How To, 50A BBSHD Controller Mod

A crude way to get around the voltage limitation would be to place a voltage regulator between the battery input and the B+ line going to the display that limits it to under 60V. The display has a voltage limitation as well as the controller logic.

This regulator would be placed in series with the BROWN wire going from the controller to the display. This line gets switched by the display unit and feeds pack voltage back to the controller. Normally this line supplies about 100mA when the motor is running.
 
fechter said:
A crude way to get around the voltage limitation would be to place a voltage regulator between the battery input and the B+ line going to the display that limits it to under 60V. The display has a voltage limitation as well as the controller logic.

This regulator would be placed in series with the BROWN wire going from the controller to the display. This line gets switched by the display unit and feeds pack voltage back to the controller. Normally this line supplies about 100mA when the motor is running.

Oh, so the logic receives all of its power through the brown wire? Sweet! I already removed the brown and orange wires and soldered a jumper on the controller. Hopefully I can just solder a to-220 regulator in it's place. The SMD components above the jump have gnd.

Gonna go look for a regulator now.................



**EDIT**

Searched Mouser, cant seem to find a suitable regulator.

How about just a voltage divider? Figure a 10ohm and 100ohm resistor would give me 57v from 63v input.
I've got tons of SMD resistors. Something like so..........
 

Attachments

  • Untitled.jpg
    Untitled.jpg
    146.9 KB · Views: 6,221
Too much load for a voltage divider and the load isn't constant. You want a voltage regulator with about 52v output. You can make one from a transistor, a zener diode and a couple of resistors. It will render the voltage display useless. There are fancier approaches you could do which might preserve the voltage display usefulness.
 
I just ended up running two 47ohm 1/2 watt resistors in series between where the brown and orange wires were connected. With battery at 63v, the controller sees 61.4v. I think I have fully pimped this controller now: Shunts replaced, caps replaced, hvc ghetto-bypassed, wiring reduced and rerouted, Arctic MX-2 thermal compound added. Also, I just got my Eggrider bluetooth programmer setup. Now, if only I had a working BMS - Speedict Neptune 15...........long story.
 

Attachments

  • IMG_20170905_193307.jpg
    IMG_20170905_193307.jpg
    203 KB · Views: 6,168
Hi guys! Thanks for sharing the knowledge.

I'm preparing myself for the controller mod to make it run on 40-50A (<60V) and I have a few questions:

1. Will it be enough if I replace the SMD resistors only (pair of 3 mOhm) or should I replace the fets as well to keep it cooler?
2. How are your controllers? Did you have any issues so far? Is it getting much hotter than on 30A?
3. What did you use to refill the sealant? Thermal silicon paste?

Cheers!
 
Most people have just modded the shunt resistors and leave the stock FETs. It is very difficult to replace the FETs on this board without damaging anything.

Check to see which FETs your controller has. The newer ones seem to be coming with IRFB3077s, which are pretty good. The older ones with 75NF75s will generate significant heat at high current but the motor will be contributing more to the heat than the FETs. A temperature monitor would be a good idea.

Even running 30A for any length of time is going to get the motor hot. It will get hot faster at higher current. Normally you only use the max current when starting out or going up a hill. For long hills, you will still need to gear down to keep the current reasonable.

I just left my board unpotted but it could be an issue if I rode in rain. Some kind of conformal coating works well and is a lot easier to get off if you need to rework something.
 
Thanks fechter! Now it's all clear :)

I'll take out my controller in the next few days so I'll check what FETs I have.

Which components are generating the most heat? Shunt resistors or fets?
What do you think about adding small radiators on those parts?
 
You can do the math and see the shunt resistors throw off a few watts of heat, but guys have just stacked additional resistors on top of the stock ones and it seems to work. I guess they dissipate to the PC board. I wouldn't bother with trying to add heat sinks to the resistors. They are also easy to work on if one does fail.

