GoldenMotor.com Forum
General Category => Magic Pie & Smart Pie Discussions => Topic started by: Rob9771 on September 24, 2010, 07:39:19 PM
-
HI
I just set up a MPie 1000w 48v on a rear hub. Straight out the box 40kmh on flat - plenty of torque.
Covered about 30 miles and decided to try it on some rough tracks. No problem for about 3k then motor cuts out.
I still have all lights on trigger throttle and loads of battery. Switched off and back on and no sounds.
Rode home and put hub on USB to try and find out what was wrong. Plugged in again and get no sound , As soon as hub moves I get 2 beeps but no power.
Any help greatly appreciated - I was having great fun until then !!
I have no pedelec, no brakes and no cruise set up.
Thanks
-
Hi Rob and(https://i.imgur.com/evDSMvT.png)to the forum.
Two beeps is usually an indication of hall sensor failure, but the motor should still run in sensorless mode with this particular fault.
You haven't mentioned whether you have the internal or external controller, so I'll try and cover both options.
If you have the external controller, double check the connections on the five thin wires going to the motor, as a poor connection here may cause this fault.
If you have the internal controller, it's worth double checking the connections to the motor to ensure they are properly plugged in, but I think the fault is more likely to be a failed hall sensor.
If the motor was bought from a dealer, you will need to contact them direct, but if it was purchased direct from GM in China, email tom@goldenmotor.com and explain that your Magic Pie gives two beeps , but no power at all.
Try and keep the email short and to the point to avoid any confusion with the translation.
Please let us know what the outcome is.
Alan
-
Hey Rob
Did your pie get soggy at all (wet)?
I rode mine about 20 or so km's in the rain, some water must have hit a hall sensor so I disabled them with the USB software (changed the phase) it eventually must have dried out and is working awesome now.
Have you tried riding the bike for a bit then trying the throttle? It should go into sensorless mode, I was riding on that for a while until I tried the hall sensors again
good luck!@!
-
Tried riding again but no change. Latest development is that if I leave it switched on the throttle lights go out, then turn off and on again
and it tries to move about an inch before dying again. Can I buy hall sensors in the UK as I get the impression GM are not really quick
to respond to problems (not my experience yet but from what I read on here)
Rob
-
There are a few places that do Hall sensors. The HoneyWell SS41 is the better part as it has good temp, voltage and current ratings.
Ebikes.ca have a lesser version that seems to work just fine with their controllers.
You still should be able to get going in sensorless mode. The LED lights out could be a sign of the BMS cutting and or a bad battery connection. Still not related to hall failure.
As the BMS and controller are not integrated to know what each other are doing, A failed hall sensor could have the motor phase work against each other and trip the BMS before auto sensorless mode can work. Maybe reduce you max and continuous amp setting in the software to maybe avoid this problem of BMS cutting you off..
What about charging? You mention numerous things about range but don't mention you actually charge the pack,
You may need to charge the pack right up and make sure it is fully balanced. Even a 5 hour charge may have your pack reading enough voltage to get you going but if the cell blocks are all out of balance the BMS still may complain and cut you off.
A new pack could take three days to properly balance and you are set. Still leave the pack as much as you can on the charger when you use the bike often.
Try leave the pack on the charge for a couple of days and get the pack balanced perfectly then see if it cuts out on any load. Reduce the max and continuous amps to 20 amps max and 15 amps cont to see if you can get sensorless mode running. And as Alan has already said check all your connections for and intermittency especially on you battery and throttle connections.
Is your controller internal or external?
-
Ohhh and BTW sensorless mode may require you to pedal first and get the bike moving for it to work.
-
In detail - Internal controller -
I have always needed to pedal start from day one - I have not connected cruise, brakes or pedelec.
I have charged the battery but it usually switches off after a couple of hours - should I leave it connected so it can switch on again or is that the end
of the cycle?
