If you're willing to accept the weight, the big issue with these scooter motors is the controllers.

Well the small issue is the battery to feed that sucker!!! And yes, the controller now pretty much becomes the weak link.

You can buy 24 Fet controllers, base models, for around $100 delivered. The trick then is to get the best FET's into them. What was failing in your controller? Was it just the heat from the cheap FET's?

[Harold in CR]

Did we lose a bunch of photos with the latest Site updates ??? I find many "File does not exist any longer" notices when I click on a download to see a photo ????????

[John in CR]

My main hubbies, comparable to the new Xlytes, but with a 40mm stator and a higher Kv at 16rpm/V even blew the super Methods controller, 18 by 4110fets and beefed up everything, the 2 days after LFP left last year. It's the current limiting that kills them staring down the barrel of the low resistance windings that need almost 30mph in a 20" rim before BEMF gets strong enough to help control things.

I don't know what's wrong with Hubmonster's factory controller. I thought I could get away with doubling the current with a 200lb lighter vehicle and the controller much more exposed to the wind, but it couldn't handle it and I wasn't even riding it near max performance.

Everyone seems to have trouble once you cross the 100A barrier, and further evidence is the Chinese going to monster size motors so they don't have to cross it. Arlo1 has had the best luck pushing them well beyond that, but that's with an X5304 which is a far easier load. I have an X5304 that was running fairly conservatively, and I took the controller off of it to run on one of my main hubbies in a 20". It took off way stronger than the X5 was running, but the controller blew the first time I gave it full throttle, a result that still baffles me.

The Sevcon's Luke has for brushless are apparently up to the task, but the setup is a mountain I don't have the tools to cross.

I've had luck with my Lyen 24fet 4110 controller...realiable since Sep of last year, but I did tone the settings down to 80A after running at 110A for a little while to make sure it stayed reliable. I beefed up the traces, doubled up the 10ga phase wires, and halved the shunt resistance to double the regen current. I was planning to add to blowers to bring in fresh air and keep the insides cool as well as almost double the surface area for FET cooling, all with plans to run Hubmonster with it, but now I need to put it back on my daily rider since I'm out of controllers. Blue is down for a quick overhaul (better batts, new wiring harness and bearings, and some beautification) while we're in the meat of rainy season, but I'll keep it conservative until I get another bike going.

I thought I was done with controller problems after learning hard lessons about current limiting, partial throttle, and hills. A year of zero problems really spoiled me, but now that I'm reaching for more performance controller popping is back and it sucks.

[John in CR]

Did we lose a bunch of photos with the latest Site updates ??? I find many "File does not exist any longer" notices when I click on a download to see a photo ????????

If any of mine are missing, I have backups this time. I'll wait until they finish trying to recover the database before going to the trouble.

[Harold in CR]

Most of the ones on pg 5 are missing. Didn't go through the whole thread.

[The Mighty Volt]

Cheers John for that breakdown.

What I don't understand is this.......why do the Chinese have an 8000w rated hub....which eats controllers???

I mean....this thing has an aluminium body and weighs over 40lbs. Thats a lot of copper and magnets. Why bother...if it cannot be properly fed.

What are they using them for? High voltage, low-amperage applications....???

What does 50A feel like in one of these things???

Cheers.

[Arlo1]

My main hubbies, comparable to the new Xlytes, but with a 40mm stator and a higher Kv at 16rpm/V even blew the super Methods controller, 18 by 4110fets and beefed up everything, the 2 days after LFP left last year. It's the current limiting that kills them staring down the barrel of the low resistance windings that need almost 30mph in a 20" rim before BEMF gets strong enough to help control things.

I don't know what's wrong with Hubmonster's factory controller. I thought I could get away with doubling the current with a 200lb lighter vehicle and the controller much more exposed to the wind, but it couldn't handle it and I wasn't even riding it near max performance.

Everyone seems to have trouble once you cross the 100A barrier, and further evidence is the Chinese going to monster size motors so they don't have to cross it. Arlo1 has had the best luck pushing them well beyond that, but that's with an X5304 which is a far easier load. I have an X5304 that was running fairly conservatively, and I took the controller off of it to run on one of my main hubbies in a 20". It took off way stronger than the X5 was running, but the controller blew the first time I gave it full throttle, a result that still baffles me.

The Sevcon's Luke has for brushless are apparently up to the task, but the setup is a mountain I don't have the tools to cross.

I've had luck with my Lyen 24fet 4110 controller...realiable since Sep of last year, but I did tone the settings down to 80A after running at 110A for a little while to make sure it stayed reliable. I beefed up the traces, doubled up the 10ga phase wires, and halved the shunt resistance to double the regen current. I was planning to add to blowers to bring in fresh air and keep the insides cool as well as almost double the surface area for FET cooling, all with plans to run Hubmonster with it, but now I need to put it back on my daily rider since I'm out of controllers. Blue is down for a quick overhaul (better batts, new wiring harness and bearings, and some beautification) while we're in the meat of rainy season, but I'll keep it conservative until I get another bike going.

