GoldenMotor.com Forum
General Category => Magic Pie & Smart Pie Discussions => Topic started by: Geoff on April 27, 2010, 12:06:50 PM
-
I have 2 MagicPie's with 36V16A LiPo batteries purchased in late Feb 2010. The batteries finally arrived last week and I'm generally very pleased with the results. However I have 2 questions...
1. I have installed PAS detectors on both bikes. And PAS operates by default on both. Without specifically turning the PAS on. In fact there seems to be no way to turn it off. After 2-6 turns of the crank the motors kick in... and power up to full straight away. Its a bit disconcerting. I'd like to be able to turn the PAS on and off without physically disconnecting the sensor.
2. Occassionally, after using regen braking one motor will stop functioning. If I turn it off and back on, (at the key), the hub will beep 6 times, the charge lights will come on but the throttle will not respond. If I leave it for about 5 minutes it will turn on normally and function fine. It doesn't seem to be overheating. I can place my hand on the motor housing and it will just feel warm.
Any ideas on these?
-
Geoff,
If you haven't used the throttle switch for lighting, it should be possible to use it to turn the PAS on and off by cutting the Red +5V supply lead going to the pedelec sensor, and connecting the two cut ends to the brown and yellow switch leads (make sure the joints are safely insulated).
The cutting out you describe does sound like it's temperature related. Although the hub casing may feel only warm, the stator inside the hub will be much hotter, but as is separated from the cover by air, it will take a while for the air to conduct the excess heat away from the stator. Hence the 5 minute wait. When it cools down to a safe level it will then work again as normal.
Is the one that's cutting out being ridden faster or does it have have smaller wheels or carry a heavier load by any chance? As all of these factors will have a noticeable effect on the amount of current (and also heat) being generated by the regenerative braking.
Alan
-
Thanks for the response Alan,
If you haven't used the throttle switch for lighting, it should be possible to use it to turn the PAS on and off by cutting the Red +5V supply lead going to the pedelec sensor, and connecting the two cut ends to the brown and yellow switch leads (make sure the joints are safely insulated).
I'll give this a go. Though I wonder why its on by default. It seems to be counter to the documented behaviour. It would also be great if the amount of power applied to PAS could be controlled. My wife is getting back on her bike after sometime and finds it scary when it just takes off at full speed with no way of moderating the power.
The cutting out you describe does sound like it's temperature related. Although the hub casing may feel only warm, the stator inside the hub will be much hotter, but as is separated from the cover by air, it will take a while for the air to conduct the excess heat away from the stator. Hence the 5 minute wait. When it cools down to a safe level it will then work again as normal. Is the one that's cutting out being ridden faster or does it have have smaller wheels or carry a heavier load by any chance? As all of these factors will have a noticeable effect on the amount of current (and also heat) being generated by the regenerative braking.
I suspect that you are right here. I'm pretty heavy at present (110kgs) and the bike has 700c wheels. I'll ease back on using the regen. I've put the regen on a seperate switch instead of using the supplied brake levers.
Can anyone confirm if there is a way of changing the amount of pedal assistance supplied by the motor?
-
Hi Geoff.
I can confer with Bikemad.
Those red buttons on the throttle have been used as PAS bypass on a few different bikes.
I would rid of it all together.
The point of PAS is to make people rely on the pedals by making the throttle unreliable. Its a case you must pedal the bike because its got pedals but if you ever need the motor to save your dear life ever forget about it. They say its safer not to have people rely on a motor that may fail.
Not using PAS IMO doesnt create a rely upon the motor at all times situation.
You just pedal when you need, the absence of PAS doesnt cut peoples legs off.
Ouch 110 kgs on a 28" wheel. Your combined weight and bike = more than mine with 35kgs of lead strapped to my frame.
But I can tow another 50 kgs on top of that with a 26" wheel on a HBS..
Alan.
The regen on 28" wheel is perplexing me here.
The larger radius of the wheel obviously relates to torque vs regen current supplied to the FETS gate. The 28" wheel would be more wheel torque vs regen, where as a 16" wheel would amount to more RPM vs regen.