AT present I have 3 full lights on throttle and 2 beeps if I move the hub slightly. Tried pedalling up the road but no power.
I have changed settings in controller ( I think) Not entirely sure whether it is changing anything but it does beep when I push connect.
Do you know of any instructions for the USB controller software or an explanation of how you know it is doing what it says ?
I have just ordered a front hub so I can duel it so I am committed to getting this working.
Thanks for your advice
Rob
-
Sensorless mode may indeed trigger the cut outs if the hub gets jammed up with no pedals to start, but this I am not entirely sure of. I would advise keeping the current down to the motor until you get this sorted. Any overload will make the BMS kick off the controller. This will be indicated to loss of throttle lights.
Some of the solder points maybe poor inside this controller and this could create a low drop on fet side of the shunt. Again we are not sure as the controller is over complicated IMO with all theses weird beeps that really seem to mean nothing if the controller hasnt been soldered properly to begin with.
Is there some beeps that indicate crappy controller soldering? Otherwise the beeps are presumptuous to the fact that there is a normal malfunction such as hall sensor failure. I would rather a controller that doesnt fail over a controller that always fails and has beeps to tell you something as obvious as hall failure or fet failure as usually the solution to all and any other problems is to check your connections or get a new controller. So why do we need beeps to tell us the same stuff.
100v fets and 100v caps for the regen controller would be a start, IMHO the regen voltage between the motor and the shunt can exceed the capacitor rating making it fuse. On a full charged pack there is only about 7v difference between a full pack and the 63v cap. This is designed to fail IMO. THAT(S WHY THEY ARE FAILING!
My suggestions are always ignored so we still see much more failures.
You may need to get an external controller with a reputable track record. Yes something with at least 100v caps.
I get the feeling the hub isnt giving you anymore power. Maybe something inside the hub has shaken lose or a poor solder point has failed during your off road use.
If GM doesnt make good on support youre are going to have to deal with this problem yourself.
Meaning open your hub, remove, the internal controller, run hall wires and phase wires out side to an external controller, with the right gear the motor should be a pearl,
-
If the controller shunt is limited to 20 amps or even 25 amps and regen outputs 30 amps or 35 amps at any stage. There is 5~15 amps remaining between the motor and fets with no place to go, so the voltage will increase between the motor and shunt due to the inductive behavior of the motor. Your guess is as good as mine how many volts over your packs voltage this could equate to, so REGEN spikes are a good enough reason to suspect why we are seeing the failures.
Edit.
Its good practice to limit your max amps to the maximum your motor can pull other wise the controllers regen can allow more energy from motor before the shunt than will go back to the pack.
I have two hubs in route to my home. Pre tested by GM CA, Gary reports that one hub delivers no more than 20 amps despite the settings in the software where as the other hub can give 25 amps.
The problem hub will be the 20 amp hub. Other than pull the 20 amp hub apart and look for the bottle neck, it is best I limit the max amps on the 20 amps and set regen to not give more amps than 20 amps or else the voltage can and will spike above the pack...
-
so REGEN spikes are a good enough reason to suspect why we are seeing the failures.
But can you explain why controllers also fail on minimotors that don't generate any voltage whatsoever?
Alan
-
Do they. Where?
I thought most folk burn the mini motors over excessive voltage.
-
Hi everybody, and thanks for your nice info!
Seems to me Gm has a quality problem with the controler: Same problem after 1 month of normal use: 5 beeps, wheel hard to move.
The day it stopped the controller was quite warm -not hot. Temperature problem? I was driving flat and very slow.
Does anyone know
1. Other controller in the market fitting their motor? I read somewhere Gm one is a copy from Crystalite
2. A guide to replace the damaged part (if its a known one, mosfet or so)
I'm runnig minimotor kit, 250watt, 36v, 10Ah. Good results till now! >:(
Thanks everybody
Hello guys,
Just like it happened to you, I had the same problem: 5 beeps, wheel hard to move.