The FETs are already bolted to the case, so nothing more to do there. I have seen a custom CNC'd outer case with extended cooling fins made by Luna. That might be worth while or figure out a way to add some external fins.
 
I've been hammering mine for about 6 weeks now @ 63v/50A. I weigh around 265lbs. in gear, and my bike is a single speed 30t/16t. The controller/motor definitely gets warm, but never too hot to touch. I added Arctic MX-2 thermal paste in between the fet's heat spreader and the case. It's a non-electrically conductive so no worries about using "too much".
 
Epyon said:
I've been hammering mine for about 6 weeks now @ 63v/50A. I weigh around 265lbs. in gear, and my bike is a single speed 30t/16t. The controller/motor definitely gets warm, but never too hot to touch. I added Arctic MX-2 thermal paste in between the fet's heat spreader and the case. It's a non-electrically conductive so no worries about using "too much".

The thermal paste is a good idea. The fit between the heat spreader and the case isn't very tight.

Do you know which FETs yours has?
 
Very interesting thread!

At the moment I do not plan to modify it, but this thread gives valuable info about the internals of this controller and its limitations, so very much appreciated the efforts and risks you guys take for maxing the performance of BBSHD with the stock controller.

Regarding the FETs: I bought my BBSHD in July 2017 from UK. Is there a way of telling what type of FETs in the controller, without pulling it apart? Like by serial number, or if not, by production date?

HVC: At the moment I run it stock with 13S battery, but if possible would like to run it with 15S@30A, but if the controller/LCD cuts at 60V the charge would be a bit low for 15S, so 14S would be better. It's mentioned in this thread a HVC of 61.5V, but others have said that it will not run above 60V, so if anyone have tested this limit in real life, that would be very useful info for me and others, that wants to push stock controller to max without modifying it. I wouldn't mind soldering a SMD resistor or two if it was easy accessible, but removing potting etc runs too much risk of damaging controller.
Bard
 
bardbe said:
HVC: At the moment I run it stock with 13S battery, but if possible would like to run it with 15S@30A, but if the controller/LCD cuts at 60V the charge would be a bit low for 15S, so 14S would be better. It's mentioned in this thread a HVC of 61.5V, but others have said that it will not run above 60V, so if anyone have tested this limit in real life, that would be very useful info for me and others, that wants to push stock controller to max without modifying it. I wouldn't mind soldering a SMD resistor or two if it was easy accessible, but removing potting etc runs too much risk of damaging controller.
Bard
Mine works at 61.4v. You can just add a resistor into your wiring harness on either the brown or orange wire. The voltage passes from the controller through the display and back to the controller's logic. I used 2 47ohm resistors in series. I am not using a display though, so you may have to use a slightly lower resistance due to the increased load (resistance increases with higher current). Like fechter mentioned several posts back, a voltage regulator would be better. All this will, of course, throw off your display's voltage display, and you'll need to drop your LVC in the controller configuration.
 
Epyon said:
bardbe said:
HVC: At the moment I run it stock with 13S battery, but if possible would like to run it with 15S@30A, but if the controller/LCD cuts at 60V the charge would be a bit low for 15S, so 14S would be better. It's mentioned in this thread a HVC of 61.5V, but others have said that it will not run above 60V, so if anyone have tested this limit in real life, that would be very useful info for me and others, that wants to push stock controller to max without modifying it. I wouldn't mind soldering a SMD resistor or two if it was easy accessible, but removing potting etc runs too much risk of damaging controller.
Bard
Mine works at 61.4v. You can just add a resistor into your wiring harness on either the brown or orange wire. The voltage passes from the controller through the display and back to the controller's logic. I used 2 47ohm resistors in series. I am not using a display though, so you may have to use a slightly lower resistance due to the increased load (resistance increases with higher current). Like fechter mentioned several posts back, a voltage regulator would be better. All this will, of course, throw off your display's voltage display, and you'll need to drop your LVC in the controller configuration.