How does the differences relate in the real world to the end user?
Those fets are a wide channel for current each.
Could it be when in regen mode, high RPM (16") stresses the controller and high torque (28") stresses the motor?
-
Hi Geoff
I also would say just unplug it. I had it hooked up on my first conversion and also found it difficult to use. It would kick in at full throttle after a little peddling and the scary part was it would continue to go full throttle for a few seconds after releasing the throttle.
I vote, unplug it!
Gary
-
Ouch 110 kgs on a 28" wheel. Your combined weight and bike = more than mine with 35kgs of lead strapped to my frame.
But I can tow another 50 kgs on top of that with a 26" wheel on a HBS..
This weight should still not impede the use of his Pies. I did a rough calculation of my weight, bike weight, conversion kit with 2 batteries and gear carried to go to work and it came out to over 160kgs. Thats on a HBS 48v1000w , the pie should still have more torque and should still be able to handle this. I'm not recomending this but sometimes reality takes over.
Gary
-
Ouch 110 kgs on a 28" wheel. Your combined weight and bike = more than mine with 35kgs of lead strapped to my frame.
But I can tow another 50 kgs on top of that with a 26" wheel on a HBS..
This weight should still not impede the use of his Pies. I did a rough calculation of my weight, bike weight, conversion kit with 2 batteries and gear carried to go to work and it came out to over 160kgs. Thats on a HBS 48v1000w , the pie should still have more torque and should still be able to handle this. I'm not recomending this but sometimes reality takes over.
Gary
Thats what I meant. If my meagre HBS36 500 watt (200 watteva) can do the weight it pulls home some days a 28 inch pie should find it easier. And doesnt he have dual drive?
Right now I think we should look at a few things more carefully and explore all the possibilities of whats going on here.
Not a bad idea but some variables cn be explored. Like the rubber brake for the front rim's brake switch slows the rear and the reversed.
Crossing the braking action can make it smoother for both wheels. If regen and brakes are applied over the frame evenly the regen maybe shared more evenly.
Geoff
I have 2 MagicPie's with 36V16A LiPo batteries purchased in late Feb 2010.
Interesting combination.
Geoff can I ask a few questions.
These are Lipo not lifepo4?
Two packs or one and how are they configured to the controller?
This is a dual drive?
If, so it wouldn't be by any chance its the front wheel cutting out? ::)
And what is the C rating on you lipo's?
Thanks.
-
Hello, just for the record, I've got my Magic Pie just a few weeks now. My bike with 700C front wheel Magic Pie and 4 12V SLA 10Ah batteries is just over 50kg. Add my weight (75kg) and the kids trailer (50kg) and the pie has to pull over 170kg. Fully loaded at 24V I'm getting 20km/h and a range of +50km. At 36V 30km/h and +30km and at 48V easily 40km/h and +20km.
At the start the batteries give 54V and I must say the torque is huge. Without trailer and with heavy pedaling I easily reach speeds of +45km/h. I didn't install the PAS and find the regenerative braking hard to use as it slows down the bike real quick (almost immediately to full stop). Nevertheless it's so much fun to pass by and leave behind a group of trained cyclists with two kids laughing at them in the trailer...
-
I would not hesitate to go for the Magic Pie, because it has plenty of torque making it ideal for your needs.
The Pie may be heavier, but its large diameter and extra poles/magnets give it much greater torque.
I must say the torque is huge.
I was just going to say that you made the right decision in buying the Pie, but it sounds to me like you might already know that! ;)
Alan.
The regen on 28" wheel is perplexing me here.
The larger radius of the wheel obviously relates to torque vs regen current supplied to the FETS gate. The 28" wheel would be more wheel torque vs regen, where as a 16" wheel would amount to more RPM vs regen.
How does the differences relate in the real world to the end user?
With a small wheel, the regen is far more effective at actually slowing the bike down quickly, so it only works harder, but for a shorter period of time.