I've used an ancient controller and the motor is fine. The problem really comes from the controller.
GM STAFF: Can you guive us a guide to replace the damaged part? I'm sure you already know the problem. I see that it's very usual.
My setup:
250W Mini motor
Magic controller (broken)
2*5S LiPoly batteries (37V)
Thanks for an answer,
Luís
Both of these are minimotors!
Alan
-
2 posts.
Just because a fet fails doesnt mean its got to be REGEN.
Now can you tell me where the regen energy goes if there is a bottle neck in the controller?
What happen if some one sets their controller to 50 amps max, BTW is that 50 amps default, and regen to 100% and there is a 20 amps shunt in the way.
Just an exaggerated scenario but why has the regen controllers with reputable track record all using 100v fets and caps.. Why everyone with a brain on ES who develop this tech recommend desoldering 12 75v and 63v caps and replace them with 100v parts.
Maybe they at ES don't know what they are talking of and thesze two threads are more evidence of the 75v and 63v parts reliability.
Even a diode over a relay coils is recommended to short the spikes from inductive voltage spikes!
-
BTW can someone inform me,
What happens to the mini motor stator bell fly magnet ring when regen braking is engaged for motor cut out. Does its high RPM disappear into magic thin air?
What is the rpm of the inside of minimotor and does it stop suddenly when the brake switch is applied?
I think just because the rider doesn't experience the back torque of the regen doesnt mean that there is no current coming back from the motor.
Being a geared motor I suspect the mini motor has high RPM and the sudden stop could cause a spike. But not knowing anything about the physics, a non owner of this product, I can only guess the RPM going down hill could be surprisingly higher than any other direct drive motor and the inertia could add up to a short but high current spike meaning this could add up to volts.
-
I read somewhere that these geared mini-motors can generate spikes in high speeds. I have a mini-motor with a CA on my winter cutie, I’ll test this as soon as I got a nut to my rear wheel. It looks impossible though to find a nut whit the Sram thread FG 10,5.
Peter
-
I read somewhere that these geared mini-motors can generate spikes in high speeds. I have a mini-motor with a CA on my winter cutie, I’ll test this as soon as I got a nut to my rear wheel. It looks impossible though to find a nut whit the Sram thread FG 10,5.
Peter
The forum thanks you and loves you.
Sorry again for my indiscretions, but it makes me mad when people like myself, you, Alan and the many others here grind their fingers to the bone when GM cant even answer Emails.
-
Im not sure regen is the culprit here either. As some of you know I have 2 pies on my bike and even with a rolling weight of over 400 pounds (yes 400 counting me the bike, 2 motors, 2 batteries my trailor and all my gear ) when I hit the brakes the most amps I have seen going back due to regen is about 8 amps. This is a reading I am getting by watching the Cycle Analyst while I brake. Of course the CA my not be able to read quick spikes but I should see one once in a while if it is happening. Does anyone else with a CA have anything to contribute on what amps you are seeing during regen?
Gary
-
Wow Gaz, yes I would be suprised to know that too
I was wondering this myself running 2, can I ask you what you ended up setting your regen for both wheels for?
-
Wow Gaz, yes I would be suprised to know that too
I was wondering this myself running 2, can I ask you what you ended up setting your regen for both wheels for?
The 8 amps I saw going back in was when I had them both set at 100 percent. I have since changed the front to 40 percent and the rear at 80. I had to do this as I almost wiped out a few times when they were set so high. The highest amps I see going back at this setting is about 6 amps.
Gary
PS who is Gaz?
-
The culprit I believe.
Is
1: Poor connection to the battery.
2: Poor configuration of Regen and cont~max amp setting
3. Poor solder points in the controller or BMS power train.
With only some milliamps to the fet gates Regen is variable to many amps I believe via the software settings.
Like, when you see these posts with "I cant get more than 15 amps from my MP" this, is the warning sign and keep your eyes out for them guys because it will be "My MP is broken" in the next day or so.
