EEVblog 1485 - PedalCell CadenceX Bike Generator LOL FAIL!

[@Vishmoney15]

Huge fan of your channel! How I learned all of my EE fundamentals. I’m actually the engineer that designed the electronics. Would love to chat more

[@EEVblog]

Wow, small world. Did you work for Pedalcell or did you do it as a contractor?

[@EEVblog]

@@Vishmoney15 Thanks, but sorry, not interested in reviewing such a product. It's better to simply strap a small USB battery bank to your bike with a bit of velcro. Is there any notification on your website advising people of the bad batch and replacements?

[@uni-byte]

@@EEVblog Wholeheartedly agree. I use a dual LED helmet light on my enduro motorcycle that has a small Li battery that straps to my upper arm. Lasts 3 hours on medium power, which is plenty for a dirt bike in the woods late at night doing race speeds. On a bicycle you could get away with low power and get 5-6 hours. Simple as can be and no installation required. It charges up over USB overnight ready for the next night adventure.👍

[@gumbi79]

@@Vishmoney15 lol covid such an easy excuse for lazy work and shit QA/QC

[@radman999]

@@EEVblog Looks like radio silence from here on in. Always a hallmark of a great company.

[@ashleywest7443]

It's a sensored brushless DC motor being used as a generator. They are commonly used in upmarket RC cars as the sensors give good starting torque. Chinese made ones are known for losing their magnets. We use unsensored ones in RC planes and were in the habit of pulling them apart when new and super gluing the magnets in properly before first use.

[@KerbalLauncher]

A lot of skateboard motor sellers advertise "battle-hardened" where they go in and add extra epoxy to the magnets. Some of them go even further and try to balance the motor.

[@preddy09]

Generally in RC prop type planes/drones, outrunner motors are used where even basic glue will work due to magnets being on the outside. Where inrunners are used such as EDFs, helis and cars, the in running stator with magnets are wrapped with kevlar or carbon fiber as any glue alone isn't a guarantee.

[@debochch]

If it is brushless DC then why is the output AC?

[@shazam6274]

@@debochch No such thing as a "Brushless DC motor" they all operate from 3 phase AC. It is "market speak" for the fact that the BLDC motors have a built in multiphase AC generator allowing the motor to operate from a DC power source.

[@mrlazda]

@@debochch bacouse brushless DC motors are not DC motors at all they are AC motors with electronic computation, or more precisely they are permanent magnet synchronous motor with trapezoidal back emf. Brushed DC motors are also AC motors but with mechanical computation (brushes convert DC to AC). There is only one motor as I know that is realy DC motor but it produce so little torque that is near useless.

[@HwAoRrDk]

I noticed from the previous Mailbag video the "do's and don'ts" sheet included in the box said "don't ride over 40 mph". I guess they already realised that excessive speed was a failure mode for the generator! But simply telling your users not to ride too fast isn't really a proper solution to poor manufacturing...

[@MidnightSt]

wtf

[@volvo09]

Yep! I have no idea about this brand but that "generator" looks horrendous for even low speed reliability.

[@TheKnobCalledTone.]

Unless you're on enough drugs to participate in the Tour De France, there's no way any normal person will ever exceed 40mph on a pushbike.

[@EEVblog]

@@TheKnobCalledTone. Bikes can go down hill.

[@RK-kn1ud]

@@TheKnobCalledTone. I've done 40mph down a hill on a $900 road bike. I am not a super athlete either. I bet I could approach 45mph if I had a bigger hill available.

[@MrSnoots]

"someone skimped at the glue factory" made me laugh out loud. I just imagine someone at an actual glue factory being pissed because someone at a janky motor/generator factory did a shit job.

[@xjet]

That's what's called an "inrunner" configuration (where the magnets are on the rotating shaft). They can fly apart as this one did. The other configuration (where the magnets are on a rotating bell and the coils are mounted in the middle) is called an "outrunner". The benefits of the outrunner include the fact that the magnets are being flung in-place by centrifugal force when things are turning. They still throw magnets sometimes however -- if they're poorly glued.

[@Pillowcase]

Yup, I've glued back many a magnet in the outrunner motors of my racing drones. They still worked quite well, especially after rebalancing.

[@Guineh76]

I was wondering why they didn’t use an out runner design. In runner seems more prone to flinging it’s guts out. Eh, not a motor or generator designer anyway. Seems like something that would really only be usable in low rpm situations.

[@Sonny_McMacsson]

@@Guineh76 Or you can put a proper mounting system instead of just gluing it like a cheap POS and it would never be an issue.

[@xponen]

​@@Guineh76 in out-runner design the outer casing have to free-spin and only attached to the static part only by a single shaft on a bearing, while a in-runner design the mechanical configuration is typical, ie; like in the video, a load can be attached to the motor by a shaft while the casing is rigid and can be securely mounted to a larger structure. So imagine how to design a plastic mounting for a spinning casing (out-runner)? on a DC fan it is done by making the fan as the casing for magnets and the base of the motor is mechanically supported by a circuit board.

[@WacKEDmaN]

imo the better design is with the windings on the armature and the magnets fixed...altho this requires a commutator and brushes to get the power off the armature and out to the housing.. and that can fail (not as quick as this junk tho!)

[@I.____.....__...__]

Well that's a shame, and a waste. The rest of it was pretty well-made, then they blew it with the very first, most fundamental, integral part in the most bone-headed way. There were so many ways they could have done that, why they went with a drop of glue, is beyond me. 🤦

[@psionicxxx]

Take into consideration that in China, one manufacturer makes PCBs, the other one makes the motors, and the last one assembles everything. Maybe it's how it is done in other countries too, except that other countries have QC department, and are not using the absolute cheapest to bring down the prices...

