How To, 50A BBSHD Controller Mod

[fechter]

This old BBSHD controller is fulfilled. It is my fourth tuned controller and it got the full overhaul. Tuned to the max and ready for 22S 92V 55A.

It will be an absolute beast and probably the most insane BBSHD with an original controller .

What LEDs are you using for the voltage dropper? Do they light up?

If you have the plastic gear, that will be the weak link if you run it hard.

 

[E-mil]

Hi Fechter, the LEDs are just standard 10W and 1W white or red, as in my example here.
In an old post I measured the Vf of different LEDs at different current draw. Seems like both the display and the controller logic are fed thru them, so display illumination level affects the draw.
Consensus was that current draw is around 75 mA, and a typical white 10W LED drops 8.9V while a 1W drops 2.9-3.3V (all depends on type/quality/bin).

Pictured is the sandwiched LEDs I used for the test, here powered at about 29V 80mA. It's enough for a backlight and the LEDs are running way under spec, albeit without a heatsink.


I have a spare nylon gear. Let's see what goes first.

 

[ChristianKi]

Hi Fechter, the LEDs are just standard 10W and 1W white or red, as in my example here.
In an old post I measured the Vf of different LEDs at different current draw. Seems like both the display and the controller logic are fed thru them, so display illumination level affects the draw.
Consensus was that current draw is around 75 mA, and a typical white 10W LED drops 8.9V while a 1W drops 2.9-3.3V (all depends on type/quality/bin).

Pictured is the sandwiched LEDs I used for the test, here powered at about 29V 80mA. It's enough for a backlight and the LEDs are running way under spec, albeit without a heatsink.
IMG_20201208_230729.jpg

I have a spare nylon gear. Let's see what goes first.

What about your bike? Does it run as you thought? Or are there any issues?
How much km/miles did you drive after this tuning?

I read nearly the complete thread. But what I don't understand is the voltage drop with the led's. Is this a reason of the controller or the display?
If it is because of the display, does it work at a 20S 72V without led's if I want to run a eggrider V2?
At the eggrider app I could select a 72V battery.

 

[fechter]

With the stock display, it will cut off at about 61.5V. So it looks like maybe the Eggrider will get around that limit. What I'm not sure about is whether the 61.5v limit is monitored in the display or on the controller board. If it's on the controller then the Eggrider won't help.

The voltage regulator in the motor controller is rated for 80v max and as I recall, the stock FETs are rated for 75v. Even with the Eggrider, one of these other parts might not be happy at the higher voltage.

 

[ChristianKi]

Sounds logical. I know about the fets. That isn't a big issue. Changing parts is the simplest part of the game. Destroying Prozessors is way more painful.

So electrically it's like this? The full battery voltage goes to the display (brown) . The voltage for the Prozessor comes from the display (Pl-orange)???
But this normally couldn't be. Because if you aplicate the controller with the bafang tool, there is only the UART adapter (max. 5V) at the processor. If the processor (does mcu mean processor?) needs only 5V from the USB UART port, it normally doesn't get more from PL/Orange.

The Vdrop with the LEDs is only at the brown cable to the display, right? I bet there is a separate voltage regulator with 5/3.3V out for the processor.
If so, what does PL/orange from the display do?

Why not using a voltage divider to drop the voltage? Then I could divide it to exactly half. The display shows 36V instead of 72. I could calculate the range like 36V battery and with a 60A shunt mod the power is shown exactly divide by 4. All pretty simple to calculate in the head.
If my 13S battery is build, I have to measure the voltage at the orange cable. I'm waiting for cells and charger for now.

 

[fechter]

The orange wire is switched pack voltage going from display to the controller board. On the controller board there is a voltage regulator that drops it to 12v and a cascaded regulator for the 5v.

The reason a voltage divider doesn't work well is the load current varies significantly so the voltage will be all over the map. If the voltage seen by the controller gets too low, it will trigger a low battery cutoff.

 

[ChristianKi]

The reason a voltage divider doesn't work well is the load current varies significantly so the voltage will be all over the map. If the voltage seen by the controller gets too low, it will trigger a low battery cutoff.

Why should this be? There isn't any heavy load. Just a display, the controller processor and the processor at the display. I don't understand the alternating current.

But now I have another idea.
To be continued.....

btw.... What about the eggrider? Das this work out of the box with 72V 20S at 84V full charged? You could select 72V in the app.