These people for the love of just being on the road for longer must set their software within the limits of the hardware and find the bottle neck along the power train if they want the full MP experience.
I have measured amps to / from the pie with a clamp on ac/dc tester. It is 0-100A 1,5%
Results are a bit suprising
Forward: Max amp 15A (checked this by setting max continous to 16A. No noticable change in performance!)
Regen 80% : 16A
Regen 60% : 10A
Regen 40%: 8A
Regen 20%: zero , nothing
It an all GM setup; Magic Pie rear, internal controller, 36V16Ah GM battery.
(I also have to add the Hioki clamp on does not have a min/max register; Hard to get an accurate reading....Reading given with no decimal)
As you see here there are more amp coming back to the battery than going to the controller, this dictates that the motor voltage is higher than the battery volts.
At 42v 15 amps
42/15= 2.8 ohms.
2.8 ohms * 16 amps = 46v
If Jensenff were to set his Regen to 100% he may get 18 amps
So 18 amps * 2.8 ohms = 50.4v from the motor.
50.4v/42v=1.2 multiplier.
A 48v pack is 54.6v
54.6v * 1.2 = 65.5v This is over the capacitor voltage rating but not over the fet rating.
I really can not nail the 48v pack without the same bike on a 48v12ah pack, as the resistance could be the same or even is a little higher on a 12ah pack its hard to do the math.
But in essence you can see how a bottle neck in the system and or incorrect software configuration can indeed lead to higher volts on the motor side.
Even a higher voltage capacitor would make the Magic controller much more reliable IMO. Higher voltage caps do have a lower ESR meaning the resistance is lower, to voltage spikes this is better.
-
Folks that buy the 16 cell lifepo4 pack need to be especially careful. Be it GM brand LFP-4820 or any other brand, as the max charging voltage on such is a a few volts higher.
Why cant GM listen to me and the others I gathered this info from for once.
I think for a few more dollars in build quality ot wouldnt hurt anybody but results in controller that costs $5 more that could be charged to the customer upon purchase. Nobody would lose.
The customers get better parts GM get a better reputation, and it would be a win win situation.
All I asked for was 100v fets and caps. Not even close to what other have got at request.
.
-
I getting a 24 inch pie and hooking it up to a mighty little 24 inch MBT MX thing...... wow they look awesome.
(http://www.goldenmotor.com/e-Bike-DIY/MagicPie Series.jpg)
^^^^^^^^^^^^^^^^^^^^^^^^^^
SICK.
Can we get 100v fets please. ??? ??? ???
Request denied!!!!
-
Leslie mate I think you may be a bit confused, I'm sure Alan has always 'liked' you silly!
You have some good points about the battery etc, my thoughts is that maybe there is a switching problem with possibly the caps discharging back through the phase when the brake is released, enabling the throttle. I can remember clearly as stopping at the lights I let off the brake switch and the wheel started cogging and I thought at first I had a flat, then I heard the beeps...
Could that current or fast switch from whoah to go be the cause of the problem I don't know. What I do know is that I ride my bike everyday and since really miss 2 pies!! having 1 now is like having 1 leg
-
Leslie mate I think you may be a bit confused, I'm sure Alan has always 'liked' you silly!
when the brake is released, enabling the throttle.
I think Regen works by reversing the gate driver function to the opposite switching. So when normally the gate driver is open it is closed and vice-versa. The percentage of regen is delivered by the duty cycle of PWM.
Maybe at this point of brake release, the resistance between the motor and shunt raises at a sharp peak due to internal capacitance of the fets draining off in a millisecond and the phase current cap should be full of regen energy offernig little resistance for any spikes to be absorbed.
After a few shots or more the phase cap has energy punch micro holes into the insulator between the plates and resistance between the phase cap positive and ground increases until the cap fuses, this in my mind results in 100% regen to ground before the battery yet regen on the next operation parse takes out the capacitor and intermitency makes this happen.