I thought I was done with controller problems after learning hard lessons about current limiting, partial throttle, and hills. A year of zero problems really spoiled me, but now that I'm reaching for more performance controller popping is back and it sucks.

John I have been able to run over 250 amps through my x5 during acceleration and the spike in my sig is from my X5 24fet powered bmx too! Note anything past ~200-250amps in a X5 just turns into heat and does not make it any faster infact it makes it slower because of heat! But thats an X5304 it has more inductance then the hs35 motor! Inductance is the major key to hi amps what it does is slow the time it takes for the amprage to increas this is what saves my controller.
On that note I have to double check but I think the x5 was ~260uH here is a spread sheet that shows inductance of collossus (motor A) and the X5 (motor B) And it shows how long it takes for the current (amps) to build which is a good thing in the x5 because it gives the controller time to control it with the PWM!
The X5 actualy has 20x the resistance but I put it down to the same resistance as collossus because resistance doesnt effect it as much.

[John in CR]

Beautiful info Arlo1. I haven't been following your work on inductance as closely as I should. Did you measure inductance or was it easily calculated. So it's the inductance that's been killing my controllers, since I'm sure mine are even lower than the HS motors. Have you been able to source any big cores to wind our own toroidal choke coils?

I've got enough magnetite for a few, and it's easy and fun for the kids to collect with a magnet at the beach. I was thinking epoxy+magnetite doughnuts to wrap my phase wires through could be a cool DIY low loss, low force solution. What do you think...worth a try?

[Arlo1]

Beautiful info Arlo1. I haven't been following your work on inductance as closely as I should. Did you measure inductance or was it easily calculated. So it's the inductance that's been killing my controllers, since I'm sure mine are even lower than the HS motors. Have you been able to source any big cores to wind our own toroidal choke coils?

I've got enough magnetite for a few, and it's easy and fun for the kids to collect with a magnet at the beach. I was thinking epoxy+magnetite doughnuts to wrap my phase wires through could be a cool DIY low loss, low force solution. What do you think...worth a try?

I got 2 cheep meters from ebay I just made sure they ready to 1uh or lower. I did get a set of proper inductors from another member that were 27uH each and when you put them in series with each phase they get added together. So if one phase has 8uH and you put 1 27uH inductor inseries with each phase you get 27 + 8 + 27 for 64uH per phase! That helps a lot it will retard the timeing though because it slow the current build up but also doesnt quit flowing current instantly they basicaly take as long to build amps when the phase is energised as it takes for the amps to quit flowing after the fets are sut off!

But first things first try to mesure yours.... I mesured a small collossus and its 48uh with a 75kv rating and the turnigys are probably lower with the lower turn count for higher kv. I just chaged my rewound collossus to WYE and it is now 25-26uH with only lowers the kv by 1.73 times so I get 4x the inductance and only loose 42% rpm. I am going to try to make it up with more voltage! As a rule you need ~100uH or more for a controller to be happy!

[John in CR]

Just to share what Arlo and I were discussing by PM in case anyone else has input, I think I may have an easy solution to the low inductance motor problems making it so hard on controllers. If 100uH is enough inductance then this is a piece of cake using toroidal inductors. The problem is finding one that can handle our high phase currents. My idea is to cannibalize 3 speaker drivers to get the big ferrite ring magnet out. Then heat it up past the Currie temperature of 480°C (860°F) to demagnetize it leaving a perfect large ferrite core. I found a toroidal inductor calculator online at http://lcbsystems.com/InduCalc.html, and while I'm not clear on the materials, it should take only 10 turns or less for 100uH. If that's true then the only remaining issue is whether we'd be hitting core saturation at our current levels in the hundreds of amps, which I hope isn't the case, but I can solve by robbing magnets out of larger cheapie speakers, which I have.


Questions:

Flux saturation- Will it manifest as heat, so we can easily identify it? Is it as unlikely as I think due to the large cores, the circular shape preventing leakage, and the relatively low inductance we're looking at?

Might we be able to get away with a single core by winding all 3 phases on one?

Am I correct in my understanding that there's no added controller risk by trying this approach?

Am I missing the boat wildly somehow?

Will the added inductance hurt motor performance noticeably?


A cheap and easy way to make hard to drive motors easier on controllers is the holy grail to anyone in this hobby chasing performance. I think it was Fechter who suggested coils on phases to me first nearly 3 years ago. I just never could come up with a suitable coil, and couldn't find a calculator to rea;ize how few turns are needed. Most of us have seen these coils in a smaller form, probably just as a filter in many electronics devices, in the form of a small metal ring with a wire wrapped through it a few times.

John

[Arlo1]

John you dont need 100uH on each inductor. You need <50uH each because you will have 2 in series pluss your motor which has some, and they add up when in series!

[The Mighty Volt]

John, just curious, why is there a need for 150Amps???? Are you just pushing the limits for the hell of it, or is there a need to run those suckers at those amps, or do they just draw that sort of amperage of their own accord once you apply a pertinent voltage at WOT?