[@conwaylai8562]

When I was working in a tape drive manufacturer, we were surprised that the failure from the drive unit were so high. Went down to the factory in China, and after some auditing; the operators were not curing the loctite glue under the UV lamp (they just dab the glue and move it along). Apparently, no one understood the process instructions, so that part was not translated to mandarin for the operators.

[@uploadJ]

PLUS no one had seen or used UV-cured glue before!!!!

[@conwaylai8562]

@@uploadJ Yes, it wasn't the norm in China then.

[@volvo09]

Almost makes me wonder if the company used a bunch of discount known bad motors as generators thinking they'd suit... What an odd failure and a crappy design, that motor does not look like it would have survived high speed rotation either way...

[@SubTroppo]

If only there was a way to quality-control company directors.

[@arthurmoore9488]

Worst part is this is easily fixed.

[@uploadJ]

re: "What an odd failure and a crappy design" Ignores possible abuse by the owner, WHO might have dropped these units as well, jarring the magnets loose?

[@jhoughjr1]

If you spin anything fast enough it will fail.

[@T2D.SteveArcs]

I remember seeing a "dynamo" on my grandfather's bike as a kid, I asked him about it and he said "oh that thing !" "It just makes it harder to peddle, they don't use them anymore now we have good batteries available" that was in 1987 😅

[@pauldeboer]

In hindsight it shouldn't feel smooth as silk when you rotate it I guess. Also when you opened it I noticed the strange distribution of the magnets with a gap between 2, see @5:41. That should have been a clue what happened but I didn't get it either until you got one of the magnets out. I wonder if there are different forces in play when it is used as intended, eg as a motor. Not sure, more coffee needed.

[@EEVblog]

Yeah, was quite confusing when first opened. Can't say I've ever seen nore heard of rotor magnets coming apart before. The feel after repair is kind of a bit grindy, not a magnetic slip, the back plate which handles the centre alignment didn't go on fully.

[@ninjaneer0]

Aha! I knew that gap was strange as well! Great catch.

[@HPD1171]

as a motor i would expect even more forces to be exerted in the rotor as the coils would be actively powered applying more force to the magnets along with the centrifugal force. so either they are operating it far above its intended rpm range or they seriously need to audit the factory making the motors as I would never pass a construction like that relying solely on glue and soo little of it. there should be a physical thing holding the magnets in alignment such as just molding it in plastic which would be far cheaper and stronger then the laminated steel and glue they used in the rotor which is doing absolutely nothing. Usually Chinese factories are good at copying other designs that work but that often means they copy the mistakes too and combined with cutting corners is a recipe for disaster which is likely what happened here. we see this sort of thing all the time at my company and we are constantly having to audit (babysit) the factories in china to make sure they are using the correct manufacturing methods that we specify and every time they seem to find a way to change something to make cheaper that ends up biting us in the ass if we are not constantly vigilant. perhaps this company does not have the experience of dealing with china and are learning the hard way

[@deezelfairy]

It's a brushless dc motor and they never feel smooth - they fell notchy as the magnets align with their next coil lamination. In the rc model world we call it 'cogging'

[@spr00sem00se]

That's gap between the magnets was the first thing I saw. I still didn't realise what had happened though. Piece of junk for sure. I'd rather an old 70s dynamo .

[@jarnohonkanen4321]

This was interesting, not because faulty product but it relates to topic of my masters of theses "Bicycle dynamo powered charger for electronic devices" (paper is in finish) where I studied HUB generator powered USB charger, done some measurements about HUB generator, and done some simulations. 3-phase generator is definitely better choice than 1-phase generator what usually is seen HUB generators. More steady output voltage after rectification. Quality of construction of generator really matters, it's harsh environment. All that vibrations and etc.. but here rotor construction had extremely bad quality 🤦

[@johnrobholmes]

Laminations are used on the PM rotor because it's practically free to tool the lam mold to cut the stator and rotor support (back iron) at the same time. There is some amount of recirculating current in this "back iron" but laminations aren't really needed there.

[@yowieP51]

Found this problem a number of times Dave in model aircraft motors its quite common for the magnets to let go.

[@GannDolph]

I had one of those as a kid - manufactured over 50 years ago at this point It was basically a perm magnet brushed DC generator (motor) you'd wire directly to the lamp module which contained a standard torch bulb (6v as I recall) . There was only one wire - they grounded by clamping down and gashing thru your paint to the bike frame. The on off switch was simply the mechanical lever that would engage/disengage the drive cog against the side of your tire. It had horrible magnetic cogging and placed a surprising load on the wheel which you could really feel . The light intensity would go up and down as you peddled and went off when you stopped. No capacitors or batteries or regulators. It ran quite dimly most of the time , to leave some headroom in case you went really fast so it wouldn't blow the bulb..

[@GannDolph]

@@Okurka. yup. no three phase rectifier , s capacitor , regulator etc. But mechanically speaking looked / worked the same .

[@jb9652]

Haha - perfect description of the components and the installation! One good thing about them: AFAIK, they never got stolen! 😅

[@Wineman3383]

I'll be, the glue was what gave up in the motors. Soon as they got a little warm.

[@uploadJ]

OR dropped - OR over-sped. Maybe the glue gave up in +100 degF Aussie summer?

[@Wineman3383]

@@uploadJ that's most likely what is really happening. Over heating.

[@shreddder999]

Perhaps it could wirelessly transmit the power to the grid through the solar roadways.

