Golden Motor Magic Pie - first taste

Hyena

10 GW
Joined
Aug 13, 2008
Messages
6,222
Location
Sydney, Australia
Read on for a slice of the magic pie action, something you can sink your teeth into.
OK, enough pie jokes :p

Below is a fairly long review of my initial findings. See my second post below for a summary.

So I received a magic pie to review today. It had arrived while I was at work today but I only had 1.5hrs after work before dark to test it so this is just a preliminary review so far. I'll make a video on the weekend. I got a 26" rear but it came with no sprocket and no brake handles so I'm unable to test regen braking or comment on the ridability for now. My brake handles have the gear shifters built in anyway so I wouldn't have used them anyway.

So first impression ? It's big! Yeah we knew it was going to be bigger than the old GM offering but it really dwarfs it in comparison.
The spokes are quite short as a result and you'd probably have to lace a 24" wheel radially. They're quite a heavy gauge too, I'm not that up on spoke specs but they look very tough. The big golden motor writing on the side of the rim is just a sticker that is easily removed.
The rim is is also noticably thinner than my current GM one at 1.5". Not that it matters for this build as I'm running quite a narrow semi slick road tyre

sizecomp.jpg


spokes.jpg


It comes with a throttle with a red button on it and 3 charge indicator lights. I'm not sure what voltage these are supposed to be set to as all 3 illuminated when tested on 40 and 60v. The LVC for the integrated controller is set for 24v so maybe so are the LEDs ? That's if they are LEDs, I notice after 10 minutes of having the power connected plastic housing was quite hot to the touch. Unless there's a bit resistor in there trying to hold back most of that 60v ? The button didn't do anything but I haven't had a chance to have a play with it yet. This may not be the final throttle that comes with the kit anyway, as this was just sent to me for review rather than the final product. Pictures on GMs website show cruise control buttons too which this also didn't come with.

throttlewire.jpg


The cabling is very neat with one fat cable exiting the hub and then splitting into a power lead and another multipin lead for the throttle, cruise, brakes voltage indicator etc.
It's a nice touch that GM includes a seperate set of the pins and plugs for those handy enough to cut the cable to the right length (it's waaay too long out of the box) and reterminate the leads.

So... pop the hood ? Yeah, I know that's what everyone wants to see :p

I took a guess on which side would have the controller circuit accessable and unbolted the side with the cable exiting. Wrong answer. Doh.

inside1.jpg


Ah well, you can the caps from this side, they're rated at 63v as expected hence GMs upper limit of 60v.
Not alot of coppery meat in this pie... :|
Flipping off the other side gave a better view of the controller. It's mounted on an aluminium plate about 2mm thick. Also as you'd expect given the space constraints its a little 6 fet job. Wiring looks to be about 14g. There's a thread showing the insides of the magic controller floating around, but I'd say its slightly different given GM have implemented a few new tricks in this motor such as thermal and over current protection. The bulk of the controller goodies are hidden but I didn't want to unbolt that section because I don't have any thermal paste and even though it's remote, don't want to do anything that would influence the operation and longevity of the motor given GM is trying to test them before officially releasing them.

inside2.jpg



I didnt see any hall sensors in there but theres hall wires leaving the controller so I guess they must be just buried.
There also seems to be a few spare output terminals on the board - no doubt for future goodies to be added. One that's clearly visible reads 'lock' and I assume that's feature that's been mentioned before that effectively shorts out the 3 phases preventing, or atleast making the motor difficult to turn.

controller.jpg



OK so lets get it back together and on the bike!
I know some people don't like the pattern on the side plates but I don't mind em.

bike-pie.jpg


OK battery time, as I was half way though soldering on the 4mm bullet connectors to mate up my lipos I noticed someting...

