How To, 50A BBSHD Controller Mod

[Ivanovitch_k]

Sorry guys, don't want to go too much off-topic here.

But I just received another Bafang Ultra wich I bought of Aliexpress from a "reliable" company.
Price was quite good.
It arrived today and I opened it up immediately to compare it with my Luna Ludicrous version.
The Ludicrous is a 1.3 version and the new one I got from China has "version 1.5" stamped on it.
Controller looks a bit different but I noticed the Hall sensors were not connected.
The connector is missing? WTF I don't think this will work.

the V1.5 still have a hall sensor but it's located on the mosfet board. The magnet is on the rotor tip. Bafang must have had some leftover parts from v1.3 as I also have the hall board in mine left unconnected.

See my picture in this other thread

 

[E-mil]

Hi all,
It's been a long time since I or anyone for that matter contributed to this topic, and I've got some updates and good ideas.

My stable BBSHD 18S 45A with NF75 FETs is still going perfect, and took the summer heat alright with up to 30 C ambient temperature.
In the meantime I have modded two brand new BBSHD controllers for 18S 50+ A. The first one got the caps replaced in standard upright position, but on the second controller I removed the lighting circuit board. This made it possible for the caps to lay sideways, and then it was no hassle to refit the controller onto the BBSHD.
View attachment 4

With or without the lighting circuit in place, I repurposed this two lead wire for voltage drop mod simply by cutting the brown display wire and solder the lighting wires in series with it. I can then choose whatever Vdrop i like as a plugin just by inserting it into the plug.


I don't recall ANYONE who dared to replace the FETs actually posting back about their maximum voltage; 19S or 20S for example. What are your stats?

Since I now will have a spare controller with the old NF75 FETs I might consider replacing them for 100 V versions. The internal BBSHD PSU chip crap out at 80 V or so according to posted specs, but it occurred to me you just might simply do the very same LED v-drop mod this chip, as done with the display/controller battery voltage sensor. It could open up for, say 22S in theory!

 

[Couchy]

Is anyone selling these upgraded controllers to order or is everyone just hacking them themselves? Im not good with electronics

 

[ichiban]

I also have a BBSHD with 14s7p currently.

Anyone achieve this 50A-mod retrofit module that can easily and directly plug into the original BBSHD motor, I'm interested. Count me in. +1.

Thanks in advance.

 

[fechter]

At one time, Luna Cycle had them with nice CNC machined heat sink housings. I don't know if they are still available.

 

[ichiban]

Thanks fechter.

This is what I found on Luna site now, they should be stock controller from Bafang. Not those 50-A heavily mod like we discussed here in this thread :

BAFANG BBSHD 1000W REPLACEMENT CONTROLLER $99.95 + shipping
https://lunacycle.com/bafang-bbshd-1000w-replacement-controller/

But Luna's price is pretty good comparing to those from AliExpress starting @ $150.- and up (free shipping)

 

[Couchy]

Looks like Luna are not selling there high performance ludicrous controllers separately unless you own one of there bikes and need a replacement.

Might have to buy a normal one from them and take it to an electrical repair shop or something similar.

https://m.youtube.com/watch?v=Gl9SALEkoGA

This looks a bit difficult for me.

 

[fechter]

It's really not that hard to do the mod even with the potting. The whole board can be pulled out with the potting and the shunt resistors are on the bottom where there should be little or no potting. The trick is to separate the potting from the housing around the edges and prying up on the screws that hold the FETs on. The potting is pretty easy to dig out with a small screwdriver where needed.

Once the board is out, changing the shunt resistance requires soldering but it is pretty straightforward.

 

[Couchy]

Ok fechter I have ordered some shunts and some fets.
WSLF25123L000FEA
IRFB3077PBF
Not sure if they are any good or not.

Will let you know how I go.

Thanks for the push

On a side note will also be ordering another controller to play with doing this.

A little while back (Post Apr 01, 2017 10:41 pm) you said something about a switch to lower the current.

"If you wanted a remote switch to lower the current, you could tap into the little resistors going to it."

How would this be done?

 

[fechter]

A little while back (Post Apr 01, 2017 10:41 pm) you said something about a switch to lower the current.

"If you wanted a remote switch to lower the current, you could tap into the little resistors going to it."

How would this be done?

It's possible but not so easy given the size of the little resistors.

