Bike lives again! Soldering Shunt.

[dozentrio]

My bike has been out of commission all winter, due to problems charging batteries and drive-train alignment. I haven't had the motivation to fix it, since it was too hideously cold out to ride it much anyway. But now! Now spring is here, and yesterday I fixed my drive-train up. I accidentally melted a connector connecting my battery in reverse polarity (see connector rant) and in the process, found out that I had a bad connection at my connectors which was limiting my current draw. Put on new connectors, made sure they were tight, and I suddenly went from 1500 W to 1900 weee!! noticeably more torque. Having decided it isn't good enough, I'm now soldering the shunt on my ecrazyman 48v 800w controller. I hope it doesn't burn up on me. Hope to break the 2kW mark!!! My good old 450W Unite my1018 motor is screaming for mercy but it won't get any. No siree.

I am using 14 AWG stranded copper wire. Would there be much benefit for me to go to thicker wire? I am drawing close to 50 amps.

 

[amberwolf]

Are you getting any voltage drop across the wires from one end to the other at high currents? Or wire heating? If not, I wouldn't worry about it.

 

[dogman dan]

Even the best connectors tend to be the problem, rather than wire size. In construction, we used a lot of 100' extension cords, and often noticed that a saw might run better on one 100' 14 guage cord than on two 50' 12 guage cords, because of the fried connectors on the plugs. But wire size does matter, if you are soldering shunts, and pulling 2000 watts, 12 guage wire from controller to battery, and from controller to the point where the wires are about to enter the motor is not a bad idea. Big fat connectors too, instead of the 30 amp andersons.

 

[icecube57]

I have 30A Andersons and 12G all through my bike they dont heat up or feel warm. I just did 25-30m the other day pulling 2200w constant full WOT and everything was ice cold. My controller is 45A with the phase current set at 120A. I think that if you use andersons it does help to pry the clip up just a little bit to increase spring tension to make sure you have a good solid connection that wont accidently pull apart and increase in resistance on you.

One thing I have seen is a perfect crimp but to small of gauge wire. I was using my inverter as a dummy load and the inverter was pulling a solid 40A continuous load. The 16g wires I recycled from another project got hot enough for the insulation to become tacky. This later caused the connector housing to melt.

On the previous version of my build I was using 10G spade terminal connections. Id figure the are cheap enough to replace if they go awry. The connections got so much impedence that it caused my spiral wrap to melt. It caused the connectors to deform. It cause the phase wires to melt together. This only happened within the first 3-4 inches leading away from the connectors on both sides. I was pulling 4200w also. Despite all of that the actual rest of the harness itself was 16G it got warm but it didnt melt the insulation or feel tacky. So good connections can mean success or failure.

Just upgrade to 12g. For the amount of power you are running and from my experience it should be fine. Its only a 15-20 dollar investment. Crimp on andersons for all your power wires. I know i said i used the 30A on mine but just do the 45 for peace of mind. Methods uses 45A anderson on his 100v 100A build and they held up nicely. They are so under rated.

 

[dozentrio]

Wow, thanks! This helps a lot. I was about to get the 75 amp Andersons, hehe. Just in case of future upgrades. But if the 30-45 A ones hold up so well then I will definitely be using those.

But I am confused now. Sometimes I am able to get 44 amps from my bike, most of the time it is limited to about 38, and after I've drawn about 3-4 amp-hours from my pack, I can only get around 20 A. Is it possible that this variation is entirely due to the charge level of my cheap lipos? I have 12s 10Ah of cheap ebay RC packs.

I guess it makes sense, if voltage sag is 5 volts then the motor won't draw as much current. Something that annoys me though is I think the low voltage cut-off on my controller is putting a damper on things. I see the voltage drop to 42. The LVC on my ecrazyman is 41.5V, so maybe it starts to throttle current down or something. I may have to disable this. Although I guess it does help me conserve battery. Hmm.

Conclusion: Need more and better batteries.