(http://goldenmotor.com/SMF/index.php?action=dlattach;topic=2499.0;attach=3455;image)
The full circuit above is like massive so I can only give a screeny of the part that matters.
There are a few things that come to mind but to trace it and simulate it is quite a challenge.
BTW that spike is reading 40v over the 52v battery voltage and no disconnection is happening to the charging 52v battery. It was only on a few time I "'released" the short to the cap I get this spike too sometimes it was a drop of 40v.
You know what happens when cap fail they behave like they work and then present a short circuit. When and how is anybodies guess. They can even test like there is nothing wrong with them until you put them to work too.
To be honest im confused as all hell now as there isnt much going for the regen idea and a lot of thing happen that can make it fail. Maybe not a single thing but a multitude of events causing failures,
The above sim is of a pulse similar to a regen pulse from the motor generated by a 555 timer 12v output to the gate of a fet, rail is 52v to a 1000uh inductor (motor) then to a fet drain, to sim pulses, and then the fet output is transimtted to a fet bridge config like our controller, with a short swtich over the cap onto a 52v battery. The 2.8 ohm resistor offers some idea of the resistance in circuit to the batt.
Im not going to simulate this idea to 100% accuracy as I am not 100% sure of all the electro mechanics involved but I put to the pro engineers, who is?
If this was a proven technology and everything is known in a single thought of tracing and all inconsistancies were 100% understood then we would not be here discussing why the magic controllers are failing.
All I can offer up is that the higher volt parts must be more resistant to what we are seeing in this simulation.
-
' you silly!
People have always said we both work well on different ends of the table, IMO covering each others weaknesses in support.
But thats the problem I feel, while I am chesse he is chalk. I don't think we will ever be the same person nor replace one another.
But here is the $1000 question where do you fit in Monkey. So much skill your posses. Maybe youre where the opposites meet..
What got up my nose is I was made a tech official here with absolutley no up to date information on the newer products nor the funds to aquire them. I felt like a fast runner with a missing leg waiting to be shown up for a fool at the starting line..
-
What got up my nose is I was made a tech official here with absolutley no up to date information on the newer products nor the funds to aquire them. I felt like a fast runner with a missing leg waiting to be shown up for a fool at the starting line..
I don’t understand why GM isn’t taking care of such a good support / advert / salesperson as Leslie.
If they could provide him with technical information and/or some of the products, they would have it back greatly.
This forum would be nothing without Leslie and Alan, even if there’re a few more sharp guys in here (or girls).
GM Please listen to this
Peter
-
Hmm I guess I just fit in because I like this forum, and the fact that I'm the only person I know that has a ebike!
I think GM are even a bit too busy ATM to do anything... Maybe National Day holiday is big one as some of my contacts took this whole week off. I've noticed a fair lack of comms from china in the past week or so.
But still, understand where you are coming from Les. And what can ya do hey :( frustrating
-
Monkey.
Well a simple mod would be to replace two caps with 100v parts at least with the external controller. Im sure this would be hard to mess up. Im sure I could manage to fail at it.
Frustration?
Funny thing is Im not as good with my hands as I am with my mind. Pins and needles in my fingers from child hood, loss of eye sight to age sometimes makes life difficult for me. I have an exceptional memory like an elephant the organization skills of a derelict homeless and cant find my glasses half the time., computer skills and understanding how to learn is not an issue for me, This is why theory, science, physics is so attractive to me.
Peter thanks for the compliment.
I don't know, :-\ I think if someone offered me free products I would refuse, weird I know. I always like being free, and getting things free always seems like a ball and chain, I could take them but I would be trying to offer 10 times in return, You got that much right. But I would suffer much in real life.
I always have problems in employment, as I work myself into the ground. I can never pace myself in duty in any correct manner and always end up at odds with someone. I always believe that the workers time is such value that money can not buy, loyalty and respect from the employer is worth more.