[@f.f.s.d.o.a.7294]

Solar Freakin' Bicycle Generators

[@ratbag359]

🤣🤣🤣

[@SianaGearz]

Germans formed a "Forumslader" enthusiast working group decades ago, because any attempts at electronics coming out of bicycle accessory companies were garbage. Their DIY series of devices have several design goals. Main is compliance to legal norms - you simply HAVE to have dynamo lights in Germany, so you don't ever run out of energy. When you do, it's nice to be able to run them independently of pedalling, to give you extra safety when you need it, so they added hold up power which can be activated manually. Also they made it stealthy, you hide it entirely in the frame, so the authorities don't go around questioning you and people aren't so inclined to nick your stuff. Plus you get the ability to top up your GPS (this is when the project started, before smartphones were around) and phones and gadgets. Many members go on bicycle tours hundreds of km with intermittent camping and whatnot, and having true energy independence is great for a peace of mind. Then of course they got carried away and added Bluetooth and whatnot, which is nice as well, people love their bicycle tachos, but they are a little finicky, and the hub dynamo gives you regular pulses anyway. You've got to use an actual bicycle quality dynamo to get that power. Tyre side dynamos work but aren't entirely pleasant. Hub dynamo is fairly nice though, and when it's shorted out, you can barely feel it at all, and those are reliable, they'll last for decades. That's the one sort of part you want from bicycle parts companies where they know what they're doing and are held to relevant standards.

[@mosfet500]

It only takes one magnet to come lose and stick to the coils. Once that happens it jambs the whole motor up and then all the rest of the magnets break loose. The magnets change polarity from one to the next - NSNSNS. You can use a spare magnet to get the orientation, it doesn't matter if it N or S just as long as they run opposite from one to the next.

[@mikeselectricstuff]

I wonder if the hall sensor is so they can adjust the charge current depending on speed

[@EEVblog]

Possible. I would have thought you'd just dump as much energy as you could into the supercaps any way you can and not worry if it drops out or whatever?

[@mikeselectricstuff]

@@EEVblog I think you'd want to adjust the loading, if nothing else to reduce risk of slipping. The patent ( AU2021250917A1 ) mentions use of a threshold speed below which it doesn't draw significant power

[@EEVblog]

@@mikeselectricstuff I think what might happen is that mostly the USB charge current doesn't come from the supercaps, USB output is only enabled when the RPM/voltage from the generator is high enough. At lower speeds it just scrapes what charge it can for the supercaps to power the circuitry.

[@nickldominator]

I'm pretty sure the hall sensors are used when the generator is used as an actual motor, not a generator. That configuration is extremely common for sensored BLDC motor control. The generator is most likely just a repurposed off-the-shelf sensored BLDC motor, and my guess is they use one of the hall sensors as a handy speedometer gauge for the bike.

[@chrispollard6568]

No, it's just a depopulated standard brushless motor being used as a generator

[@byronwatkins2565]

Nice catch, Dave! This was never destined to be reliable as a motor either. Adhesive doesn't stick well to ceramic magnets... especially under tension. I would have wrapped the ends of the magnet assembly with a few turns of fine wire and covered that with adhesive. Even better would be to specify grooves to fit the wire.

[@Randrew]

Hard disk drive head voice coil servos all use glued, plated neodymium magnets like these. But I'll admit they probably don't see anywhere near the Gs this pedogen delivers. I still think it should be quite doable as the epoxies commonly used for such things are very strong in tension, though weaker in shear. In fact the best way to remove hard drive magnets is a sharp strike of the servo magnet bracket against something hard - striking with the shear line of the glue perpendicular to the hard object being struck against. As my son discovered and videoed when he was about 13. How to Remove Hard Drive Magnets: kzbin.info/www/bejne/lXrQYal_q8-DhLM

[@leocurious9919]

The glue isnt sticking to any ceramics, its a metal coating. Also, its an alloy, not a ceramic. Also also, glue has no problem holding on to ceramics.

[@byronwatkins2565]

@@leocurious9919 And yet it falls apart almost immediately.

[@leocurious9919]

@@byronwatkins2565 Yes? Because its a metal surface, cheap adhesive and only a tiny bit of it.

[@caerffili_callin]

These type of products are popular among long distance touring cyclists who may not have access to mains electricity to charge a lower bank up. Usually the electronics is connected to an already installed hub dynamo on the front wheel, but they have become very expensive and I assume this product was designed to be a lower cost easily removable alternative.

[@jcxtra]

Hi Dave! Love the channel, you, Big Clive and DiodeGoneWild pretty much taught me everything I know. When I was a kid we used to have a dynamo generator on my bike for powering tungsten lamps for riding at night. It made the bike tougher to ride, but it meant that I didn't have to keep buying batteries when I only had my pocket money. These days though I'm guessing given how efficient LEDs and lithium batteries are, it's probably easier/cheaper to just strap a decent sized "smart" power bank to it and run a bunch of USB cables, or if you wanted to go 12v (like SLA maybe), you could wire in a cigarette style 12v to USB adapter. I guess this product is in the gimmick / convenience market, rather than solving a problem?

[@harshbarj]

$299 for what's incorrectly called a "bottle dynamo" (All that I have seen are actually generators as they produce AC, not DC). I paid $80 for my Sturmey archer generator hub with built in break. For less than $50 I could easily add a battery pack and get the same thing as this. As a bicycle commuter having always on light is great and having them powered by a generator mean never worrying about batteries.

[@randomviewer896]

This BLDC motor would work just fine if the magnets were fully potted in epoxy like you did at the end for your repair. It blows my mind that a company would ever sell such a motor with internal magnets that only held onto the shaft with glue. Even if this was used for its intended application, it would still be a matter of time before the magnets flew off.