 

[icecube57]

Normally the lower the voltage the more it will pull. Controllers arent perfect in limiting amps. It also depends on how much Back EMF is in the motor. That will limit amps. If you are already up to speed the amps will be less than if you are doing a hard start. Normal "stall" starts will show a peak and then the controller will pull back to the appropriate value programmed.

 

[ZOMGVTEK]

Im running 45A andersons on 14AWG stranded copper, pulling 60-65A at 48V and the wire only gets slightly warm if youre going up a steep hill for a while and the motor is pulling 40-60A all the way.

But yes, 14 gauge has a very noticeable voltage drop and im running 10 or 8 gauge once I get my new batteries. Probably switching to better connectors as well. The little andersons cant handle 10 or 8 gauge terribly well.

 

[dozentrio]

Does anyone know how to get rid of or adjust the LVC on ecrazyman controller? I have the 48V 800W. I have decided that it might be the LVC on my controller that is limiting me...

I suppose it would help if I showed a picture of it, but I glued it shut. Argh.

 

[dnmun]

if your infineon controller dropped below 41.5V, the controller shuts off entirely. it does not go wobbly and weaken, it shuts off the controller totally. the LVC is there to prevent the over discharge of SLA packs, and if you have a BMS on your lipo or lifepo4 pack, the LVC on the controller does not matter and will not interfere unless your BMS will allow the pack to go lower than the 41.5V cutoff on the controller.

i think the lipo guys decided to not allow the cells to go below 3V, but some say 2.7V is the limit.

 

[mwkeefer]

Not trying to be rude, but the LVC on the Infineon isn't there to protect SLA or any other chemistry packs - it's there to protect the controller itself, the LM317 rail needs the voltage to remain above 41.5v in on a 48v calibrated controller to ensure proper logic and gate voltages... sag here could cause massive failure (kentucky fried face plant in a front install).

With LiPo 3v is supposed to be the "normal" chemistry LVC but don't count on the infineon doing this for you since it's pack level and a weak cell could drop below the 3v mark with ease trashing that parallel cluster of cells from the series string... Unlike LiFePo4 chemistry which is more tolerant (not as dangerous) of overdischarge.

For best life of a LiPo pack, we usually try to leave 20-40% of the SOC remaining and that means a cutout at cell level of 3.20v - 3.4v. This may sound counterintuitive or as if it works against the chemistry energy density but what we have found is that packs will last alot longer when charged to 98-99% SOC and when discharge is terminated at 20-40% remaining.

Hope it helps!

-Mike

 

[dozentrio]

Yes, thank you both.

But if the controller is supposed to cut out completely, and voltage sag+LVC is NOT what is causing the drastic decrease in power, then why can I only draw about 2 amp-hours from my 10ah lipo pack before it drops to 2C discharge? I am getting less than 20 amps from my pack for some reason. Is it possible that they are just such shitty batteries?

I will continue to investigate. Connectors arriving soon, though I doubt that is the problem. I suppose it's possible that I'm not fully charging the pack, and so I'm going between 4-6 amp-hours in the pack when I use it. I stop charging at about 4.14V/cell because I'd rather not overcharge. Suppose it's also possible that the pack is damaged from pre-watt meter abuse.

 

[mwkeefer]

Yes, thank you both.

But if the controller is supposed to cut out completely, and voltage sag+LVC is NOT what is causing the drastic decrease in power, then why can I only draw about 2 amp-hours from my 10ah lipo pack before it drops to 2C discharge? I am getting less than 20 amps from my pack for some reason. Is it possible that they are just such shitty batteries?

I will continue to investigate. Connectors arriving soon, though I doubt that is the problem. I suppose it's possible that I'm not fully charging the pack, and so I'm going between 4-6 amp-hours in the pack when I use it. I stop charging at about 4.14V/cell because I'd rather not overcharge. Suppose it's also possible that the pack is damaged from pre-watt meter abuse.

Now we may be getting somewhere - after 2AH drawn at 40-60A, what is the temperature of the lipos?