However this is not how the real world works. I don't get hierarchy culture in the work place and think it just an excuse for classicism.
Sadly it may be too late for me here as usually when this time comes and Im talking like this I cant change course, I am hoping this time, in this forum I can learn how to settle down. This is something I am trying to learn, however when the thorn is in my side. EG I see the no email response threads, it is bugging me and making it hard for me. I will go the boss like a pit bull until I'm put out of commission.
No no!
GM are reading all of this and I am positive as I was asked to edit my posts via PM, I am wondering why they keep this forum going but I know if this forum goes so will the GM business.
As this is the customer support and the good peoples assistance here are sustaining the product sale not the flashy crappy web stuff and expo's.
Wana prove me wrong about this one too GM? try it.
Ping interacts on ES.
Keywin interacts on ES
Justin from Renaissance interacts on ES.
John from CA interacts.
Most of what I post here is in some form is collated and learned from the techs from ES.
GM opposes ES and sees them as an enemy or understudy.
Here is a thought! Swallow the pride. GM get on ES and find out from a larger community as to what to do with the controller and DO IT.
Hey why doesnt GM just make a few cheap and easy changes to some things and be done with worrying about all the warranty emails.
-
The irony is that in simple terms for those who have right brain thought domination the voltage must be higher from the motor than the battery for any current to charge the battery. The caps rated at 63v and controller limited to 60v and the battery GM battery charges to 56.4v and leaves 3.6v head room to stay within GM magic controller specs @ 60v and 6.6v to stay within the capacitors specs. Thats a Brazilian with Lazar technology close to make anything viable.
Exactly how much volts has to get up from the motor to deliver 15 amps 20 amps 25 amps to the battery? Exactly how does this EMF sensor works or could work is beyond me. From my perspective such a thing as cutting regen with 25 amps soaking the circuit would more than likely create more problems than good. As amps shows there is resistance on the motor.
Can anyone yet report regen cut off before fet damage? And if its not regen thats taking fets out then what could it be?
-
In detail - Internal controller -
I have always needed to pedal start from day one - I have not connected cruise, brakes or pedelec.
I have charged the battery but it usually switches off after a couple of hours - should I leave it connected so it can switch on again or is that the end
of the cycle?
AT present I have 3 full lights on throttle and 2 beeps if I move the hub slightly. Tried pedalling up the road but no power.
I have changed settings in controller ( I think) Not entirely sure whether it is changing anything but it does beep when I push connect.
Do you know of any instructions for the USB controller software or an explanation of how you know it is doing what it says ?
I have just ordered a front hub so I can duel it so I am committed to getting this working.
Thanks for your advice
Rob
If I lived in the UK I would offer help. Even if it was practical I would buy the defective wheel from you for an honest price as the controllers aint worth much new IMO.
-
Hi
I have same problem like you but I do not use throttle I try to drive my gm 5k pot. when I give power to controller I hear 2 beeps and no work my hub.
could you help me?
-
Hi Cagatay and(https://i.imgur.com/evDSMvT.png)to the forum.
Two beeps usually indicates failure of the hall sensors inside the motor, if you can give us a bit more information it might help:
Which motor do you have?
How old is it?
Did it work OK before, and has recently stopped working, or has it never worked at all?
Have you tried spinning the wheel to see if it then starts to run under its own power?
Alan
-
hi
? have 36 v 250 watt rear motor and it is almost 4 months old. it worked before a lott before but now not ? have ordered a throttle ? am expecting my throttle when it comes ? will try again.
I have not tried spin wheel also.
can that be any problem about motor hall effect sensor?
thanks your reply...
-
Hi,
I have the same problem - two beeps on connection the battery to the controller. Actually these are only two wires connected to the controllers and these wires are from the battery...
What's wrong? How the problem might be solved? What're next steps?
Thank you