Can I get more power from my 36V 250W Bafang (10Ah battery) ?

[Orraman]

emissios-free and daniel,

Is it possible that you gentlemen have not yet experienced the Panasonic torque sensing system?

The level of response is to the point that flecc has described it as bionic, and so it is. It matches the cyclist's power in every pedal stroke at normal speeds but at low speed in hill climbing there is a measure of continuing power between strokes . This is most evident when cresting a sharp rise when the power continues for a few feet, this is quite welcome and gives a feeling of the assistance that has been given.

My Gitane has the Shimano automatic 3 speed hub and although not best matched these two systems allow a sub standard human all the joys of cycling about town without involvement in the technology.

Dave

The Cyclone powered recumbent is a joy on level open roads but level is scarce here and throttle management for derailleur gear changing to maintain a regular human input is a continuing chore.
Torque sensing equal to Panasonic would be a great benefit, hold a throttle setting, ease pedal pressure, change gear and go.
Member jcebike offered a torque sensing crank from China some time ago and I had hoped to make a simple microcontroller to interface with the Cyclone but we were unable to communicate successfully.

My satisfaction with the Panasonic system means that although it lacks the comfort and the dash of the recumbent it is by far the most used but a torque sensor could change that.

 

[daniel.weck]

I tried a crank-mounted pressure sensor but unfortunately I haven't experienced the Panasonic crank unit yet. Based on people's descriptions, I understand how "natural" it must feel. I would own one if I could, but:

(1) I need a folding bike due to the lack of storage space, which narrows it down to the Swiss Flyer, a Gitane model, and the upcoming Dahon (although it's an untested Sunstar unit with probably a different feel and power specifications than the Panasonic),

(2) even if I had enough room, I wouldn't be able to afford such a bike (I already own an expensive Brompton, which I love, so I don't intend to replace it).

That's why I decided to give my current bike the "e" treatment, and I went the DIY way to:

(1) cut the costs down,

(2) have more flexibility,

(3) because it's a lot of fun !

I'm not convinced that the crank sensor I'm getting will improve my riding experience, but there's a chance it will so I'm grabbing the opportunity. As usual, I'll post details on here when I know more.

Cheers, Dan

 

[john]

So, can I technically get more power out of the Bafang if I increase the controller protection threshold ? (e-Crazyman / Keywin Ge controller + USB cable with ParameterDesigner.exe software). More importantly, how safe do you think it would be ?

Your 10 amps is almost certainly a controller limit. The Ping battery is considerably higher than that, as you can see from the spec and the stall current of the Bafang is much higher too. The Torq 1 has a limit of 22 amps I believe and the Bafang pulls this through most of its rpm range.

 

[daniel.weck]

Your 10 amps is almost certainly a controller limit. The Ping battery is considerably higher than that, as you can see from the spec and the stall current of the Bafang is much higher too. The Torq 1 has a limit of 22 amps I believe and the Bafang pulls this through most of its rpm range.

Hi John, I think so too...the Cycle Analyst shows a 15A spike, but really the peak value that I observe in practice is 10/11 Amperes.
I'm trying to get a compatible serial/USB data cable to check the actual limits in my controller firmware.
Cheers, Dan

 

[john]

You can get a cable here:
Infineon Chip Controllers

let me know if you find an alternative source as I am thinking of getting one too.

 

[daniel.weck]

You can get a cable here:
Infineon Chip Controllers

let me know if you find an alternative source as I am thinking of getting one too.

DIY USB to Serial Cable For $3! | uC Hobby



Endless-sphere.com • View topic - Infineon has REGEN BRAKING (and more)

 

[daniel.weck]

I'm going to purchase a regular USB - TTL 0-5V serial converter.

Many of them are based on the FTDI chipset (which has free drivers to enable a virtual COM port, for most systems):

Drivers: Virtual COM Port Drivers

------ Board from China (actually tested to work by Keywin Ge / E-Crazyman):

PC USB Serial to RS232 Module TTL Converter Adapter NEW on eBay (end time 21-Dec-09 09:03:43 GMT)

------ This one is available in the UK:

µUSB-CE5 - µUSB - USB to Serial Converter Modules

Slim USB to TTL UART serial converter FTDI on eBay (end time 18-Dec-09 09:36:38 GMT)

------ This one too is available in the UK, and it comes with a 5-prong header, ready to be plugged into the controller:

USB to TTL serial Convertor on eBay (end time 09-Dec-09 12:42:18 GMT)

------ Another one from China:

PC USB to RS232 RS485 TTL Signal Converter New on eBay (end time 31-Dec-09 16:00:52 GMT)

------ And another one:

PC USB to RS232 RS485 UART TTL Signal Converter New on eBay (end time 09-Dec-09 09:50:49 GMT)

 

[daniel.weck]

There's another chip with readily available virtual COM port drivers:

Drivers:
USB to UART Bridge VCP Drivers

------ Neat (albeit fragile) version:

USB-TTL Converter on eBay (end time 25-Dec-09 02:43:38 GMT)

------ A much more complete version (my favorite, because of the screw-on wire connectors and the choice of RS232 and USB plugs):

CP2102 USB/TTL/RS232 Serial Port Converter Transceiver on eBay (end time 13-Dec-09 09:20:52 GMT)

http://www.sure-electronics.net/mcu,display/DB-UC001_Ver1.0_EN.pdf

 

[daniel.weck]

Using the above USB-TTL (0-5V) converters, we can plug into the allocated pin holes on the controller board (this photo of a 6x P60NF06 MOSFETs controller is not mine, but the type is the same: EB806-XC-C-3), and use instructions from Endless-Sphere, or from the link below:

Infineon Chip Controllers

Endless-sphere.com • View topic - Methods 100V 100A Programmable Regen Controller

 

[daniel.weck]

A 5-way Latching Pin Header can be soldered on the PCB, and a corresponding Crimp Terminal Housing can be used to ease the flashing process.

5x 5-Way Latching Pin Header+Crimp Terminal+Housing Kit on eBay (end time 16-Dec-09 15:38:13 GMT)

 

[daniel.weck]

And finally, the flashing software can be found on Endless-Sphere or here:

http://www.e-bikemike.com/dl/ParameterDesignerSetup-ALL.zip




The latest version of the programming software for the 116 chipset (hall-sensor controllers, like some Infineon ones or the Open-Source one on Endless-Sphere) can be downloaded at the link below (with an EndlessSphere account):

http://endless-sphere.com/forums/download/file.php?id=23833

(in theory it won't work with my controller board, but I put it here for reference)

 

[daniel.weck]

I should add a handy link to NRG's tips about where to purchase connectors:

http://www.pedelecs.co.uk/forum/electric-bicycles/4984-my-first-entry-into-world-e-bikes-5.html#post64793

 

[john]

I some some built into a usb cable which looked quite neat.

 

[daniel.weck]

And finally, the flashing software can be found on Endless-Sphere or here:

http://www.e-bikemike.com/dl/ParameterDesignerSetup-ALL.zip




The latest version of the programming software for the 116 chipset (hall-sensor controllers, like some Infineon ones or the Open-Source one on Endless-Sphere) can be downloaded at the link below (with an EndlessSphere account):

http://endless-sphere.com/forums/download/file.php?id=23833

(in theory it won't work with my controller board, but I put it here for reference)

Earlier this year, the USB/TTL protocol was reverse-engineered, so there is now an alternative software to "Parameter Designer.exe" (it's open-source, works on Linux and Windows):

Endless-sphere.com • View topic - XPD: open-source keywin e-bike lab replacement

Normal builds:

https://developer.berlios.de/project/showfiles.php?group_id=12389

Experimental builds (mainly for testing braking-regeneration):

Index of /zap/pub/xpd


I like the presets for various Infineon controllers. It should work for all ecrazyman (Keywin Ge) / Ed Lyen variants (Shenzhen Sucteam hardware, or EB8XX / EB2XX series), as it is functionally very similar to Keywin's eBike Lab software.

 

[daniel.weck]

And finally, the flashing software can be found on Endless-Sphere or here:

http://www.e-bikemike.com/dl/ParameterDesignerSetup-ALL.zip




The latest version of the programming software for the 116 chipset (hall-sensor controllers, like some Infineon ones or the Open-Source one on Endless-Sphere) can be downloaded at the link below (with an EndlessSphere account):

http://endless-sphere.com/forums/download/file.php?id=23833

(in theory it won't work with my controller board, but I put it here for reference)

User manual (can be used to understand the values configured via the open-source software mentioned in the previous post):

Endless-sphere.com • View topic - Keywin E-bike Lab Parameter Designer Software Manual-ver 1.2

 

[Kudoscycles]

We tried increasing the amps(12>16) on the controller on our latest Eco bike-!0 Ah x36v LifePo4 battery and 250 Watt SB motor. It made the motor noisier,the range reduced a lot but in all honesty made little difference to the hill climbing ability. So have reverted to the standard 12 amps for production.
Dave
KudosCycles

 

[Scottyf]

It should realistically have increased torque by 25% as electric motor's have a linear torque line with increased amp's.
It should be quicker. I can understand not wanting to push the smaller motor from say a peak of 504 watt peak (42v x 12amp) appx. To 672 watt peak (42v x 16amp). Not including voltage sag.

Depends though how much sag the battery would have though srawing out those extra amps and what C rating they can provide.

Maybe you just put a little less effort in without realising...
Maybe most of the extra draw was turned into heat...

 

[NRG]

I'm surprised you did not notice any climbing difference, power at the motor with 16amps on a fully charged battery would have been around 420w after losses vs approx 310 at 12amp. Certainly when I set my alien up at 12amp it could still climb hills but there was a clear difference between v 15amp.

Also 12amp at 36v nominal battery voltage would place the motor power output in the legal range of 250w+10%, maybe that was the real intention...