Bafang ultra upgrade to 2500watt

[jpc6000]

Grin says
If you just manually short the signal to Gnd you'll see it toggle down, correct?

When I shorten the signal
SETUP PAS DEVICE only D go up and down

 

[jpc6000]

I made screenshot about ca config
The screenshot was standard, I need read first
The program is not working very good, to attemp a lot of times before succeed

This is the config

View attachment 1

 

[jpc6000]

Justin from grin technologies write:
"Be sure that you're seeing a difference between pins 3 and 4 on the PAS plug when shorting.

Shorting to ground pin 3 should be the D arrow, and pin 4 should be the P arrow."

That was wrong.

I have changed the two wires.

When I short pas now the P go down.
But when I turn crank, P still not move up and down
I have tried differend resistors but I do not succeed in moving P
Without resistor 1.6V and 4.6V
With 20K resistor I get 1.2V and 2.8V
With 10K resistor 0.9V and 2V


With 5.6K resistor 0.7V and 1.58V its working

 

[jpc6000]

picture of e-bike

 

[garyal1]

From exess-bikes:

The only problem is the controller because Bafang has made several versions which are not compatible. But we are currently developing our own controller which will be much more reliable, more efficient (higher range), better software, better protection. It will be available end of 2018. It fit in the place where the Bafang controller is mounted, so it just replaces the Bafang Controller. It comes with all bikes from January 2019 and it will be available as an aftersales part as well. So the customers having already an EXESS bike can use this controller also.

Will you be able to run more than 30A through the new controller?

yes we will be able to run more than 30A. I don’t know yet how much current we will enable, this needs to be determined.

Is this controller available yet? I have an Ultra and frame on the way and I have plenty of room inside frame for external controller mount...

 

[TorEddy]

The controller from exess-bikes is meant to replace the internal one. It's not an external controller. I don't know if it is ready yet, but send them an email and check.

 

[garyal1]

I just wanted to thank everyone who contributed to this thread. The pinout information and other helpful information allowed me to configure my new CA3 and PR to work with my new Bafang ultra 1000 build.

This thread gave me all the information I needed to have the confidence to order the parts and start the build! Worked flawlessly on the bench thanks to this thread!

GREATLY APPRECIATED!
:thumb:

 

[-JFK-]

Hi all,

I'll get my first G510 Ebike soon !
Did someone can tell me the dimensions and watt value of the stock "Shunt" resistor ?
I want to replace it by 2x2.5mohm.
And the Mosfet will be replaced by IRFS7530PbF.
What do you think ?
Thanks

 

[fechter]

And the Mosfet will be replaced by IRFS7530PbF.
What do you think ?

Those are 60v rated. OK for a 36v system but I would want a higher voltage rating for a 48 or 52v system.

 

[-JFK-]

You right, so it will be the IRFS7730PbF.
Thanks.

 

[jeron]

You right, so it will be the IRFS7730PbF.
Thanks.

Keep us posted!

 

[-JFK-]

Hi,

To reach 2500W, do you have replace the mosfet ?
Or they are good enough and it's useless ?
Thanks.

 

[fechter]

Running the stock controller at a higher current has had reliability issues, but I don't think the FETs are the weak point. It seems the voltage regulators and some 5v zener diodes across the 5v supply like to short out. Something like a big 5v TVS diode on the 5v supply might solve that.

 

[-JFK-]

Running the stock controller at a higher current has had reliability issues, but I don't think the FETs are the weak point. It seems the voltage regulators and some 5v zener diodes across the 5v supply like to short out. Something like a big 5v TVS diode on the 5v supply might solve that.

Hi,
Why increasing the output current have an impact on the 5v supply?
I can't see the link.
Do you have a photo to locate the "5v TVS diode" you're talking about?

 

[fechter]

Hi,
Why increasing the output current have an impact on the 5v supply?
I can't see the link.
Do you have a photo to locate the "5v TVS diode" you're talking about?