The voltage across the shunt is amplified by a LM358 then sent to the microprocessor A/D input. The gain of the amplifier appears to be 10x. The gain is controlled by two resistors. To have a lower current limit, you would want to increase the gain. Lowering the 1k resistor or increasing the 10k resistor would increase the gain. Lowering a fixed resistor is easier since you don't need to remove it, you can just place another resistance in parallel.

So if you placed a 1k across the existing 1k, you could cut the current limit in half. For other ratios, it is easy enough to do the math. You could also use a trimmer pot instead of a fixed resistor so you could dial it in.

The alternate approach would be to leave alone the stock shunt resistors and lower the gain of the amp by lowering the 10k resistor. If you placed another 10k across the existing one, it would double the current limit. This would be too much and likely result in smoke or other failure. If you place a 15k across the existing 10k, it would bump the current from 30A to 50A.

 

[Couchy]

Thanks fechter.

I might make a new thread about remotely limiting current if there is not already one so I don't derail this one.

Regards

 

[E-mil]

@ Fechter:
Thanks again Fechter for the details about the analog shunt hack. I didn't pay it enough attention when it was first brought up earlier.

There's possibly a small problem if you only increase the gain, rather than solder extra shunt resistors to halve the value from 0.0025 to 0.00125 Ohm. The power dissipated in the shunt is noticeable at higher amps, as 50 A and 0.0025 Ohm gives 6.25 W heat dissipation. If you double the resistor count to four, they will share the load while only dissipate a 1/4 of 6.25 W each.
I hope I used Ohm's law correctly.
[EDIT: The board has two shunt resistors, not only one as I originally remembered. Hence I edited the calculations above, and the problem might not be a problem after all, thereby no hot spots. I have let it be included anyway].

Also, thanks for the "DIY Fix for BBS02/BBSHD Pedal/Throttle Issue" https://endless-sphere.com/forums/viewtopic.php?t=93977
I'm sure my new extra controller suffer from it as well, but I have not tested it yet. It also has the newly added inline fuse on the battery+ wire.

@ Couchy:
Cool! I hope you manage to change the FETs. I would personally not change the NF75 (inferior) to 3077 (superior) when it only gives me 5 or 10 A more. For me, the NF75 has proven sturdy enough for 45 A for a whole year, even in 30 degrees C ambient temp.
Instead, I would definitely change the NF75 to 100 V FETs because overvolting gives so much more power feel. As proposed, doing the display voltage hack just might help the PSU chip as well to accept higher voltage than 80 V anyway.


---
Generally I made a plan to sell modded controllers, with all mods done except changing the FETs.
It takes me 4-5 hours to do everything resulting in 18S 50+ A capable controller, a lot like the one posted in August. It would be more expensive than Luna's 50 A / 14S though.

I broke down the time it took me to do it. Your milage may vary:
Controller extraction 15 m
De-potting 1 h
Light board removal 15 m
Caps replacement 45 m
Shunt mod 15 m
Traces beef-up 1 h
LED Vdrop mod 30 m
reassembly 20 m
IN ALL: 4h 20m

 

[fechter]

There's possibly a small problem if you only increase the gain, rather than solder extra shunt resistors to halve the value from 0.0025 to 0.00125 Ohm. The power dissipated in the shunt is noticeable at higher amps, as 50 A and 0.0025 Ohm gives 6.25 W heat dissipation. If you double the resistor count to four, they will share the load while only dissipate a 1/4 of 6.25 W each.
I hope I used Ohm's law correctly.

Yes, you are correct, increasing the current limit with the stock shunt resistors will make them heat more. I think they will be OK as you won't spend long periods of time at the limit (or the plastic gear will melt first). The main B- wire attaches right next to the resistors so will act as an additional heat sink. You could further improve this by soldering a chunk of copper wire between the B- connection and the shunt resistors.

Replacing the FETs is a real pain on these boards as the holes are tight and hard to clear without damaging the traces. I have done a couple, but the first one wasn't pretty.

 

[Andrewol]

I would definitely change the NF75 to 100 V FETs because overvolting gives so much more power feel. As proposed, doing the display voltage hack just might help the PSU chip as well to accept higher voltage than 80 V anyway.

Have you had a chance to try this out? I'm interested in 100V mod too. What 100V FETs would be the best here?
Thanks and Regards.

 

[fechter]

Have you had a chance to try this out? I'm interested in 100V mod too. What 100V FETs would be the best here?
Thanks and Regards.

