tomjasz
1 GW
staldor said:
Which motor? Date? Which battery case?
Bafang BBSxx - fault finding and fixes
- Thread starter Slowrider
- Start date
Which motor? Date? Which battery case?
Drum said:
I have now checked all battery cells, and they were perfectly balanced at 3.99V.
I also tried the two terminals between the battery case and holder. It was very difficult to get the case lose from the holder. And after re installing it, the bike lasted longer than normal. So I'm starting to believe it could something here.
Could you explain a bit more how you fixed this? Are you talking about some wires inside the battery case?
tomjasz said:
Motor is first version of BBS02, I get the same problem on two different motors (bought 2015 and 2016) and two different displays. This is the battery
https://www.greenbikekit.com/e-bike-lithium-battery/li-ion/48v11-6ah-e-bike-frame-battery-with-5v-usb-output.html
I really wonder how this flashing battery symbol is triggered.
tomjasz
1 GW
Thanks! I sounded like the problems Shark cases have with the slots corroding. The Dolphin cases don't have as many connection to base issues, but it sounds more and more like it is a bad connection. Inspect the pins and the plug and treat with some dielectric grease, or Boeshield, after a careful cleaning. 3.99V/cell is low for a 18650 cell. Is this an old battery or are you trying to conserve by lower charges?
Drum
100 W
Hi Staldor, About the battery base modification:
The bike I was working on was a Chinese hub motor bike, and it would sometimes ride fine then lose power for no obvious reason. I noticed that removing and re-fitting the battery seemed to restore power, so checked the contacts.
The battery had 4 "female" (two-sided) contacts, two of them were positive, two were negative.
The battery base had 4 "blade" contacts that slid into the battery contacts. Looking inside the battery base, I found that the wires to the controller were only soldered to the outer contacts, and the inner two were not connected to anything.
I soldered short lengths of wire between the inner and outer contacts on each side, so the battery current could flow through both contacts on each side instead of just one.
I also pinched the contacts on the battery so they "gripped" the blades tighter.
After those modifications, the problem disappeared.
The bike I was working on was a Chinese hub motor bike, and it would sometimes ride fine then lose power for no obvious reason. I noticed that removing and re-fitting the battery seemed to restore power, so checked the contacts.
The battery had 4 "female" (two-sided) contacts, two of them were positive, two were negative.
The battery base had 4 "blade" contacts that slid into the battery contacts. Looking inside the battery base, I found that the wires to the controller were only soldered to the outer contacts, and the inner two were not connected to anything.
I soldered short lengths of wire between the inner and outer contacts on each side, so the battery current could flow through both contacts on each side instead of just one.
I also pinched the contacts on the battery so they "gripped" the blades tighter.
After those modifications, the problem disappeared.
tomjasz said:
Yeah, i agree. Somehow the shark case was also very stuck when I tried to get the case loose from the holder. Wonder if it can have something to do with the problem. Anyway, after opening up the battery case to check the cells voltage, the battery actually work again. But I'm afraid the problem will come back. I have some dielectric grease somewhere which I will try out on the terminals. Could maybe also be som loose wires to the BMS. The 3.99V per cell voltage was after a hilly test ride, so it is probably closer to 4.2V when fully charged.
You were right tomjasz. It is the battery terminals. I have tried to clean the plugs. What is the best way of cleaning such terminals? Looks almost like there is some parts missing at the minus terminal (left terminal), or maybe just filled with corrosion.
I had some dielectric grease, but this did not help. How should one apply dielectric grease? I think i used to much.
Lastly I cleaned the terminals from the grease and I put some steel wool inside this hole that seems clogged. This seems to have helped. I will try it out tomorrow and see if it holds my commute. But let me know if anyone have any cleaning ideas. Permatex battery cleaner maybe?
I had some dielectric grease, but this did not help. How should one apply dielectric grease? I think i used to much.
Lastly I cleaned the terminals from the grease and I put some steel wool inside this hole that seems clogged. This seems to have helped. I will try it out tomorrow and see if it holds my commute. But let me know if anyone have any cleaning ideas. Permatex battery cleaner maybe?