[@sw6188]

China does not care.

[@andyjdhurley]

Of course the company making the motor is NOT the company making the end product, problem is the latter found the cheapest supplier and did not check the part was up to spec. I bet they did not even know how bad it was until they started getting lots of returns and then added the poorly thought out 40mph limit on their instructions... instead of recalling the product and changing out the part for one that is up to the job.

[@mikefochtman7164]

@@SporadicBenevolence I was going to say the same. In REAL generators, the magnets would be a tiny bit longer than the coil poles so you can 'band' each end of the rotor with any non-conductive 'string' to hold the magnets on the rotor. But that would cost a few pennies more. Maybe the darn thing just isn't rated for the speeds involved?

[@felixhirt9284]

@@SporadicBenevolence its not just about the price, i think. the bandaging reduces the air gap, resulting in a torque reduction. We are designing our own high-power, high-speed motors (25krpm) for our racecar - and AFAIK, without bandage. For the afforementioned reason. The magnets are embedded in the core. A colleque of mine simulates exploding rotors for high-power machines - all without bandage. We are talking about high-reliability applications :-). In this case, however, they *should* have added it.

[@losipoop]

Sintered rotor was the only logical choice in my eyes.

[@dem0n1k]

It's dodgy. We've had push-bike wheel dynos that worked fine to power lights for years with no such magnet issues. Cheers for the pull down report as always!

[@yodab.at1746]

Those generally have a horseshoe magnet rotating around a stationary coil. Very robust.

[@jhoughjr1]

Also very low power and not very clean.

[@GregBurrowa]

The evolution of Technology! My big sister had a similar device for her bicycle lights. That was 70 years ago, it worked well.

[@ct92404]

@@Willam_J Hah, as far as oldies radio stations broadcasting 80's music goes...they must be trying to appeal to air-headed Millennials or something. The Skinny Jeans Generation thinks anything from more than 3 years ago is "retro."

[@ak-em7yr]

Hi!!! I have one that failed too!!! I bought this Pedalcell/ex-cadenceX and it was almost 400 €! The same thing seems to happen to me, the controller and the dynamo still look fine, but not generating any power! On the manual they say : " do not go beyond 40 km/h" (which is a bit silly on something supposed to be used on a bicycle) because it could damage the unit. I always followed my GPS speed to never go past it, it still failed only 3 months after I bought it. A bummer for something this expensive... I will send your video to their customers service, thanks, I was about to open it and I see your video. I will keep you updated to see what they have to say about it.

[@highfidelityinc]

I think it is interesting that there are 6 poles on the stator and 8 poles on the rotor. The first time I saw that I would have thought there would be the same number on both. But after do a little searching, apparently having an uneven number makes the output sinusoidal and reduces vibration. Be interested in hearing your take on the reason for the difference.

[@tactileslut]

Thanks for noticing that and bringing it up. It makes sense to me in that the shaft is going to want to stop with a magnet as close as possible to a pole piece, and if they all lined up at the same position it would hum like a stepper while being turned.

[@SkyCharger001]

Another thing, since this was originally meant as a motor, the unequal count assures that there are always at least two coil sets that can pull the rotor into alignment with themselves, allowing the motor to self-start.

[@mikefochtman7164]

Yeah, I've seen this before. Especially with MOTORS that you like to run at very slow speeds (as in Model Trains). When the pole numbers match, you get more 'cogging' where the slowly turning rotor sort of 'jumps' from one pole to the next in a not very smooth fashion. This arrangement, as you slowly vary the three phases, the rotor moves more smoothly.

[@highfidelityinc]

Thanks to all who replied. It definitely helps to clarify why there are usually an unequal number of poles on the rotor and stator.

[@Randrew]

@@mikefochtman7164 You're right. And as a generator it should also benefit from this with reduced cogging.

[@ArtyomGalstyan]

You know what? I have a feeling that the guys at the company are 100% aware of this and not going to do with it anything. Well, probably because it helps them to sell more and more of these products when they fail. You mentioned yourself, the guy has just bought 2 of those. Not sure he will go for the 3rd one after your investigation. By the way, great job, you cracked the case in 15 minutes 😁

[@johnpekkala6941]

If you sell me a shitty product u loose me as customer! I say this is really to bite yourself in the foot if it was made like this on purpose. Customers expect things they buy to work and when they dont they will spread negative critics about the company all over the internet giving the company negative reputation for making crappy stuff (just as it should be)

[@AffordBindEquipment]

@@johnpekkala6941 and in 24 hours, about 10,000 people world wide know not to buy one of these! and at $299US, that's a lot of negative advertising!

[@MSP_TechLab]

@@AffordBindEquipment unfortunately not. Of course there are people who are into biking and electronics, but I assume that mostly not. So, it will take time until small group of electronics guys will spread this information.

[@bloeckmoep]

Doesn't matter if the "original" Brand name suffers, give it a day and the EXACT same faulty product is available on Amazon under a different generic Brand name like "XXYC usb bike dynamo" or "EXYC usb bike charger". As long as the price is cheap it will sell.

[@johncoops6897]

Artyom Galstyan - The guy didn't buy 2... he got the second unit free under warranty claim. Pay attention, and stop spreading BS > kzbin.info/www/bejne/bqq7i32QeLB8o6s The more units that fail, the more money the company loses due to warranty replacements... faulty components don't benefit the manufacturer in ANY way.

[@sysghost]

Interesting. A friend of mine had this bike generator that failed only weeks after installation. I start to suspect it failed the very same way.