- Excellent info! (Though I'm about to get the ERT Sinewave kit for 20-24S setup). 61.4 V / 15S = 4.09 V/cell , which is a sweet spot.
Wouldn't a diode be a much better solution to make a stable voltage drop instead of using resistors?

Another thing, the 43 V max LVC is already too low for 14 S. I wouldn't mind mine to be around 3,3 V/cell.
 
E-mil said:
Wouldn't a diode be a much better solution to make a stable voltage drop instead of using resistors?

Another thing, the 43 V max LVC is already too low for 14 S. I wouldn't mind mine to be around 3,3 V/cell.

A zener diode would work in theory. There is a big current spike when it turns on, which would blow a regular zener. A TVS diode might work.
 
Today I have done the series resistor mod on the BROWN wire in the display cable, and added a switch to be able to change between mod or standard.

I chose a resistor value of 68 Ohm for lowest possible watt-loss while still allowing the system to operate.
The display draws around 50 mA at illumination level 4/5, but it varies relative to the level in settings so a voltage drop between 2.5 V to 3.8 V over the resistor will occur.

Some quick calculations for that:
62.5 V - 59 V = 3.5 V drop. 3.5 V / 0.050 A = 70 Ohm

Watts taken by resistor: 3.5 V x 0.05 A = 0.175 W . This can be handled by any resistor.


Before I did the mod I made some tests:

No mods --
58.8 V : All OK. (14S max voltage of 4.2 V/cell).
61.4 V : All OK. I did not expect that to work! (15S, 4.093 V/cell theoretical).
61.5 V : "Error 07 Controller Report". Motor refuses to drive. See picture.

Resistor mod (68 ohm) --
62.5 V: All OK! Motor turns. (15S, 4.167 V/cell theoretical).

This results in a max wattage of around: 30 A x 62 V = 1860 W. Using a 20Ah 5C/10C battery with low voltage drop, this will certainly be felt.

Next step is modding the shunt and traces for higher amps, and I hope the controller contains the IRFB3077 FETs.
Also replacing the caps for higher voltage rated ones will be on the list. Is it possible to use caps of 35 V or 50 V in series, or will the Ri be too high?
I just might postpone my ERT shopping! ;)

(My BBSHD was produced 11. Nov 2016 according to label on the bottom. I guess it uses the quick-release plugs instead of Andersons).

Display cable opened.jpg
Overvoltage alert.jpg
Voltage mod switch.jpg
 
That is awesome. Great work.
For caps, look for 80v 1000yf. They were the highest I could find in the same physical size (16x27mm). The smaller is 220yf. The smallest cap should already be a 100v 10yf.
 
Following
 
Awesome information here.
I'm purchasing a bbsHD this week and will not be able to resist giving it some tweaks.
My initial plan is to buy the basic set-up and get it running but looking at the "next step" it is clear that the battery that I was planning on is not going to cut the mustard re current.
I'd like to run 40A and potentially 52V (or more). I'm not too concerned about the voltage not reading on the display initially.
Can people recommend a battery that will be better than me buying a cheap one that won't cope with the mods later, please?
I've set aside 2 bikes to look at. A specialized SX trail (old) and a Giant AX (even older). They are set up as "enduro/trail" style but both have battery fitment issues. Now...given that my electronic knowledge is at least good enough to follow series and parallel, I may have to split the battery to 2 locations and re-house them.
Trolling aliexpress etc (I live in NZ so they ship here ok but many other locations won't)....what should I look for that would go with these mods?
https://www.aliexpress.com/item/fat-bik ... 269.twZrhF
This is the one I was thinking of using.
(I did read the posting rules but have the memory retention of a brain-dead squirrel, so if the link is not legal, my apologies...I think it's legit and in no way have interest in the company etc).
(I'm assuming that a purchase this week from a high turn over seller will give a unit that has the 3077 Fet array and getting hydraulic sensors and gear sensor too and a chain ring set up that I may be able to use a front mech on[perhaps]).