Ivanovitch_k
10 W
staldor said:
Hard to tell from the picture, but It looks like the negative contact arced hard or often enough to vaporize the metal and deposit carbon (soot-like) inside the neg. post. And it seems the increased contact resistance generated enough heat to start melting the plastic housing around the pin.
You could try filing / small-grit sand paper on the pin & plug to remove the carbon...
Here is what a brand new connector looks like (slightly different model but from the same manufacturer -Reention-). Shiny.
Also, I would advise against dielectric grease / beoshield use directly on the contacts. Both are insulating materials.
Grease is sometimes used on screw-in terminals or on tight-coupling quality connectors. With them, contact force is high enough to push the grease out (surrounding the contact area with a layer of insulating grease) and still maintain a good metal-on-metal coupling.
On those cheap Reention & Hailong connectors, the grease is more damaging than useful as the fit, especially contact retention force, is far from perfect and loosens over time, even more so if the battery is moving / vibrating a lot.
If you really need additional moisture / water ingress protection, you can put sone dielectric grease on the mating flanks of the connector barrel.
tomjasz
1 GW
staldor said:
Yeah, Reention another parts pit! Negative looks carbon coated, phosphoric acid on a Q tip works and leaves 99% of the Coke to drink. I’ve found that these unattended connectors in Hailong as well as Reention cases are problematic. I find Boeshield to never insulate. But clean connections before treatment is important. I deal with this constantly especially in slushy winters, water, salt, ice and snow will take a toll if preventive maintenace doesn’t happen. Clean a treated connections make for good power flow.
Thanks for your helpful replies Ivanovitch_k and tomjasz! We had a long winter in Norway as well, so this has probably been the problem.
I will try to find some phosphoric acid or try it with some sandpaper. I also saw on the internet that baking soda mixed with water could clean battery terminals?
Anyway, i have contacted the seller, and they will send me new connectors. Not looking forward to the soldering work, but i guess I will manage.
I will try to find some phosphoric acid or try it with some sandpaper. I also saw on the internet that baking soda mixed with water could clean battery terminals?
Anyway, i have contacted the seller, and they will send me new connectors. Not looking forward to the soldering work, but i guess I will manage.
Ivanovitch_k
10 W
tomjasz said:
boeshield is an oil / spray stuff right ? it's much thinner than silicone grease and that might well be why it's working.
One could also use noalox antioxydant paste. It's a non-insulating anti-corrosion grease made for aluminium & cooper electrical connections.
Also, if the bike is stored outside, I would try to find some kind of plug to cover the connector (a plastic bottle cap ?).
staldor said:
Tomjasz wasn't joking when he referred Coke
! Just use coca cola (a non sugary kind = no sticky residue), it's packed with phosphoric acid.staldor said:
Use a powerful enough iron (40W is the bare minimum) with a non-damaged flathead / chisel tip (avoid conical like the plage, they suck at thermal transfer).
Use a lot of flux. Use non RoHS (I.E leaded) solder. Pre-tin both the connector and the wire. Let both cool down to room temp to not overheat insulations.
Reflux, and feed some solder at the interface between the iron tip and the solidified solder in the connector plug, this greatly increase thermal transfer. Assemble.
Once cooled down, the joint must be shiny. If not, not enough heat was applied and the joint is dry, just reheat it until liquid otherwise the connection will be unreliable.
Thanks for the soldering tips Ivanovitch_k! My soldering iron suck, but luckily i have a good one at work.
I found a place were they sell phosphoric acid. Sounds better than coke. But let see what I end up buying.
The steel wool hack have actually hold up my commute today. And greenbikekit.com have been kind enough to send me a new connector (battery is from 2015 so did not expect warranty). So maybe I will just wait for the new connector.
I found a place were they sell phosphoric acid. Sounds better than coke. But let see what I end up buying.
The steel wool hack have actually hold up my commute today. And greenbikekit.com have been kind enough to send me a new connector (battery is from 2015 so did not expect warranty). So maybe I will just wait for the new connector.
tomjasz
1 GW
You only need the low levels found in Coke. Phosphoric acid solutions may be a problem! Careful, no one is suggesting anything other than mild solutions, Coke!staldor said:I found a place were they sell phosphoric acid. Sounds better than coke. But let see what I end up buying.