[@KLondike5]

That was spinning way too easy at first. The fields definitely produce resistance when the magnets aren't all flung off.

[@mrbmp09]

Back in the late 60s they sold small generator/ headlight sets for bikes. You had to be moving 5-10 mph to get any light out of it.

[@uploadJ]

They used a "wire" in the bub to make the light, too. A little Tungsten wire, it got real hot, and created broad-spectrum light ... those were the 'daze' ...

[@antonypilepich8947]

I've never had and and success with gluing neo magnets with thin layer of epoxy, also most epoxies soften with heat, it was bound to fail.

[@alicangul2603]

My guilty pleasure: Watching Dave not knowing (power) electrical stuff 😅

[@SystemX1983]

They should have written in the manual: riding faster than 40 mph will most probably activate "SELF-DESTRUCTION MODE" It's not a bug, it's a feature 😉

[@MickMcMadder]

Should be safe on a Penny Farthing front wheel, but not a child's tricycle.....

[@pauljs75]

Knowing Chinese, they probably substituted MPH for km/h, so they were only good for half the speed on the instructions anyways.

[@AntzWilkz]

Yes the magnets have come away from the rotor. Some motors also have a seperate magnet assembly to drive the hall sensors. On one of the BLDC motors I worked on the ceramic hall affect magnet ring would come away when the motor got exceptionally hot. It would disintegrate leaving bits all over the place. The failure you are observing is just because of some aspect of the adhesive.

[@walterpark8824]

Fascinating! A motor designed to destroy itself by centrifugal force of spinning. ‘Works great if you don’t turn the shaft!!’ This is like that NOP box that uses a mechanical finger to turn off its own switch. An amazing diagnosis and prescription for recovery, Dr. Dave. I guess we’ll have to put the stator-stator motor effect back in history’s reject pile.

[@johncoops6897]

You realise that the motors in a "hoverboard" have a shaft that doesn't turn?

[@mozismobile]

Such a simple mechanical failure! And that does explain why some many of the things have died, because the combo of bicycle vibration and spinning fast is going to break that glue every time. I wonder if there's room for a wrap or two of fibre tape for a repair?

[@gorak9000]

it's not good enough for the magnets to be taped into a ring - they have to be mechanically connected to the shaft, so the tape won't do squat. Just need actual glue that holds.

[@uploadJ]

re: "because the combo of bicycle vibration and spinning fast is going to break that glue every time." Yup.

[@Enigma758]

Funny! I gather the polarity of the magnets needs to be considered when gluing them back on.

[@mark-]

Big engineering fail on this motor. Looks like they used a dodgy $1 motor on this... Thanks for sharing 👍

[@ruben_balea]

As for the laminated construction I think they just make the components and assemble them using the same machines that they already had to make brushed motors because that's more cost-effective, sheet metal can be punched to final shape much faster than one can see with the naked eye, for a solid piece they would have to use a lathe to face the ends, turn the outside diameter and make the hole... I don't know how fast a CNC lathe can be but I doubt it can outperform a die cutter.

[@PocketBrain]

"Stabbies." Sharp probes to stab thru the conformal. Above a certain RPM, your magnets are going to throw to the outside and part with the rotor.

[@uploadJ]

OR - just drop the motor ... onto a hard surface. Concrete might do it.

[@timthompson468]

I wonder if a thin plastic sleeve around the magnets would keep them on the rotor. Is there enough clearance for something like that? Interesting troubleshooting though.

[@josephcote6120]

Even something like heat shrink tubing might work fine.

[@EEVblog]

Barely room, but possible.

[@Brian-L]

I was thinking something like a kapton sleeve, but I’ve never seen kapton molded in tubular form.

[@bardrick4220]

REPAIRED!!! 🤣 Just needed better glue! . . . I smell some designed obsolescence here too.

[@coriscotupi]

The gap between two of the magnets seen at 05:37 made it immediately evident that something was wrong with the placement of said magnets. Since they were in a more outward position than they were meant to be, the larger perimeter was not completely filled by the magnets. I have a gasoline-powered RC model helicopter that uses a scheme not too different from this device. A brushless motor (of the kind usually employed in electric RC model airplanes) is driven at 11,000 rpm by the gasoline motor. Its output goes through a 3-phase rectifier and voltage regulator. Keeps the heli flying regardless of the state of the onboard battery, which provides great peace of mind. In fact, a few people have reported that they forgot to turn on the battery altogether and went on flying, on generator power alone.

[@MorRobots]

Hahaha Dave was staring right at it...lol... how long until he sees it... 7:11 😆

[@PF-gi9vv]

I couldn't believe he didn't see it straight away, I was thinking "whats he talking about, can he not see it, he's missed it lol"

[@donmoore7785]

"Yeah it's a piece of cr**" caught me off guard it came out so quickly. Agreed! This thing had a price tag of $299!

[@SCAPE0GOAT]

I'm guessing the glue failed as the unit got hotter & hotter. If it's a glue like super glue, it turns back into a liquid if it's heated beyond a certain temperature. Pretty shoddy. Great vid though Dave, love it 👌

[@artursmihelsons415]

That generator is big failure, but fixable with lots of epoxy at magnets ends and gaps before use.. Thanks for great video, Dave! 👍

[@tomvleeuwen]

People have been using alternators on bicycle wheels for ages to power the lights. These days it's integrated in the hub and the lights are LED, but back in the day these small alternators could also deliver a few watts of single phase power. A rectifier and buck-converter is all you need to charge your phone. Also the hub alternators are able to deliver a few watts

[@ct92404]

Yeah, there would be a small generator mounted on the fork, which the wheel would turn by friction. But the problem with that is that makes the bike harder to pedal and it always caused more wear on the tire.