polarity.jpg


Now I don't have a PhD in electrical engineering, but labelling a red wire as -ve and the black as +ve.. Really GM ? REALLY ? :?
By this stage it was nearly dark and wanted to test ride it so I had to take a leap. Most chinese stuff has notoriously poor QC, but do I trust the 'engineers' who assemble these things to have used the global convention of red for positive and assume the guy at the end lf the line who puts on these labels for noobs has just got them around the wrong way ? After a quick debate in my head I decided with all the fool proof work that had gone into the motor they probably had reverse polarity protectors as well, so went with common sense and hooked red to +ve and black to -ve. I plugged in my lipo pack and guess what...

burnt.jpg



An inch long 4mm bullet connector vaporised. Good thing my lipo pack was only 5 ah and rated at 25C so I only just discharged 125a into a short circuit at my fingertips :shock:

Take note of the colour of my fingers afterwards methods, no KFF :p
Seriously GM, you need to put a big warning label on that and bring it to your customers attention that it doesn't follow normal polarity colour code conventions.

Luckily I was right in that it does have reverse polarity protection (I figured that when the connectors blew up and there was no pop from inside the motor) and switching the leads over bought it to life.

So first impression ? This motor is alot louder than GMs exisitng motors. If I hadn't just pulled it apart myself I'd have assumed it was geared. When coasting along and backing off it has a cool electrical engine / jet sounding whistle to it. When accelerating and drawing any current at all the noise is quite noticable. At one stage when it got up to speed the noise went it was a slight decline and when I looked down at the wattsup meter it was drawing 0 amps. It's not power drill type loud, but definitely noticable and not as stealthy as my current GM motor. If you'd never heard a hub motor before I suppose it's what you'd expect an electric bike to sound like. I normally frighten the shit out of people on bike and foot paths when I come roaring up behind them at 40km/hr and they only hear me at the last second, while out testing today several people walking dogs turned their heads as I approached. I'll have to go for a longer ride before I decide if the noise will bother me or not. It was a bit of novelty for the 5 minute test ride but I suspect it could get annoying on long rides. I'll do a video on the weekend.

So what about the current rating of the controller ? GM never announced what it was, but I can reveal after testing under a few different conditions with hard starts and up steep hills that the max current is about 28 amps. For the most part it's hard to get it to draw more than about 22 amps and if I had to put a rating on on it I'd say 25a. That's not too bad for an integrated controller and infact that's all the old 12 fet GM controllers would do.

It's quite an 'economical' motor, only drawing about 5-10a for the most part bar hard acceleration. I did 2 quick test rides with a 10S lipo pack (~40v actual voltage) and a 15S lipo pack discharged a little so it'd operate under the 60v HVC. Actual running voltage with the 15S pack was about 58v. Peak power draw on 40v was 1000w, peak on 60 (58) was 1600w.

What about the speed ? My current 36v 500w motor running on 60v just hits 50km/hr - more like 48. Those running the 1000w 48v motor claim around the same on 48v, although their 48v in reality is more like low 50s. I expected the magic pie to be faster from early comments by GM until yesterday when they announced it would be slower but have more torque. On 40v it's not too inspiring, but then I'm used to running on 60v. On 60v its quite zippy off the mark and pulls up hills well (I had no option to pedal even if I wanted to because I had no chain drive) but it has no top end. I was really, genuinely disappointed by the top speed on 40 and 60v. I thought it'd be atleast as fast as my old motor, but in a sleeker package and with the extra diametre and magnets. Nup, it's a snail. It feels like a little 200w motor. Top speed on the flat was 25km/hr on 40v and about 36 km/hr on 60v. Top speed down a hill was 41km/hr and I could feel the motor holding it back (I go down this hill every day at full throttle with my other GM motor at around 60km/hr) This is probably best case scenario too, I weigh just under 70kg, my lipo pack is ~4kg and I'm running skinny high pressure semi slicks. Does this sound right Yao ?? Speeds were GPS certified too, not a $5 uncalibrated bicycle speedo.

I really wanted this motor to be great and I waited months for its release thinking it would be but I'm not so sure I can say it is. I feel like this is a top gear review, talking all about it then at the end saying "...but it's rubbish".