A larger wheel exerts more leverage, making it more difficult for the regen to slow the speed down, and it therefore has to work for a longer period of time, but at a reduced rate compared to the smaller wheel.
The velocity, gradient of the hill and the total combined weight of the bike and rider will ultimately determine how much load is put upon the regen. The higher the load, the greater the generated current (and also the heat) will be.
I think the voltage generated at high speed (before regen is used) may be more of a problem for the controller with a small wheel than with a big wheel. The voltage being generated will be relative to the wheel speed, therefore a 16" wheel will produce ~75% more volts than a 28" wheel running at the same road speed. If the voltage goes too high, something is likely to fail. (ie. the FETs or capacitors etc.)
I have 2 MagicPie's with 36V16A LiPo batteries purchased in late Feb 2010.
Interesting combination.
Geoff can I ask a few questions.
These are Lipo not lifepo4?
Two packs or one and how are they configured to the controller?
This is a dual drive?
If, so it wouldn't be by any chance its the front wheel cutting out? ::)
And what is the C rating on you lipo's?
Thanks.
The clue was in the initial post: ::)
I have installed PAS detectors on both bikes
I suspect the batteries are LiMn not LiPo. ;)
Alan
-
Hello, just for the record, I've got my Magic Pie just a few weeks now. My bike with 700C front wheel Magic Pie and 4 12V SLA 10Ah batteries is just over 50kg. Add my weight (75kg) and the kids trailer (50kg) and the pie has to pull over 170kg. Fully loaded at 24V I'm getting 20km/h and a range of +50km. At 36V 30km/h and +30km and at 48V easily 40km/h and +20km.
At the start the batteries give 54V and I must say the torque is huge. Without trailer and with heavy pedaling I easily reach speeds of +45km/h. I didn't install the PAS and find the regenerative braking hard to use as it slows down the bike real quick (almost immediately to full stop). Nevertheless it's so much fun to pass by and leave behind a group of trained cyclists with two kids laughing at them in the trailer...
:'(
Thats awesome.
I have a dirt bicycle comming together with a 20" wheel at the back.
I hope to own one soon too.
-
1 MP per bike.
1 GM supplied 36V16A LiFePO4 battery per bike.
One bike is a Giant Suede (26" wheel) http://www.giant-bicycles.com/en-au/bikes/model/suede.slate.blue/5321/39386/
The other bike is a Giant Cypress 2 (700c wheel) http://www.giant-bicycles.com/en-au/bikes/model/cypress.2/5329/39116/
The Cypress 2 gave me around 50k range over varied terrain with speeds on the flat of around 25-30kph I'm pleased with that.
It's cut out on me 4 times now with around 150K of use. Each time it dies with no warning and comes back after about 5 minutes of rest. After it fails it will beep 6 times. I can't find in GM's documentation, or on their website, or within this forum what 6 consecutive beeps means as a diagnostic code.
After it has failed once and recovered it will then continue to work for up to 30 minutes without problem. From memory each time it has failed has been on flat terrain but after using regen braking.
I'm happy to accept that overheating is the problem but some information about the diagnostic beeps would be helpful
-
Its not much point trying to convince me on the facility of PAS or the desirability of turning it off. Its a given for me that it work as advertised.
I've ridden other bikes where the PAS works very well. It was part of the MP feature set that helped me make my purchasing decision.
Currently it doesn't work as advertised. I'm wondering if anyone knows if there is a reason why this is the case and if there is a way of making it work properly.
Switching it on and off is one solution, but a kludge.
Making it work as advertised, (ie turn on by pressing Cruise twice and off by cycling power) would be better.
Best of all would be finding a way of changing the amount of assistance it provides.
Has anyone else noticed this behaviour and interested and wanted it to work properly?
-
Thats what I meant. If my meagre HBS36 500 watt (200 watteva) can do the weight it pulls home some days a 28 inch pie should find it easier. And doesnt he have dual drive?
Whoops! Sorry Leslie, I guess there were to many beer bottles in front of the screen and I didn't see your entire post.