[@tomvleeuwen]

@@ct92404 Exactly, that's why I often drove without light in the dark😅. But the hub alternators hardly provide any resistance at all.

[@g4gxl]

@@tomvleeuwen of course any generator will create some resistance but the PedalCell is virtually undetectable in use. It is actually more efficient than hub alternators. The point of this is that it can charge a phone, bike computer and lights on a multiday trip where there is no access to mains charging.

[@DavePoo2]

I think in the previous video I saw that the manual said "don't exceed 40mph", seems like they should have set that speed a lot lower given the amount of glue holding it together.

[@mikefochtman7164]

My thoughts as well. Estimating that drive wheel is about 2.5 cm, even a leisurely 15 mph would send it over 5000 RPM. Small wonder it 'spun itself apart'. Classic over-speeding failure.

[@DavePoo2]

@@mikefochtman7164 ha ha! 15mph is my average speed on a bicycle! Make sure you don't go down any hills.

[@jhoughjr1]

@@DavePoo2 Yeah it's not like they have brakes so you can control your speed or anything. I get that most drivers are incapable of maintaining a speed limit.

[@frollard]

It's an interesting segue to how the new Tesla S plaid needs carbon fiber overwrap on the motor rotor...no gear changes means it needs torque from 0-200+km/h. I forget something like 9:1 means the rotor something absurd 30cm(guess from memory) diameter has to spin at absurdium rpms...Rotor needs the carbon wrap just to prevent a nearly solid piece of metal from becoming shrapnel.

[@Blox117]

no, the rotor is different than shown here. It's an IPM design where the magnets are embedded inside of the steel. They can't come out. In the IPM design torque is provided threefold: permanent magnets, electromagnets, and magnetic reluctance Also and more importantly, this design is over 20 years old and Tesla has nothing to do with it.

[@DrHarryT]

The first thing I saw when you took the cover off was the magnets stuck to the stator coils.

[@gandalf87264]

Don't worry about the vibration. when you have a 622cm diameter wheel spinning a little 50cm diameter wheel on the generator, when that bike gets going, can you imagine the G forces on those magnets?

[@JPN76]

Guess it exceeded the max RPM I guess we could calculate this if we knew the diameter of the bike tire.

[@EEVblog]

The OP said he didn't ride it hard and broke two of them.

[@JPN76]

@@EEVblog yeah it didn't look like a good epoxy anyway. Any decent motor I have taken apart has epoxy on a bit thicker and just looks better. That looks like hobby glue or something.

[@danielthechskid]

Tire diameter is irrelevant since the edge where this is driven will always be moving roughly the same speed as the ground hence this PMA's drive wheel will also. IE the tire is just an idler and it's the diameter of the PMA's little wheel that matters. If it is a 50mm diameter it's just about 100 RPM per km/h.

[@JPN76]

@@danielthechskid yeah that makes sense to me.

[@ericgerstenberger7813]

700c x 23. It was on my indoor trainer with a max speed of maybe 20mph (36 km/h).

[@Wineman3383]

You just did R and D for cadence. Now they can have a recall and ship out working units. Yeah right 👍 Sure they will!

[@Rx7man]

As likely as Futuremotion doing something like that I"m sure

[@youdontknowme5969]

yeah recalls are for usually only safety reasons e.g. n-teen people have to die first, n-teen properties have to burn down, class-action lawsuit, etc it's sickening really

[@gorak9000]

Chinese R&D? Receive and Duplicate?

[@GodmanchesterGoblin]

@@gorak9000 A sales manager for a company I know in the far East referred to it as Repeat and Delay due to the number of unsuccessful design iterations and the delayed product launches that this caused.

[@youdontknowme5969]

Rip-off & Destroy

[@stevenbliss989]

They probably thought they were getting a bargain buying up a bunch of cheap 3p motors at surplus, but they got screwed by the supplier! ...WOW!

[@uploadJ]

Newsflash: They all did pass incoming QC though I'd bet. Time does things to various glues though ...

[@stevenbliss989]

@@uploadJ I find it hard to agree, for glue to go bad that fast...???? And let's say you are right, then it is a serious design flaw by using a crap glue that doe not last, either way it still smells of badly/cheaply made product.

[@MartysRandomStuff]

Typical case of using a component in an environment it wasn't designed for and then doing zero testing. Customers are now the quality control testers. It was hard to watch how long it took to realize the magnets had come off. The hall effect sensors provide feedback for motor controllers that have speed control.

[@uploadJ]

re: "and then doing zero testing. " Like, temperature and vibe (vibration) testing? Yes ...

[@edwardwright6961]

the laminations are to stop over saturation and inprove the magnetic eddy current flow

[@pauljs75]

Thing probably would have worked if they used a motor/generator with an external rather than internal rotor. (The force from spinning would have kept the magnets to the outside bell housing, even if the glue failed.) Thus making you wonder why they made that design choice.

[@chrisayad0]

outrunnerzzzzzzzzzzzzzzzzzzzzzzzzzzzz

[@todesgeber]

i wonder if that's why they have a 40 mph limitation in the instructions. craptastic glue doesn't bind to standard neodoodlyium magnet plating.

[@thinklist]

😆😆😆 sorry have to correct UVW. Awesome video mate 😂

[@tahustvedt]

LOL they didn't wrap the magnets. They are usually wrapped with aramide. Why did they make such a complicated PCB? All they needed was a rectifier and a large range voltage regulator. I've made similar devices myself. No need for hall sensing on a little generator.