I'm not saying that, it's not rubbish but I think GM may have tamed it down a bit too much. From initial pre-release testing comments the magic pie sounded like the best combination of torque and speed you could expect but I think after alot of failures and teething problems they've backed it off a bit too much, determined to have no failures when going public. GM has announced they have more plans for a differently wound higher speed motors and other cool features. I think they're the ones to keep an eye out for.

If anyone has any questions ask away. If you want more / other pictures let me know, I have alot more that I haven't posted
 
So, in summary

It's a neat all in one package
Wiring is neat and simple and looks nice. A big improvement over just about any other kit out there.
Rim and spokes looks study but the rim might be a bit narrow for those wanting to run fat tyres.
Controller runs about 25 amps
Noise is louder than old GM motors but not excessive. It sounds cool at first but may get annoying after a while ??
Torque off the line and a low speeds is good but top speed is very uninspiring - 25km/hr on 36v and around 36 km/hr on close to 60v.

I think it'll be a great motor for those starting out with ebikes and commercially for bike shops and the like they're a breeze to install. After I was done stuffing around taking it apart I reckon I had it installed in 10 minutes. Its simplicity as an all in one package is hard to beat for noobs and for those not looking to go fast snap one up today for $160!

For the over volting, over amping speed demon crowd here, I don't think the magic pie is for you.
That said, after I've throughly tested the pie for use as it was designed and make sure the integrated controller doesn't prove to be problematic, I can't promise the soldering iron won't be coming out. :twisted:
 
Excellent job!

That is a huge diameter motor. It seems like it should have gobs of torque if properly fed. Sounds like they might need to lower the turns count on the windings. Delta/wye switching might be an interesting idea too.

It would be good to measure the Whr/km and compare it to the smaller motor. This one might get better electron mileage.

By having the controller inside, the phase wires are so short that there will be practically no loss in them.

Upgrading the controller will be a bit more difficult.

That's kind of crazy to have the red/black reversed. If it was mine, I'd put some black and red colored heat shrink tubing or tape around the ends of the wires, but I suppose the huge + and - labels should give you a clue also.

How is the motor weight? I forget if the weight spec is mentioned on their website. It should be lighter than a Crystalyte X5.
 
Thank You so much for the peek inside.

One.
Too many windings, too small gauge. EMF is huge on this pie. The stator sings.

Backside iron is 50% too small for the poles. Heating and efficiency are at issue.
They could keep the footprint and improve efficiency if laminations were used instead of solid poles.

It's a nice low cost offering either way.

Jeff
 
Hyena said:
Top speed on the flat was 25km/hr on 40v and about 36 km/hr on 60v.
Thank you for this review.

I'm a mechanical Engineer, so I'm no expert at electrical or acoustical issue, but will still say something on both...

1) Electrical. So if we linearly interpolate for a 48 V setup, the speed would be 29.4 km/h (18.1 mph). Either this is indeed a disappointment or something is wrong with this specific motor. This test should be re-done with a motor that was not subjected to connecting a 60 V battery with reverse polarity, so we can rule-out damage to the motor/controller from that mishap. Also, the +/- labels on the wrong color wires may have been specifically labeled for this motor because they realized they goofed on the wire colors.

2) Noise. The construction of this motor (large covers, and an internal 2 mm plate to support the controller) is a almost like a drum (the kind drummers use) meaning that thin plates under vibration are perfect for creating noise. I bet that if one inserted rubber spacers so that contact pressure is exerted at the center of these unsupported thin plates, the noise would happen at a different excitation frequency or maybe eliminated altogether.

Parallel discussion: Golden Motor Forum: Magic pie - first taste
 
20 mph performance is no suprise. For many it will be perfect, especially the internal controller simplifying the wiring. But clearly not the typical ES members cup of tea.
 
I thinks its great deal for the hoards.

As this E-bike industry is still in its infant stages what better kit could be provided for a newbie upstart, Many people get turned away from E-bikes because it gets a little difficult.

The rim look goods in the pictures, better than the one I got with my GM, those throttles are great and you will find them on the currie ezips, and I use that button as a dead mans switch which Currie uses them for P.A.S.