Gary
-
As far as I know, the only PAS that GM has that actually works is on the "Cruise Ebike". If you look at the picture you can see there is some sort of hub between the pedals and the crank.
Gary
-
As far as I know, the only PAS that GM has that actually works is on the "Cruise Ebike". If you look at the picture you can see there is some sort of hub between the pedals and the crank.
"True Torque Sensor Pedelec" I wonder what that really means and if its for sale seperately?
-
"True Torque Sensor Pedelec" I wonder what that really means and if its for sale seperately?
Geoff, it basically means the power provided by the motor is directly related to the force applied to the pedals, the harder you pedal, the more the motor will assist. The torque is measured using permanent magnets and hall sensors contained within the housing on the back of the chainwheel.
I've only seen these units fitted to single chainwheels, so it might be a problem for anyone using double or triple chainwheels.
Some of these units can be adjusted via an external unit to vary the ratio of Pedal-power : Motor assist-power.
(1:0.5, 1:1 or 1:1.5 etc.)
(http://img.timesfirst.com/picture/675337/675337_39687489927118166.jpg)
The basic PAS units used with the Magic Pie are only able to sense the speed of the pedals, not the force being applied to them.
On the early Pies, the amount of assistance (although relative to pedal speed) tended to be a bit too much.
The more recent Pies have a different controller which can be programmed (using the optional interface lead and free software) to allow the selection of Low, Medium or High Pedal Assistance, but apparently there seems to be a problem getting the controller and PC to actually communicate with each other. ::)
Perhaps Yao can provide some more information on the GM "True Torque Sensor Pedelec" and let us know if the unit is compatible with the Magic Pie, and how much it would cost to supply it separately.
Alan
-
"True Torque Sensor Pedelec" I wonder what that really means and if its for sale seperately?
I'll look into that...
Gary
-
The more recent Pies have a different controller which can be programmed (using the optional interface lead and free software) to allow the selection of Low, Medium or High Pedal Assistance, but apparently there seems to be a problem getting the controller and PC to actually communicate with each other.
So how do I know if my MP's will work with this controller? They were ordered from GM on Feb 22 2010.
-
"True Torque Sensor Pedelec" I wonder what that really means and if its for sale seperately?
Geoff, it basically means the power provided by the motor is directly related to the force applied to the pedals, the harder you pedal, the more the motor will assist. The torque is measured using permanent magnets and hall sensors contained within the housing on the back of the chainwheel.
I've only seen these units fitted to single chainwheels, so it might be a problem for anyone using double or triple chainwheels.
I've got one of those on my bike. Although its disconnected and I don't use it. It came fitted on the bike I bought.
(http://i908.photobucket.com/albums/ac289/muzza_au10/newbike.jpg)
It was only 24V with a geared hub motor. So I fitted a direct drive 48v 1000w motor and 60v worth of batteries.
It looks like this now.
(http://i908.photobucket.com/albums/ac289/muzza_au10/30012010204.jpg)
But anyway, the pedals seem to be spring loaded. When you push down with your foot the pedals turns a few degrees before cog starts turning. When I'm pedalling it feals strange due to the springyness. But with 60v of power, I don't pedal too often. ;)
Muzza.au
-
So how do I know if my MP's will work with this controller? They were ordered from GM on Feb 22 2010.
Unfortunatly you would need an MP2 for that and they are not released yet. >:(
Gary
-
The pie has worked other controllers and will work with most GM controllers for Ebikes with 120 deg phase angle.
The pie has an internal controller and if you want to use the interface you need to replace this, or wait and see if GM makes an interface for MP1.
Similarly to slow the regen, place in series on your battery output lead, a low drop high current diode to allow power through to run the motor and block regen from getting to your pack.
Some resistors or a regulator to bypass the reverse current across the diode.
Like this.
Edited:
Battery output +|---->|--| Controller + input
Battery output +|----w---| Controller + input
The diode and resistors will allow much current through it one way and only allow current through the resistors when regen is charging tha pack the other way.