[@samoneil1938]

That issue with the permanent magnets falling off from the rotor due to ineffective adhesion is a pretty embarrassing point of failure. Although I could have sworn that one of the diodes was blown as well at around 2:09 in the video.

[@fiveleafcloverfpv4445]

😅 I already thought it's not going to work like that. You're right. The magnets should be stuck on the rotator. Bad glue.

[@robinbrowne5419]

The laminations are to prevent Eddy currents and resulting electrical losses which would happen if the core was pure steel, for example. The direction of the laminations and insulating varnish layers are 90 degrees to the direction which Eddy currents would happen.

[@slashkensol]

Thanks bro I just bought Pedalcell but nothing generate out!!!! Damn it!!! Will open the motor to check is it same case as your's

[@mattmoreira210]

Gosh... I'd've attached those magnets with *lots* of glue to the rotor and, just for good measure, potted them in resin as well. But I guess that's the engineer in me speaking. The accountants would've hated it. 😂

[@wjodf8067]

a strong tape on the outside would have held

[@gorak9000]

@@wjodf8067 All the mechanical power needs to be transmitted through the glue from the spinning shaft to the magnets to spin them. Tape on the outside would do nothing for you if the glue fails - then you'd have a magnet ring that's still not connected to the shaft.

[@mikefochtman7164]

@@gorak9000 Well LARGE motors use a lot of fiberglass taping. Of course the rotor poles/windings are in machined slots so the torque isn't transmitted to them by simple friction/glue. But the 'banding' is to contain the high centrifugal forces when you spin the rotor at high speeds. So if they machined notches for each magnet to sit into on the rotor, and then machined a tiny notch on the outside of the magnets for some sort of high-strength band, would have been a lot better. But I'm sure, as Matheus pointed out, 'the accountants would have hated it.' 😂

[@J0eMega]

@@gorak9000just tape it down harder, bro

[@01glenn0101]

still remember the old generator on the front wheel... not the best light output ...

[@creepyloner1979]

the entire surface of the rotor is covered in glue. the "beads" are just the extra glue in the gaps between the magnets. the reason there's no residue on the magnets is because of the glossy nickel plating which is difficult for glue to bond to. typically those magnets would be soldered in place rather than glued for a use like this, or wrapped in something.

[@pocoapoco2]

In days past, those magnets would have been glued to the spindle, then screwed to the spindle straight through the back of the magnets and thread into a tapped hole into the spindle. Finally those screw heads would have been chisel/punch-set so they couldn't back out.

[@robroysyd]

Now that what is going wrong is known it doesn't look too difficult to repair with a little JB Weld. Just need a couple of rubber bands to hold the magnets in place while the glue hardens. The magnets could probably themselves in place I suppose now that I think about it.

[@azazeldeath]

Only at the 7:20 mark, and your spot on those magnets should be glued or attached to the center part, my guess is the makers of this used something pathetic like superglue or just hoped the magnets would hold themselves there. Source: Was a mechanic, have rewound a couple of alternators in my day, so to me the failure the second I saw in was pretty clear but thats just because of my experience. Love your work and keep it up.

[@StuffJason437]

I was waiting for follow up video, as were anticipating what went wrong with it. I'm always interesting in different failure modes. At 1st thought could've been a voltage spike but dammn if the bike went over ramps, and had a ton of vibrations magnets twerkelate themselfs loosey goosey, and boom no more electricity for you! Hats off to who designed the electronics, and the housing for it ...But whoever designed that generator didn't take into consideration the sheer amount vibrations created when cycling. Also, glue wouldn't be solution only temporarily. But, I'm no design expert on this, just have experience re-building the bloody things.

[@Rx7man]

the centrifugal force FAR exceeds the vibrational forces I'm sure. I bet if you spin a new on up to speed it'll die right there, unless this was just a bad batch of unglued motors

[@codebeat4192]

Don't understand why they don't use a plastic wrap (shrink tubing) around the magnets to keep them together and in place. Cheap and easy or is this designed to fail?

[@nigozeroichi2501]

As soon as you pulled the end off WTF?! Ah, I had a similar problem with a motor, worked fine made a weird noise then no workey no more, can't wait for the magnetic mine looking thing video🙂

[@gamesbychris]

Wow! They charge $300 for one of those!!! Think they could afford a bit more glue

[@jdmccorful]

Buyer beware!

[@eightysevenmoore]

UVW = 3 phase commonly known as phase 1,2,3 or A,B,C. It’s a 3 phase PMAC motor. Delta wound.

[@jr0th]

After seeing the design choices, I'm pretty sure the hall sensor output is there to "log" the speed, and when (not if) it breaks, deny replacement under warranty...

[@johncoops6897]

No, dummy. The hall sensors are used when it is used as a motor. And the guy got the replacement for free under warranty, which was not denied. kzbin.info/www/bejne/bqq7i32QeLB8o6s

[@johnwright8814]

Proper industrial servomotors can lose their magnets if they spin fast enough. When the magnets fly off, the coils are destroyed. This is how I destroyed a big and very expensive industrial servomotor. Heat-cured tape is wrapped around to hold the magnet assembly together, but a few seconds at speed was enough to break them off.

[@Arek_R.]

6:22 That is some super wanky trace layout, does their PCB CAD software not got option to draw circular traces? At work I've got proteus and I would draw whole 90 degree arc and then remove segment by segment to get any other angles.

[@markperry222]

That looks like a different brushless motor would work. It looks like the surface of the magnets are too smoth for the glue to adhere to. The pedal cell bit seems good, but the motor needs to be changed.