Lurkin
100 kW
My dolphin pack connect socket did exactly the same thing. I trawled the internet for hours to try and find an exact replacement - no dice. Molex was a close as it would get that was attainable directly
I gave up, contacted Em3ev to obtain replacement parts. It was caused by the socket coming loose from the pack over time and going unnoticed, presumably the loose connection created a poor enough connection to ark/spark causing the melting.
However, this is part assumption - I have always assumed the socket is rated and capable of the current drawn. Should it be of a lessor ability, it could also contribute to premature failure.
In any case, Em3ev were good to deal with and it the bike is back up and running without issue now. 8)
I gave up, contacted Em3ev to obtain replacement parts. It was caused by the socket coming loose from the pack over time and going unnoticed, presumably the loose connection created a poor enough connection to ark/spark causing the melting.
However, this is part assumption - I have always assumed the socket is rated and capable of the current drawn. Should it be of a lessor ability, it could also contribute to premature failure.
In any case, Em3ev were good to deal with and it the bike is back up and running without issue now. 8)
Hi guys!
I have a strange problem on a BBS02. The speed shown on the display varies quite much, just when I'm on the 8th and 9th modes. By varies, I mean it's not stable, it goes from 30 to 50 km/h in a second. I don't have that problem in the other modes. Dou you know what it could be?
Thanks in advance for your help.
Didier
I have a strange problem on a BBS02. The speed shown on the display varies quite much, just when I'm on the 8th and 9th modes. By varies, I mean it's not stable, it goes from 30 to 50 km/h in a second. I don't have that problem in the other modes. Dou you know what it could be?
Thanks in advance for your help.
Didier
dustNbone
10 kW
Make sure the speed sensor is reading properly. You might need to adjust the magnet/sensor to get them passing closer to each other, or move the sensor forward or backward on the chainstay to get it in the right position. I remember mine was a bit picky about position.
The sensor measures wheel rotation in pulses, if it's missing some of these pulses it's going to show a lower speed. It's more likely to miss pulses at higher speed.
The sensor measures wheel rotation in pulses, if it's missing some of these pulses it's going to show a lower speed. It's more likely to miss pulses at higher speed.
Animalector said:I replaced the FETS on the BBS02 controller without success I thought I would post here.. The first failure was an impact with a rock due to bottoming out... replaced the FETs on the shorted phase, and then the symptom was that it ran on throttle for around 5 seconds, then when I closed the throttle, the motor kept spinning. I pedalled backwards or something to get it to stop, and once it stopped it was dead, wouldn't spin again. this time when I pulled out the FETs I decided to measure them all, and discovered that it was only the high side that was cactus.
So this is the photo of the controller...
Going from memory (it's been a week or so since I looked at it) there is a single FET on the 'high' side ( the one on the left), and parallel FETs on the low side. It's the high side FET that keeps blowing, which is on the left.
I thought perhaps there was a gate resistor missing, and it was a floating gate, from the looks there are no components missing except the pads on the far right of the pic. . Can any guru's suggest why perhaps a high side FET might keep blowing? I should probably measure gate resistances. etc etc. maybe there's an issue with the driver and it's feeding 48V to the gate.. don't know..
Thanks
Andy
I have very similar problem. For some time controller works ok, but when I release throttle it seems like some mosfets not get closed fully and are shorting a winding. I can even hear humming sound. Controller get very hot at this time. This is hapening even if brakes are pressed. It must be something with gate driving. At first original P75NF75 fet shorted, so I changed all of them to IRFB7730. After some km fet went again. I can't see nothing obvious: pull-downs are ok, mosfet driver fan7388 supply voltage is 13V. Maybe someone analysed how this controller works and could help with this problem?
Animalector
10 kW
i gave up and bought another controller
it was under $100 and not worth my time to fix it
Good luck
Andy
it was under $100 and not worth my time to fix it
Good luck
Andy
Justas said:
That doesn't sound good. If the FETs or gate driver were bad, it would not run OK at any time and probably trip the BMS. This sounds more like a processor issue. There should be no drive to the FETs when the brake is on.
Check the 5v supply on the hall or PAS sensor connector. If it is running low, it probably indicates something is shorted on that supply.