-
I had been thinking of doing that but I wouldn't have a clue what size resistor to use. also I would be a bit worried about to many volts on the controller in regen. I understand that you should not disconnect the battery when using regen as this can damage the controller to high a resistor may have the same effect. I did wonder if a zener diode across the pairs may prevent damage.
-
Unfortunatly you would need an MP2 for that and they are not released yet.
sigh
I'll learn to live with it.
-
I had been thinking of doing that but I wouldn't have a clue what size resistor to use. also I would be a bit worried about to many volts on the controller in regen. I understand that you should not disconnect the battery when using regen as this can damage the controller to high a resistor may have the same effect. I did wonder if a zener diode across the pairs may prevent damage.
Thats easy.
E/I=R
Your pack voltage at full charge divided by your desired current.
Say 54v/5 amps is 10 ohms.
The regulator works well here.
The LM338 can limit the can limit the current by limiting the voltage when the load creates a voltage drop over a resistor on output and its adj leg.
(http://www.reuk.co.uk/OtherImages/lm317t-current-regulation-circuit.gif)
I would limit the current to 4.5 amps per chip but here it states with a big heat sink it will take 8 amps.
http://www.reuk.co.uk/LM317-Current-Calculator.htm
However I'm unsure how the LM338 output is going to take the the battery input. So a Schottkey diode between the LM338 output and pack may be a good thing.
The above circuit is a current following mode design. A way to impede energy and power from regen fet to pack, and heat can be created away from the main circuit.
Works good for the 28" wheel that gives high current and lower RPM regen voltage,
A 16" wheel would benefit from a voltage regulator.
A more efficient approach would be to use an inductor. The motor is inductive power source and pulses energy power into the pack.
Boosting the voltage by creating magnetic fields from excess current can work as a current converter for regen to get the volts up to actually charge the pack.
And bucking can convert voltage into current.
Say a 16" wheel is spiking the fets with high volts in regen. You may need the topology right. So you put a buck on the phase wires to only work in regen.
Or the 28" wheel is creating too much current and its hitting on the motor, a boost can get your voltage from current.
How does regen work.
Interesting how even when the motor isnt spinning fast enough to create a viable charge voltage for the pack, it still causes resistance on the motor.
Wouldnt the lower RPM/voltage of regen in this situation cause current to still flow from the pack into the motor and regen causes current to flow from the motor to the controller?
IMO something in between is going to cop it.
Voltage only travels down hill regardless of the presence of current. And the current has nowhere to go unless it has a lower voltage before it. Yet we still see regen resistance on the motor at a snail pace.
Where is this current going if its not going into the pack?
This is where the NChannel may lets us down, It has a diode that provides a path the ground even when its off or on or reversed.
Using a Pchannel fet for regen may separate the motor from the pack when the regen voltage is lower, as the circuits path to ground should be blocked by the diode before the motor not after.
.
Something in the below picture doesnt look right.
Between the motor and pack there is a 16v difference and the fet when closed is allowing current to flow to ground. The voltage must manifest on the N channel gate or in the motor coil.
The resistance of the fet is lower than the pack and higher than the motor coils and the 120 amp fet should take much more current than the motor. I vote the coils are going to cop it.
This is set regen. So should it not deliver the same amps at low RPM but at a much lower voltage?
Im trying to understand the picture I drew below, can anyone fill in the blanks.
(http://goldenmotor.com/SMF/index.php?action=dlattach;topic=2119.0;attach=2043;image)
-
I had been thinking of doing that but I wouldn't have a clue what size resistor to use. also I would be a bit worried about to many volts on the controller in regen.
Yes.
The volts can get up over pack voltage in the controller if the resistor isnt allowing the pack to beat down regen voltage.
The issue at hand must be understood.
Is it voltage spikes, require voltage regulation or too much current requires current limit?
I understand that you should not disconnect the battery when using regen as this can damage the controller
this is why I am working on this.