 

[fechter]

The current draw on the low voltage supply does vary with operating conditions. Most of the change is likely due to the FET gate drivers. I measured between 25mA and 40mA depending on conditions, so a pretty significant variation.

The Eggrider itself may be fine at the higher voltage but I think you will still get an error if the controller sees over 63v.

 

[ChristianKi]

In 5-6weeks I get the 750C display for the sabvoton controller. Maybe this would work.
I'm gonna test it.

 

[fechter]

With a Sabvoton, it should be fine at higher voltage. Then you'll have enough power to melt the plastic gear.

 

[ViriViri]

Hello all!
I have been lamenting over the new 28A 52V BBSHD controllers being severely nerfed, only peaking at around 27A and rolling back to 25A. They are noticeably weaker than my old (now dead) controller from 2017. Yesterday I soldered a single 5mohm resistor from that dead controller to my weakling new one and all I can say is... I should have done this sooner. It was easy for me since I've pulled apart a dead one before. No need to pull the potting other than from the backside where the shunt is and the fet heatsink. There was no thermal paste, just the rubber potting between the fet heatsink and controller cover. Seriously, Bafang don't care.

Eggrider now reports max 1509 watts so 26A@58,5V which should translate to 39A/2281W peaks in reality. I need to get an ammeter for accuracy, those numbers can't really be trusted. It makes a nice and noticeable difference, but now I'm left wondering about that second resistor and if I should have gone for it. My battery can do 44A continuous and 55A bursts, so I would need to dial it down to those numbers so nothing crazy powerwise. The motor is coming apart for maintenance next week and I'll have to pull the controller, so I might as well solder a second resistor on the one I installed yesterday. I'll probably redo the phase wire plugs and add some conformal coating to the exposed board and around the wire harness entry.

Since there are no pics of two resistors soldered on top of the old ones, Is it as straight forward as dropping another 5mohm resistor on top of the one I already added?

But yeah, If you're on the fence about performing this simple 45A mod, go for it. Here's to hoping mine won't go up in flames any time soon, I have a spare but really don't want to use it anymore. I gained noticeable torque on all gears and almost 10kmh more top speed .

 

[Baron]

Has anyone bought SMD resistors for this mod recently? Where did you buy them from?

I found this 5 mOhms from the video: Mouser #: 588-PCS2512FR0050ET
https://www.youtube.com/watch?v=mjAZ9ix7tp0&feature=youtu.be

For the 3 mOhms part is this correct? Mouser #: 71-WSLF25123L000FEA

Another question, do you need to have a programming cable to change the max amps to the motor? For example, for a BBSHD that came with max 30A set, would you need the programming cable / software to change it to 25 or can you do it on the display?

 

[Rydon]

So electrically it's like this? The full battery voltage goes to the display (brown) . The voltage for the Prozessor comes from the display (Pl-orange)???
But this normally couldn't be. Because if you aplicate the controller with the bafang tool, there is only the UART adapter (max. 5V) at the processor. If the processor (does mcu mean processor?) needs only 5V from the USB UART port, it normally doesn't get more from PL/Orange.

The Vdrop with the LEDs is only at the brown cable to the display, right? I bet there is a separate voltage regulator with 5/3.3V out for the processor.
If so, what does PL/orange from the display do?

I am using PL to power more powerful lights that can take up to 60v without using a step-downconverter. I could run them directly off the battery but I had a bunch of cables made that pick power up near the bars to make it easy to add more powerful lights on conversions. PL has the same power as P+ but is only switched on when the display is powered up. In fact, you can power up the controller and run the motor by connecting P+ and PL without a display.

 

[ChristianKi]

LM358 Layout.jpg

As far as I can tell, the offset resistor is 30k. The offset resistor is on the other side of the board. At zero current, the input to the MCU is about 1.70V.

So, for example, if you placed a 10k resistor across the existing 10K gain resistor, the amplifier gain would be 1/2, which would double your current limit. But its not quite that easy. The gain will also affect the offset. If the MCU sees a zero current signal that's out of a specified range, I'm sure it will trigger an error. If the offset is in the required range, then changing the gain resistor will change the current limit.

Well, it wouldn't be too hard to try paralleling a 10k resistor across the gain resistor and see if the MCU boots up with an error or not. If so, then the offset resistor would also need to be tweaked. With some math, we could figure out exactly how much.

Why would you go this way instead ox the shunt mod?

 

[fechter]

Why would you go this way instead ox the shunt mod?