I could buy a new controller but this is a matter of principle and a fun way learning stuff.
Checked 5V rail, it's about 4.7V. Removed chain and applied minimum throttle - heat sink on the blowing FET side getting hot very fast and it takes about 50W to spin at minimum speed. Need to borrow an oscilloscope and look what the hell is going on with that gate drive. For know will try to see MCU drive output with logical analyser and maybe hall output.
Another thought: maybe motor wingdings partly shorted? But motor seems to run smoothly and not getting hot.
Checked 5V rail, it's about 4.7V. Removed chain and applied minimum throttle - heat sink on the blowing FET side getting hot very fast and it takes about 50W to spin at minimum speed. Need to borrow an oscilloscope and look what the hell is going on with that gate drive. For know will try to see MCU drive output with logical analyser and maybe hall output.
Another thought: maybe motor wingdings partly shorted? But motor seems to run smoothly and not getting hot.
If a motor winding is shorted, it will make the motor very hard to spin. This can be tested by trying to spin the rear wheel backward. Normally there is a fair amount of resistance, but if a winding was shorted, it would be very difficult to roll the bike backward. You may need to disconnect the phase wires to eliminate the controller as a possible short.
Overheating FET sounds like it is not fully off when it's supposed to be. If you put a scope on the gates, you can see if they are all going to zero. If one was stuck fully on, it would cause shoot through which I think would instantly blow something up.
Another test is to measure resistance from each gate to the ground and 12v feeding the gate driver. None should look shorted.
When FETs blow, it is not uncommon for the gate driver to be damaged by it. A FAN7388 is around $2 from DigiKey.
Overheating FET sounds like it is not fully off when it's supposed to be. If you put a scope on the gates, you can see if they are all going to zero. If one was stuck fully on, it would cause shoot through which I think would instantly blow something up.
Another test is to measure resistance from each gate to the ground and 12v feeding the gate driver. None should look shorted.
When FETs blow, it is not uncommon for the gate driver to be damaged by it. A FAN7388 is around $2 from DigiKey.
What if windings of same phase shorted to itself? I mean the turn count of this phase would decrease.
All gates to ground measures the pull-down resistor: 10k for low side and 100k for high.
Tried various programing parameters in case something was messed up, but nothing helped. Will try to diagnose a bit more and when will change the fet driver.
All gates to ground measures the pull-down resistor: 10k for low side and 100k for high.
Tried various programing parameters in case something was messed up, but nothing helped. Will try to diagnose a bit more and when will change the fet driver.
Justas said:
If a winding shorted to itself, it would still have a huge resistance to turning and the shorted winding would get really hot.
You can inspect the windings and look for burned spots.
Gate stopper resistor burned or bridged? Something to check,
especially if the Mosfet that it was driving has been replaced.
Bridged stopper may work, sometimes fail to stop oscillations.
Burned stopper might not allow Mosfet to switch quick enough.
Sometimes the stopper isn't a resistor, but a ferrite bead on
the gate lead of a Mosfet. Lossy type ferrite, not intended
to be an inductor, but a resistor at high frequency. Wrong
kind replacement for lost bead could make oscillations too.
I havn't seen the schematic to confirm that you even have such
a resistor or bead. More of a general rule of thumb thing...
especially if the Mosfet that it was driving has been replaced.
Bridged stopper may work, sometimes fail to stop oscillations.
Burned stopper might not allow Mosfet to switch quick enough.
Sometimes the stopper isn't a resistor, but a ferrite bead on
the gate lead of a Mosfet. Lossy type ferrite, not intended
to be an inductor, but a resistor at high frequency. Wrong
kind replacement for lost bead could make oscillations too.
I havn't seen the schematic to confirm that you even have such
a resistor or bead. More of a general rule of thumb thing...
It's a little resistor on those. Good point to measure those. Measure resistance from the gates to the output pins on the gate driver. If one went open, the gate might not turn off fully.
I don't have one handy right now, but later I can measure and tell you what the resistance is supposed to be. Usually they're around 100 ohms.
I don't have one handy right now, but later I can measure and tell you what the resistance is supposed to be. Usually they're around 100 ohms.