IRFB4110 has a good track record and is fairly inexpensive. TK72E12N1 is another good option. Just be sure to get "genuine" ones from a reputable supplier.

You would still need to figure out what to do about the main caps. The stock 63v rated ones are barely enough capacitance. Higher voltage ones will have even less capacitance for the same size package. Adding more externally on the main battery wires is about the only way. I've seen this done on other controllers and it's very common on RC stuff.

 

[E-mil]

Fechter, I actually managed to fit 1000 and 470 uF (both 100 V) inside the controller, as shown in my earlier post. They are both laid sideways, and it got MORE free space now than before the mod, and even the 470 is a step up from 220.
Key is to remove the lighting circuit board (and reuse the wires for V drop mod). Even then I'd put an extra external cap, as you recommend.

If I ever get my hands on a controller with busted FETs, I'll sure replace them and see if it can be pushed north of 20S. 98 V BBSHD, anyone?

 

[Andrewol]

That's great. I think the best solution here can be Onsemi FDP2D3N10C mosfet. It's 222A Id, 2.1mOhm Rds, and only 214W Pd.
Could anybody compare this moded controller with phase runner? What are pros and cons?
Also the capacitors should be 105-120 C rated, it will be hot inside the controller case.

 

[Cobain Windsurfer]

Nice to see this thread is still alive! :thumb: I want to discuss what soldering equipment to use for the controller mod. This is very important because the components are tiny and without the right tools, this mod can(will) be a pain in the ass. I used a butane soldering iron, and a electric one and managed to swap out all the mosfets, but it was not easy to get the solder to stick 100% to the pins. Where the cables enters the controller, there are a lot of tiny cables (white, red, black ...) attached to the board, and two of them got loose. I have not been able to solder them back on, they are just to small. I thought of getting this project completed, but I want some input about what equipment to buy to get it done. Any suggestions, what equipment did you use?

 

[Enric]

Thanks for all the contributors of this great post.

I just make de resistor mod and wooow..... my ebike was transformed completely.

In fact after the mod I just forget my plans to find a most powerful ebike, and I spare a lot off money.

I will use a small part of the spare money to improve the suspension and store some bbshd spare parts like plastic gear and clutch.

At the moment I will stay with the weak 7505 fets, but I have the 3307 ready for change.
Thanks,

Enric

 

[fechter]

Where the cables enters the controller, there are a lot of tiny cables (white, red, black ...) attached to the board, and two of them got loose. I have not been able to solder them back on, they are just to small. I thought of getting this project completed, but I want some input about what equipment to buy to get it done. Any suggestions, what equipment did you use?

I have an assortment of electric irons. Big one for large pieces and several smaller ones. Where the wire broke off, you want to clear the hole first so you can place the wire before soldering it back. I use a hand held solder vacuum for that but there are other ways to do it. The crudest is to heat the spot until it's melted, then quickly bang the board against the work bench and knock the solder out of the hole. Solder wick is another way. Once you get the hole clear, you can strip the end of the wire and place it through the hole, then solder the back side. Avoid bending the wires near the board.

They are tiny and it helps if you can see what you are doing well. Avoid bridging adjacent traces. Inspect well before testing.

 

[ions82]

Where the cables enters the controller, there are a lot of tiny cables (white, red, black ...) attached to the board, and two of them got loose. I have not been able to solder them back on, they are just to small. I thought of getting this project completed, but I want some input about what equipment to buy to get it done. Any suggestions, what equipment did you use?

I have an assortment of electric irons. Big one for large pieces and several smaller ones. Where the wire broke off, you want to clear the hole first so you can place the wire before soldering it back. I use a hand held solder vacuum for that but there are other ways to do it. The crudest is to heat the spot until it's melted, then quickly bang the board against the work bench and knock the solder out of the hole. Solder wick is another way. Once you get the hole clear, you can strip the end of the wire and place it through the hole, then solder the back side. Avoid bending the wires near the board.

They are tiny and it helps if you can see what you are doing well. Avoid bridging adjacent traces. Inspect well before testing.

The trick I use for clearing solder from holes in a PCB is pretty simple. I just push the tip of a wood toothpick on the hole I want cleared (maintaining light pressure). Then, I'll apply the iron tip to the same hole but on the opposite side of the board. As soon as the solder melts, the toothpick pops through and clears the hole. Just wait a second or two for the solder to re-solidify before giving the toothpick a little twist. It's a very quick/cheap/easy way to get the hole opened back up. Far better than wicking braid or the little tool I have that is supposed to suck up molten solder.