The spokes, if they are thicker than the old GM spokes then they are 12ga. I would honestly try the motor on a different controller before judging too soon. It has plenty of poles, 56 I counted, lost count so you could confirm that please. If the winding core was a little thicker 96v would be a blast. Can you measure up the winding core to the other GM motor. Eg how many winding in parallel and who thick they are

More series windings equates to longer windings so dont count the pie out on volts too early, with a nice fast PWM and high volts you fit plenty of watts into them. The stator sing maybe too low of a controller PWM for the length of the series windings and they need to bump it up or find a better harmonic. I would try get this onto your 12fet controller at 60v.

I think we will see alot more come out of this wheel in the future, but do you think this is a good offering for someone new to E-biking? I know a few wouldnt go for this here, but I would of 2 years ago.

Edit:
It would be sensorless controller and hub as it goes with the no need to fix it theme GM is trying to push with this motor.
 
Jerome Daoust said:
Hyena said:
Top speed on the flat was 25km/hr on 40v and about 36 km/hr on 60v.
First, thank you for this review.

So if we linearly interpolate for a 48 V setup, the speed would be 29.4 km/h (18.1 mph).
Either this is indeed a disappointment or something is wrong with this specific motor.


Maybe something wrong with the post, I think he ment 25kph on 36v and 36kph on 48v. Even my golden woops its ass at 48v.

If its speed is retarded is it efficient?
 
I guess instead of Magic Pie we should call it the Singing Pie Pans. Thanks for the detail Hyena... you saved me some bucks.

John
 
It must be electrically speed governed. They are likely trying to keep it in compliance rules by design. I do love the diameter of that motor though!
 
liveforphysics said:
It must be electrically speed governed. They are likely trying to keep it in compliance rules by design. I do love the diameter of that motor though!


Gm stated it was voltage limited to 48v and anything over would have a cutoff. Ummm this could have a few meanings but my guess is it's no action what-so-ever. I would await for the OP to confirm some more stuff over a few posts, I think Hyena did a pretty good scoop in the small time he had to tinker but to overvolt the internal controller to 60v isnt indicated in his post so Im predicting there may of been some errors. 36 kph at 60v on a 48v limited sensorless controller, if it is sensorless, the dots just are not connecting in this first review.

Edit:
We need to have impedance tested on the windings and winding turns tested and a big mofo sensorless 72v controller set to it. It wont need to many amps though as it seems to have plenty of torque for a gearless by ratio alone.
 
Whats this doovalaky for? It seems to be more than a wire junction housing.

sizecomp.jpg

Maybe cut the thing off and see if it blows up the internal controller at 60v. :lol:
 
Just echoing what I wrote in the other magic pie thread.

gilnet said:
If Hyena was "paid" for review,I can't wait for general customer response

I assure you I was absolutely not paid to do this review.
Like many of the members here I'm an ebike enthusiast and did the review out of my own curiosity and for the benefits of everyone else who is interested. I have another GM motor which I've been very happy with but I'm by no means a GM fanboi and if something better comes along I'll buy it.
I've been a member of this forum for a while now and if you search my posts you'll see I'm not a secret GM spokesman taking a different approach to the others that have visited here!

I've been keenly following the development of the magic pie out of my own interest and was pleased GM has allowed me to pre test one before they're made available to the general public. I believe others have had them for testing too but have not posted their findings,most likely because they are still a work in progress by GM (although now they're going semi-public). I don't doubt that the final or future versions of the magic pie will be tweeked and GM would be wise to take onboard the comments and suggestions from the many experts on this forum and their own.
 
fechter said:
That's kind of crazy to have the red/black reversed. If it was mine, I'd put some black and red colored heat shrink tubing or tape around the ends of the wires, but I suppose the huge + and - labels should give you a clue also.
It's a one off mistake, as soon as I put andersons on there it'll never happen again and I assume it'd be likewise for anyone else who's wiring one up. Once you've actually soldering it to the battery the GM connectors are fool proof (male and female bullet connectors)

I think this guy on the GM forums hit the nail on the head. His acronym is spot on too! :)
bikemad said:
it looks like the same two pin plug and wiring has been used for both the battery and controller side of the connection. This is where the red and black are being transposed, as they are soldered the right way round on the controller end.