[@ZaPpaul]

I wonder if there is enough clearance to just wrap the magnets inplace with a large bit of heatshrink tube? That would negate the need for glue and as long as there is no friction, would last forever.

[@dejayrezme8617]

I've wondered before how you would glue those coated neodynium magnets. Must be difficult. Also don't you generally want very little glue but with enough pressure to ensure a good bond? Generally the glue will be less strong than the glued parts so if you have thick glue it can break more easily.

[@grndkntrl]

Most likely some sort of epoxy or cyanoacrylate. Looks like too little was used going by what we can see on them @10:35

[@dejayrezme8617]

@@grndkntrl I guess you want just enough so that it will be squeezed out and a thin film remains. Probably a messy job in the manufacturing line.

[@lasskinn474]

epoxy potting would work as mechanical fastener. they sure had enough potting for the electronics. I guess the cumbersome bit is in balancing the epoxy really.

[@RavenLuni]

What about putting some heat shrink over the magnets to hold them to the rotor?

[@thesunexpress]

Looks like the accounting department at PedalCell was in charge of ordering a bulk load of brushless motors.

[@sbreheny]

The rotor laminations do matter because the magnets act like air (in terms of inductance effects because a magnet is fully saturated) and the inductance seen in the windings involves the entire magnetic circuit (back iron, stator cores, rotor iron, and then back out)

[@Blox117]

it concentrates the field lines from the p. magnet right?

[@sbreheny]

@@Blox117 It concentrates both the field from the permanent magnet and the electromagnets. The field strength matters most in the air gap between the rotor and stator. You maximize that by creating a "magnetic circuit" with the lowest reluctance (i.e. highest permeability), which means that you want most of the field line loops to travel within iron.

[@Blox117]

@@sbreheny neat. so how do reluctance motors work then? It's the shape of the steel that matters? And it tries to move to create the lowest reluctance? It's kind of hard for me to understand why that produces a rotating force

[@nikize]

One of the diodes on the board seems to have had their magic smoke escape.

[@redsquirrelftw]

That does seem like a bad design to rely on glue only, those magnets are under lot of force, they want to attract to the coils and they are under centrifugal force too. Almost should be some kind of plastic sleeve that goes over it too. (don't want to use metal or it will mess up the magnetism)

[@marcoriviera06]

Dave, STRICTLY speaking, the silver bit is a brushless synchronous motor (not dc) It can be turned into a brushless dc motor with the appropriate control circuitry I’m sure you knew that, but I thought I’d bug you a bit!

[@cha1rtech]

good review, thanks. BTW if you still have the Pedalcell I'd be interested in buying it, have a project that would use only the generator, wouldn't need the circuit portion

[@johnsonlam]

It's absurd that the magnet falling of inside the motor, what's the price is that motor? I guess a few US dollar.

[@thorhallurragnarsson2381]

Take a second look at the diode bridge rectifier, to me one of the diodes appears to be defect or missing?

[@EEVblog]

Yes, see the teardown in the mailbag, the diode ripped off because the board was potted in the box.

[@4dirt2racer0]

the motor manufacturers attempt to affix the magnets is absolutely pathetic.. they should genuinely b ashamed of that

[@ipfmcd]

LOL, just got one of these cheap on eBay with the same issue, gonna repair and get my money out of it. Saw people suggesting superglue, will that cut it? What’s everyone’s best suggesting for adhesion of this magnet to the core?

[@kokodin5895]

to fix it you probably should do it in 2 steps 1 epoxy the core and slip on the magnets but separate them with teflon/polyetylene spacers and wait for glue to set 2 remove spacers and fill the gaps with more epoxy glue it could be better than factory with cheap metal to metal epoxy :P

[@creepyloner1979]

if you use paper spacers they become saturated with epoxy and you save yourself a step. some hard plastic heat shrink tubing around the whole thing while the epoxy's still wet couldn't hurt either.

[@kokodin5895]

@@creepyloner1979 fact i used epoxy mixed with aceton to make pcb patch out of paper napkin and it worked, still it might not be as precise as with solid spacers in this kind of rotor if it was able to fall apart by just excentric force the whole rotating assembly might need to be weighted to reduce wobbling

[@Brian-L]

Helmsman Lojur: She'll fly apart! Captain Hikaru Sulu: *Fly her apart then!*

[@McTroyd]

Well... it's a good source of neodymium magnets, if nothing else. Maybe enclose the stator coils with the rotor of magnets? (Rotor outside stator, instead of within it.) That way centrifugal force would tend to push the magnets against the case, rather than at the coils.

[@jb9652]

£32/magnet!! Regarding having the magnets rotating outside the stator coils, enclosed in a cylindrical tube (which I presume you mean): I like that idea. Effectively no max rotational velocity then. Maybe the cylindrical tube could be the outside of the alternator, turned directly from the bike wheel or whatever?

[@McTroyd]

@@jb9652 You presume correctly. I wonder if this is what drone motors do, since the whole outside housing tends to spin with those. Coupling directly to the wheel would help with power generating efficiency, but might come at a cost of constant drag.

[@Sir_Uncle_Ned]

Jesus wept. PedalCell really need to choose another supplier for those motors that actually builds them properly!

[@jameswest848]

I restored an e-bike motor with this issue. I glued the magnets back in place with loctite 638 and replaced the brushes.

[@jdmccorful]

Couldn't find 638 but found 641. Its app. notes says it would work with this kind of problem I believe.

[@jameswest848]

@@jdmccorful I bought some on Ebay in the UK based on recommendations for this issue.