 

[fechter]

The trick I use for clearing solder from holes in a PCB is pretty simple. I just push the tip of a wood toothpick on the hole I want cleared (maintaining light pressure). Then, I'll apply the iron tip to the same hole but on the opposite side of the board. As soon as the solder melts, the toothpick pops through and clears the hole. Just wait a second or two for the solder to re-solidify before giving the toothpick a little twist. It's a very quick/cheap/easy way to get the hole opened back up. Far better than wicking braid or the little tool I have that is supposed to suck up molten solder.

That's a good one. Seems like it should work and toothpicks are easy to find.

 

[Cobain Windsurfer]

Where the cables enters the controller, there are a lot of tiny cables (white, red, black ...) attached to the board, and two of them got loose. I have not been able to solder them back on, they are just to small. I thought of getting this project completed, but I want some input about what equipment to buy to get it done. Any suggestions, what equipment did you use?

I have an assortment of electric irons. Big one for large pieces and several smaller ones. Where the wire broke off, you want to clear the hole first so you can place the wire before soldering it back. I use a hand held solder vacuum for that but there are other ways to do it. The crudest is to heat the spot until it's melted, then quickly bang the board against the work bench and knock the solder out of the hole. Solder wick is another way. Once you get the hole clear, you can strip the end of the wire and place it through the hole, then solder the back side. Avoid bending the wires near the board.

They are tiny and it helps if you can see what you are doing well. Avoid bridging adjacent traces. Inspect well before testing.

Thanks for the tip, going to clear the holes and get those wires back in for soldering :thumb:

 

[ichiban]

This is not directly related to the BBSHD 50amp controller mod BUT it is close relevant and we have to open up the motor anyway. Why not do it together.

Another thing that should be improved is the grey silicone potting. Since it is "in-between" the process. So, I am thinking about getting rid of it from the equation for once and for good. Have not seen anyone mentioned this yet in ES.

As we know, BBSHD (including BBS01+02 ?) need periodical maintenance from greasing with red Mobilgrease 28, especially on that nylon gear as recommended in several webs including Luna. I have done my BBSHD greasing even before the very first connecting it to the battery. That red grease does a good job, mine run smoothly and quietly for at least 1400km - but starting to make noise now. I don't know how long will it keep running smoothly and quietly with my considerably gentle riding. Without that grease, nylon gear has a high risk of being damaged especially if the owner riding style is harsh. Bafang did a poor job by designing their product to require maintenace fairly oftenly, and also putting very little grease on the internal geartrain - as far as I saw from my own BBSHD.





View attachment 2


However, to access the needed parts to grease include removing the controller and phase wires (that held in place by grey silicone potting). So, we will need to pick the silicone clearly to get to the nylon gear. If you have done that before, you know exactly how fun picking and repotting that silicone are. Worse than laundry chores when you are exhausted from months-long, past-due big project before. Unless your hobby is to spend time picking silicone potting to discover your true inner self. :lol:








I am thinking about moding the silicone potting area, especially on the phase wires, to add another set of plugs (may be that single 6-mm XT-150 or anything that is small enough and can handle the high current suitable for BBSHD) on the armature/stator side, in-between the motor coils AND the metal wall that phase wires need to run through, and repot as usual. We will have 2 sets of plugs (3-wire each), one factory installed and a new set on motor coils side (for not to re-pick/repot silicone again). Following this process, we will have to do the silicone picking & repotting again but for the last time. The next time we open the motor, silicone works is not needed any more - save tons of work and time. No more silicone works. Since we can disconnect the new plugs installed on the phase wires side and the controller piece can be out of the way. Maintenance works can be done as usual. Putting things back together is simply in a reverse order.

I have not done mine yet but should be soon after I finished assembling the new power packs.

Hope I made my point clear enough. Try to draw some picture to illustrate that.
Anyone tried this out, please post pictures for the following others.

 

[ichiban]

I really wish that future posts will be with pictures. Since we are dealing mostly with complicated technical details here in ES.

Trying to explain technical detail by texts are sometime misleading, hardly brief, not to the point, and not fully-covered.

So if possible, put some pictures in. A picture paints a thousand words.