I don’t have any pictures on my phone but I can try to find one later.
My theory is that at higher current there are bigger voltage spikes on the battery input and these make their way through the 5v regulator and blow up the diodes. There are several ways to prevent this but adding the right TVS diode across the 5v line is probably the easiest.

 

[-JFK-]

Hi,
Why increasing the output current have an impact on the 5v supply?
I can't see the link.
Do you have a photo to locate the "5v TVS diode" you're talking about?

I don’t have any pictures on my phone but I can try to find one later.
My theory is that at higher current there are bigger voltage spikes on the battery input and these make their way through the 5v regulator and blow up the diodes. There are several ways to prevent this but adding the right TVS diode across the 5v line is probably the easiest.

I have still not receive the motor, so can't search by myself, thanks to try to find the right diodes.
I'd like to improve the TVS diode when I'll replace the Mosfets and caps.

Do you know the reference of the green arrow capacitor ?

The red one is Rubycon YXF, 63v, 470uf, so 12,5x25mm.
I guess a 1000uf, 16x25mm can fit.

But about the green one, by measuring on the photo, I find 18mm diameter.
Based on Rubycon catalogue, there is no 18mm in YXJ range, it's seem's to be 1000uf, 16x25mm max.
I'd like to know the dimensions of this cap.
I guess I can put two side by side (green circle).
Thanks.

 

[-JFK-]

Is it the faulty TVS Diode ?

 

[fechter]

The ones that like to blow up are the little orange ones across the 5v supply. They are zener diodes. They should be TVS diodes instead. Still looking for pics of the exact ones.
Larger/better main caps will likely help this problem, but it seems like at some point there is a strong spike of regen that puts the input voltage over some limit where stuff blows up.

 

[-JFK-]

The ones that like to blow up are the little orange ones across the 5v supply. They are zener diodes. They should be TVS diodes instead. Still looking for pics of the exact ones.
Larger/better main caps will likely help this problem, but it seems like at some point there is a strong spike of regen that puts the input voltage over some limit where stuff blows up.

Thanks Fechter,
Once you have located the faulty one, I'll changes it for prevention.
Thanks a lot for your help.
I have found Panasonic caps with very low ESR, I'll put 2000uF at "left/green" and 1000uF at "right/red".

 

[-JFK-]

Did yo find the correct diodes ?

 

[fechter]

These like to blow. I don't remember which one of the two is the zener, but on mine it's shorted so hard to tell.



When it shorts, it often takes out the 5v regulator. Replacing the zener with a TVS diode might solve the problem. A big TVS diode across the battery input (with the proper voltage rating) might also do it. You might get away with leaving the zener in and put a TVS somewhere else, like the holes marked 5v and GND.

I've seen TVS diodes across the battery line on a Bosch controller.

 

[-JFK-]

These like to blow. I don't remember which one of the two is the zener, but on mine it's shorted so hard to tell.

Blown Zener.jpg

When it shorts, it often takes out the 5v regulator. Replacing the zener with a TVS diode might solve the problem. A big TVS diode across the battery input (with the proper voltage rating) might also do it. You might get away with leaving the zener in and put a TVS somewhere else, like the holes marked 5v and GND.

I've seen TVS diodes across the battery line on a Bosch controller.

Thanks a lot!

 

[Jimmydoerty]

Anyone willing to comment on reliability at this power with the ultra motor?

Possibly worth getting a brass or composite replacements for the nylon gear inside if running this for a long time?

Thanks all

 

[Deafcat]

There aren't any nylon gears, unless you have a very early version of the Ultra motor.

Mechanically the motor is completely capable of reliable operation at 2500W, with good riding habits, avoiding high instantaneous torque load on the sprag/one way bearing in the middle of the gear train (heavy throttle from standstill, or up a hill, especially when combined with too large a chainring).

Bafang has mitigated some of that throttle from standstill risk in their latest stock programming, the bike must be moving to use throttle now... But this can be reprogrammed.

 

[fechter]

Other than blown controllers, the only other reports I've seen were burnt windings from overheating. A temp sensor inside the motor that rolls back the current when it gets hot is a good idea.