Thanks in advance :)
 
Somebody needs to design an open source circuit...
The circuit would serve to trick the bike Bafang display into thinkink it's always below 61.5V.

The circuit would let any voltage pass between zero and 61.4V. That way, you can still have an accurate voltage reading on the Bafang display in the normal 0 to 61,4 V operating range.
But then, anything 61.5V and over, the circuit would regulate voltage out at 61.4V so that the Bafang Display would not hitn the HVC, show error code, and lock the controller to an OFF mode.

I'm curious to know...
At 30A normal.... What volts (watts) can the controller handle if the Bafang display HVC is not an issue anymore.
Provided we can find suited capacitors, can we pass 100.8 Volts through the rest of the controllers circuit boards components ??? If we can, then I dont see why we would need more amps anyways, avoiding big gauge cable is an advantage.... And I'd rather have more rpm/speed, than higher torque but low speed...

So 62.4V OK....
What about a 20S Battery, with 72.0V nominal and 84.0V fully charged. --> at 30A, that's 2160-2520W ! (For the same power, you'd need 42.9A on standard 14S battery voltage)
What about a 24S Battery, with 86.4V nominal and 100.8V fully charged. --> at 30A, that's 2592-3024W ! (For the same power, you'd need 51.4A on standard 14S battery voltage)
I'd rather have more top speed (aka more motor RPM) by upping the controller volts, than have more "not so usefull»" torque by upping the controller amps (higher amps means higher heat and need for large gauge cables).

Just at 48V, the BBSHD already provides >160N.m of torque. I can already pop wheelies... I don't really see the need for more torque (aka more amps)...
I think higher top speed mod (aka more volts) would be way more beneficial (of course a suited bicycle frame, tires, brakes, suspensions is a sine qua none for higher speeds).

I'm concerned that some of the IC chips in the Bafang BBSHD Controller would just fry if the BBSHD is overclocked to 100.8V (max, fully charged on 24S battery).
This is just my guess, but I have no knowledge/experience to be honest on that level of electronics (I don't know much on chip technology) ....
Would the chips in the BBSHD controller fry at 100.8V or not ???? What about at 84.0V ???
In other word, are the Bafang Display HVC and the capacitor rating the only limiting factors to upping the voltage on Bafang BBSHD systems ???
Have we succeeded to push the system limit higher by bypassing the HVC and swapping capacitors to higher voltages or are there other bottlenecks with the original Bafang controller ?
The whole idea here is to keep the integrated look of the Bafang controller, and still having most of the fonctionality of the Bafang display.... But having higher performance.


Matador
 
Other than the display unit and voltage regulator, you have to worry about the FETs and the main caps. If you replaced the FETs and main caps with 100v rated parts, it might work except you can't fit enough caps. One possible workaround for this would be to mount an additional cap outside the controller right where the battery wires come out of the case (like some RC controllers).

A regulator to keep the display voltage under 60v is pretty easy. I'll post a schematic when I get a chance.
 
Wow, that'd be awsome !
Could mount a big capacitors inside the seat post tube :D Swap some same-size higher voltage FETS...
But the "might work" question is still puzzeling. Will it predictably fry, or could it work relatively reliably for some time.

All these computer chips on the controller board .... will they just realease magic blue smoke right away or doubling voltage is no problem ?...
I know some people have experience with standard controllers. Are Bafang controller less tolerant to voltage overclocking than standard controllers ?
 
All the voltage sensitive microprocessor stuff is protected by the voltage regulator (in theory). Just the FETs and caps will take full pack voltage.
 
So, I made this recap of the voltage mod options. Turned out to look too good not to share here.

What are your thoughts?

BBSHD Voltage mod evaluation of options (Screenshot 2017-10-19).png
 
Back
Top