(http://i29.photobucket.com/albums/c251/soulelectronique/Regenmodule2.jpg)
Its only in flow plan ATM but I am starting to see the idea of having regen braking as a different part of the whole thing.
IMO regen belongs to the battery not the controller, it just so happens you don't need the controller to make power from the motor.
-
Thanks for the reply. Your knowledge on this surpasses mine, I think I will just letting my pie cool down for a while when it over heats.
If you do make a working regen control device I would be Interested.
-
Thanks for the reply. Your knowledge on this surpasses mine, I think I will just letting my pie cool down for a while when it over heats.
If you do make a working regen control device I would be Interested.
Thanks for the thanks and your compliment. But this is all still new to me really. Often my long posts contents are too much to respond to and often filled with flaws. I like people to tell me I am wrong as I learn nothing if I think I am always right.
Take a talented musician that has been into audio tech for 25 years and give him an E-bike and I am the product.
I want to able to build a controller and I'm not even scratching the surface of a PCB yet.
BLDC was invented by the mid 70's, I don't think anyone has a total grip on every aspect of this tech, I think Phillip Yao has some to do with this development too. 40 years has seen much change and really its not even one mans life time. ICE tech is way more developed in comparison and it is a century more mature.
I still ask a lot of question that often I have to find the answer by myself, but I am often unsure that I am right.
So many GM owners have good experiences. We don't always see this here because they are too busy riding their bikes and what-not . We do see the few people turn up here on the worst day of their lives.
I do tend to get caught up trying to help those and find myself working on project for no reason..
However the regen module is a great idea and worth the effort.
How can switching a controller off in regen effect the controller if its part of the battery and always connected? Separating the two functions, regen and controller, we can allow a delay of 30 seconds or more to a relay for regen to switch off.
You can use a 12v supply through the brake to switch the controller off and power the regen fets so you cant operate regen without the brake switch wire switching the controller off first .
The only thing regen needs to relate to the controller with is the brake switch.
With a depletion mode fet you don't need PWM for regen and there is not much to be gained with PWM with regen anyway. PWM is a good way to control the amount of regen, so too can a depletion mode fet with variable 12v rail voltages at the gate can control the amount of regen.
I do believe that the phase angle detection isnt needed to get regen.
The old method of variable regen by backing off throttle is fine to assist brakes but hall sensors for electronic braking in the throttle can not be legally trusted exclusively for braking. Audio engineer sliders used on pro mixing desks last for decades of heavy use, beer, idiots sliding them hard and fast. I even cleaned out a slider pot that had sand jam it up and it still works, so maybe my idea is a step closer to reliable electronic brakes.
Communities like this, V is for voltage, Endless Sphere and a few other places are divine, They aid accelerating the development of EV tech. Bravo to all the ev peeps. We have our own non profit organisation here in Australia that is dedicated to the EV cause of making this great method of transport commercial viable.
GM may find an Aussie dealer here. Please find an Aussie dealer GM. ;)
http://www.aeva.asn.au/ (http://www.aeva.asn.au/)
http://www.aeva.asn.au/forums/ (http://www.aeva.asn.au/forums/)
-
yep, I agree, exactly the same thing on my pie. I would like to have more control on the amount of assistance. I don't like the current implementation of kicking in to full speed and after pressing the brake it kicks in right away again. And after stopping to pedal it does not stopp for another 6 seconds.
btw: I have also the 6 beeps for no reason just after a 2 minutes ride and I would like to know what it means.
Its not much point trying to convince me on the facility of PAS or the desirability of turning it off. Its a given for me that it work as advertised.
I've ridden other bikes where the PAS works very well. It was part of the MP feature set that helped me make my purchasing decision.
Currently it doesn't work as advertised. I'm wondering if anyone knows if there is a reason why this is the case and if there is a way of making it work properly.
Switching it on and off is one solution, but a kludge.
Making it work as advertised, (ie turn on by pressing Cruise twice and off by cycling power) would be better.
Best of all would be finding a way of changing the amount of assistance it provides.
Has anyone else noticed this behaviour and interested and wanted it to work properly?