10C10AH lipo (2P right?) should handle 100A continuously throughout the discharge process but... 14,16G it's kinda pushing it...

The only thing I can think of for a DROP in current would be a hugh increase in the internal resistance of the pack since it doesn't sound like you have anything else but the controller limiting the current.

Do you have cell level low voltage protection? If not then you may have skunked a cell by going down past the minimum voltage. What do you use for balancing and charging the system? Does the pack have a BMS of some sort?

I would go for a ride while you get your 40-60AH... then when the 20A stuff begins... simply check the cell level voltages for one which has begun to decline faster than the others... it seems like (imho) one or more parallel clusters of cells (if you paralleled at cell level, if at pack level... it may be a quick fix) got skunked so they are now only 2.5AH of capacity instead of 5AH and thus you get your 2AH at good C rates and then the stronger of the 2 cells is making up for the difference but I tell you each time that happens you are likely killing that low cell or parallel cell group.

Try this too... Reconfigure your controller for 37.5A battery current and 75A for phase current (2x I know but try it) and see if it rides the whole bit without issue....

Also how long are your power connectors from your pack... and how long are your phase wires.

If you want to really test this out - drop the connectors and use copper crimp collars then solder them for the phase lines atleast... you can then eliminate that side...

Finally - how hot does the controller feel when you get this 40A to 20A drop.

Keep in mind I believe it's a bad cell or parallel cell group in your pack but these are other possibilities in addition to just bad connectors.

-Mike

 

[dnmun]

my intention was to educate him about how it appears, when the LVC cuts off the controller, since he is not familiar with it.

i don't know about his batteries. hopefully he had some protection.

 

[Russell]

Yes, thank you both.

But if the controller is supposed to cut out completely, and voltage sag+LVC is NOT what is causing the drastic decrease in power, then why can I only draw about 2 amp-hours from my 10ah lipo pack before it drops to 2C discharge? I am getting less than 20 amps from my pack for some reason. Is it possible that they are just such shitty batteries?

I will continue to investigate. Connectors arriving soon, though I doubt that is the problem. I suppose it's possible that I'm not fully charging the pack, and so I'm going between 4-6 amp-hours in the pack when I use it. I stop charging at about 4.14V/cell because I'd rather not overcharge. Suppose it's also possible that the pack is damaged from pre-watt meter abuse.

My experience with hitting the controller LVC is the controller does NOT cut out completely. When my battery weakens to the point where it hits the LVC of the controller it will limit the current draw from the battery to keep the voltage right at the LVC voltage. Only when a battery is well below the LVC and can not maintain any current at the LVC voltage will the controller cease to function.

I also take exception to the statement that the controller LVC is NOT there to protect the battery...that IS the whole idea behind the LVC and the voltage ranges were initially selected because 12V SLA's are basically dead at 10.5V.

-R

 

[dozentrio]

First I suppose some background on what exactly I'm using might be helpful.



I have a 450W 24V Unite brushed motor clamped to my rear swingarm, (more pictures to be shown in Photos and Videos) and a blower motor pulling air through it to cool it. I built my pack from ten 3s 5000mAh Lipo RC batteries. Cheap ebay ones. So, in the end I have 12s 2p, putting me at 44.4V nom, 10Ah. They are rated for 15C but supposedly can only handle about 6C. 60Amps is still sufficient for me, even 40 would be fine.

Now, because I was eager to get out and ride, I didn't wait till I could afford a BMS, so the cells are completely unprotected in any way, and not paralleled at cell level. Actually, they're not paralleled at all except at ground and 44.4V. They stay relatively well balanced though. I do want to improve on this though once I have the time/money. I am a little concerned that I drew too much charge from my pack before I had my watt-meter. I was badly underestimating how much charge my system drew

The cells don't get warm, the controller doesn't get warm, the cables don't get warm. I do have a fair length of wire that the power is running through. Battery - watt-meter - controller - motor. I have my battery pack on a baggage carrier behind my seat, and have to run cable all the way up to my handle bars and back, for the watt-meter.