You could potentially make it adjustable or switchable, but in the case of the BBSHD, you're right, there is no real advantage. You can program the current limit and adjust things that way.

 

[Theo128]

Hi, i´m stupid and think I`ve broken my controller by doing the shunt mod. My Bike Display is showing now error 11h. I removed the resistence but still the same error message. Do I have any chance to fix it? Kind regards, Theo

 

[fechter]

11h indicates temperature sensor failure. The temperature sensor is a little black bead on a pair of black wires that attach near the PAS sensor wires. I would pull it apart and look for that sensor and check the wires. Maybe you pinched or broke a wire. Other possibility is a tiny spot of solder landed on the board somewhere and is shorting the temp signal.

 

[Theo128]

Thanks for reply, which PAS Cable you mean? The two resistor i measure 0 ohms through it. Is that right?

 

[fechter]

A zero reading for the shunt resistors is correct. The temp sensor is buried in the potting on the top side of the board near where the blue phase wire goes in. Seems unlikely you hit the wires for it since they're buried in the potting. I'd still look carefully around that area. The temp sensor wires are skinny black wires. The sensor is attached to the curved heat spreader bar next to the FETs.

 

[E-mil]

With the stock display, it will cut off at about 61.5V. So it looks like maybe the Eggrider will get around that limit. What I'm not sure about is whether the 61.5v limit is monitored in the display or on the controller board. If it's on the controller then the Eggrider won't help.

The voltage regulator in the motor controller is rated for 80v max and as I recall, the stock FETs are rated for 75v. Even with the Eggrider, one of these other parts might not be happy at the higher voltage.

I'm a little late to the discussion but I'm almost positive it's the controller that sets the voltage limit of 57-61.5 (depending on the controller firmware) since it spits out an error code to the display.

As long as the feed thru the brown/orange wires is below this voltage by means of a V drop (the LEDs), the board logic is ok. I've tested this at 96 V after replacing the mosfets for 4110's.

 

[jwunder]

Hi, I tried the shunt mod. Well, to prepare for it anyway. I got as far as the depotting and removal, without the soldering the resistor, and then retested it on my bike. It gave a me a 30h error. Nothing works, except the display, and the voltage reported is about 6 volts lower than the actual battery voltage. Then I tried the same thing with another (buggy) controller which was giving me a 12h error, and it now also gives me 30h error (but the motor still runs inconsistently, throttle and PA, and the voltage is fine}.

Any ideas? I'm guessing I screwed up both controllers while prying them loose, although I was pretty careful. Strange that both controllers would both give me the same 30h error. Could be a harness or display suddenly gone bad, but that doesn't explain why the motor runs on one controller and not the other, or the lower voltage on one controller. For prying, I used screwdriver leverage against the mosfet bolts, and also gently underneath the paper semicircle. I depotted only as much as I had to. I tested continuity on all the controller wires and they're fine

 

[fechter]

30h is a communication failure. The display is unable to communicate with the controller. Hard to say what caused that, I would carefully check the connections between controller and display. Look for bent pins. 12h is a current sensor fault, indicating the current signal voltage coming from the shunt is out of the specified range.

 

[jwunder]

Thanks. I now think something I did to the first controller damaged the display (possibly by touching two wires together?). That's why I'm seeing the same 30h on the second controller instead of 12h. I have another display and controller ordered, but I'm reluctant to test the new display with first controller in case I damage another display. I guess I'll have to be more careful with new controller. Wish I could find some way to pry it loose without depotting.

 

[tomjasz]

I've seen a few pinched harnesses that have created an error 30. Zip ties were so tight the really thin inner wire (wires) got damaged. Not real common but it happens.

 

[jwunder]

I soldered the extra resistor on my buggy controller, just as a test, and I was impressed. Blistering acceleration up the country hills where I live. I could probably keep up with most cars and motorcycles around here. The rear suspension actually kicks back, which I never noticed before. The top speed is probably increased, although it's hard to tell with this damaged display since speedometer doesn't work anymore. My setup is, I think, torque limited, since I use a 65t chainring, so the top speed is helped by more current. Too scary to test around here. Although the throttle works, the PAS doesn't work right on this buggy controller, so I can't test it. If I go with a two extra 5ohm resistors on a new controller, I assume I'll need to cut the PAS current levels in half to get the same levels I was using. I use PAS 90% of the time, but the throttle is useful for quick trips to the store. I have considered Phaserunner, but after this test I'm not sure I need it.