It would have been more acceptable if they were wrong on the controller where they remain unseen and correct on the battery end where you can see them. But this might mean the plug from the battery would have an exposed male positive connection when unplugged, which is not good practice either. Better to have the two plugs wired differently for the controller and battery.
Hopefully this will be put right on the final production versions of the Magic Pie.

MAGIC PIE: Must Avoid Generally Inappropriate Colours Polarity Incorrect Externally

How is the motor weight? I forget if the weight spec is mentioned on their website. It should be lighter than a Crystalyte X5.
Definitely. According the website it's 6.2kg, up from 5kg for the old motor. Weight on the shipping consignment for the whole package was 10.Xkg (10.2 from memory)
Delta/wye switching might be an interesting idea too.
Yes, and it's something GM could implement themselves if they chose to in the future. Although you need a bit of space for cooling, there's half a motors worth of space for fitting other goodies :)

TylerDurden said:
Controller looks to be a variation on their BAC-21:

I would assume so. I just checked the specs on the GM website and apparently the pie is supposed to have the wheel locking feature. Maybe mine doesn't have that connection yet because it's a pre-release model ?

dogman said:
20 mph performance is no suprise. For many it will be perfect, especially the internal controller simplifying the wiring

Indeed. It'd be great for europe where they have such restrictions but GM needs to make this feature optional. Better still, have a switch for it so you can elect to limit your speed for legality or economy purposes. As I posted earlier I think GM has alot more plans for this motor and its features.
 
I must be on igmore or something. I havent rode a GM for two years and sat here reading this stuff for nothing. 8)


Can you explain why GM said there was a 48v cutoff and you were using 60v. In your post you explained you were in a hurry and metion no mod to the controller and I figured as you had not found the hall sensors you must not of used your GM controller unless it was their new one and they mention plenty about the performance being down in sensorless mode.

What do you think is going on with this 48v hvc? Does it exsist or is this what the speed slog is about?

Any ideas please as to what is going on here would be nice for all of us to hear.

Or do we just forget this hub for off road use as it only advantage is just being street legal.

Thank you
 
And another thing, missing the wheel cassette was pretty up to GM standards. On their first review too. :roll:
 
Does this explain this type of in your face kick a dog that bites you post

GoldenMotor said:
Hi guys,

try your best to avoid GM when you see this:

Video!
1. Built-in function-rich reliable cruise controller
2. Super thin body (60mm only) for most existing bike frames
3. Large diameter means higher climbing torque and smooth riding
4. Single motor can take multiple battery voltages: 24V, 36V, 48V to get desired power and speed, only need to add more batteries.
5. Plug-and-Run, use easy and professional connectors
6. Adaptable to disc-brake and 6-speed sproket (gear wheel)

It is designed for any existing bike frames, one for all.







Flamers/trolls like KazBlueSky and DannyRay... competition is fierce but... wish you good luck when we launch our next product. See if that hits you hard.

I think not.

I was expecting a lot more from this hub given the gusto GM came in here with. I am reflecting on how GM made me feel being a big supporter of theirs with this attitude and I felt relieved that they really had something.

The review was stating how the hub was pulling the amps and not putting out the speed and seems like its inefficient as hell.
 
Urrrgh, I wrote a massive reply then my broswer crashed :x

317537 said:
I would honestly try the motor on a different controller before judging too soon.

Trying a different controller kinda defeats the purpose, plus it's more easily said then done given it's integrated and all. I'm going to use and test it as GM intended. You dont start out reviewing a new car by ripping the engine out and putting a bigger one in. That's not to say that the soldering iron won't be coming out later on :wink:

It has plenty of poles, 56 I counted, lost count so you could confirm that please.

I spent a few hours pulling it apart and writing the review, the pictures are clear enough count them yourself :p
GM states 56 on their website, I have no cause to doubt them or a desire to count and make sure every last one is there.

It would be sensorless controller and hub as it goes with the no need to fix it theme GM is trying to push with this motor.

I would have thought so, but where are the hall wires going on the controller ? I don't know much about sensorless controllers, if they normally have hall wiring that goes elsewhere in the circuit rather than to actual halls.

liveforphysics said:
It must be electrically speed governed. They are likely trying to keep it in compliance rules by design. I do love the diameter of that motor though!

I hope so, I'd really love to get more out of this motor. Given what is now achievable with programmable infineon controllers GM should be able to make this thing fly if they so choose. (and I believe their earlier testing versions did)

317537 said:
Gm stated it was voltage limited to 48v and anything over would have a cutoff. Ummm this could have a few meanings but my guess is it's no action what-so-ever. I would await for the OP to confirm some more stuff over a few posts, I think Hyena did a pretty good scoop in the small time he had to tinker but to overvolt the internal controller to 60v isnt indicated in his post so Im predicting there may of been some errors. 36 kph at 60v on a 48v limited sensorless controller, if it is sensorless, the dots just are not connecting in this first review.

Check the posts from GM and the website again. They have always stated that 60v is the maximum voltage. Not a 60v pack, 60v MAX. I didnt bother trying my lipo pack off a full charge as I believed their HVC claims. As I said the caps are only 63v so running any more than that is playing with fire anyway. I stated the exact voltages I used for the sake of accuracy - 40v and 58v - you can translate those into what ever chemistry you're running and deduct your own conclusions.

317537 said:
Whats this doovalaky for?
Maybe cut the thing off and see if it blows up the internal controller at 60v. :lol:
I dont think theres anything in there, I suspect its just a junction box to keep it neat. Voltage limiting would be done in the controller itself.

317537 said:
plenty about the performance being down in sensorless mode.
that's true, and a possibility I suppose. If for some reason one of the halls isn't working properly (assuming it isnt sensorless) it could be automatically running in sensorless mode. Mind you it launches off the line without any stutter or hesitation.

The missing cassette may have been an oversight, or they may have been supplying just the bare minimum for me to test it (ie motor and throttle)
It does make testing a bit difficult, I'll try and pick one up on the weekend.
 
Dear Members,

Ignoring my rant yesterday, I have been brought to attention, the ridiculously slow speed of Hyena's magic pie.
This should NOT be the case. Our engineers said to me 5 minutes ago, that the Pie he received was of High RPM, it should go at least 47KPH on a 48V.
I can confirm this myself as I was the first rider of the Magic Pie.
Hyena, you would need to check the battery, how long has it been used? Voltmeter tests and whether the speedmeter is accurate.

Edit:

HOW TO DISTINGUISH BETWEEN HIGH RPM AND EUROPEAN VERSION MAGIC PIE:
The HIGH rpm has FLATS on the axle
The low RPM high torque has NO flats on the axle.
 
GoldenMotor said:
This should NOT be the case. Our engineers said to me 5 minutes ago, that the Pie he received was of High RPM, it should go at least 47KPH on a 48V.
I can confirm this myself as I was the first rider of the Magic Pie.
Hyena, you would need to check the battery, how long has it been used? Voltmeter tests and whether the speedmeter is accurate.

Yao it's definitely not the battery - after testing the pie I put it straight back on my other bike and it was instantly ALOT faster.
Speedmeter was GPS, very accurate and voltmeter is a wattsup meter which is also quite accurate.
We need to sort this out as currently I'm the only one reviewing this. I want to get the most out of it and I'm sure you don't want this down side being the only representation of this motor if its supposed to be capable of more.
 
Engineers doing a test right now for it. This seems strangely abnormal. I shall find out more about it. It was tested by TOM before it was shipped out and it was confirmed to be high RPM. Just one question, did you open the hub cover before or after riding?
 
Back
Top