Mosfets don't have diode junctions for drain to source inside the mosfet. I'm not referring to the back diodes either. The mosfet itself is not forward or reverse biased drain to source or source to drain. Current flows either direction through the mosfet. This is why connecting to C- for power works. You are correct, from a bipolar transistor perspective, "flowing backwards" . Bipolars do have a diode junction from collector to emitter that allows current flow in only one direction. From the perspective of convention and general logic, there is a forward direction and a backwards direction for mosfets, but in reality, no not really. I'm referring to pure mosfets with no back diodes. Add the diode and then you have design limits that may effect how you use the mosfet. Also, even the craptastic HY3410 probably has a lower Rds than the back diode so if the charge mosfets are on, they are going to be doing the lion share of conducting current despite the fact that the back diodes are forward biased in the C- mosfets in a BMS.Fionn wrote: ↑Feb 21 2018 7:56amI'm interested in using the charge termination to limit charging to lower than the standard charger output (or above what it's set to as an additional safety net).ElectricGod wrote: ↑Feb 20 2018 2:17pmsee those 2 solder holes at the middle, right hand edge of the board? That's P-. Connect your EV here. If you care about having mosfets that control charge then run your charging port to C-. I'm never concerned with charge current. AT it's best, it's maybe 50% of discharge current so I never use C-. Charging and discharging for me always go through P-. Also, if you use regen, that happens through P-...that's just a form of charging. If that's safe through P- and it is, so is regular charging through P-. For the super safety conscious types, C- might have a place in an EV, but more most people it's completely extraneous. Never power your EV through C-. It can be done, but you double the mosfets in series and the related losses. What's the point in doing that? Why they show powering the EV from C- is a mystery to me. I sure wouldn't do that! I'd prefer to have all those wasted C- mosfets on the P- side where they actually matter.
Of course I've connected up BMS wrong before too. This is a couple of 16S 50 amp BMS that are currently in use and connected via the C- port, not P- (center solder hole)...DOPE! It works, just not optimal.
I also have an interest (like one or two other posters here) in using these with lithium based renewable energy systems where having separate load and charge disconnects is very useful.
I hope to try and trace out a partial schematic for this board over the weekend to make sense of the implementation.
As you say, they seem to be recommending running power through both FET banks, which is not an optimal solution.
Also, it seems strange that both FET banks drains are connected together while the load current is expected to pass through them in series.
Does this not mean the charge FETs are reverse biased?
N channel fets.jpg
This reminds me of another detail. If the C- mosfets turn off, the back diodes do NOT turn off and they continue to conduct since from C- to the internals of the BMS, they are still forward biased. IE: Getting the BMS to actually stop charging in an over current situation wont happen. It will just be limited somewhat by the higher resistance of the back diodes. I'm going to guess that the diode compared to the 100mOhms of the mosfet is 50% more resistance? I really don't know for sure since I've never tested this and haven't looked at the data sheet to see if this is a listed spec or not.
For your set up, you want to connect your charger to C- and the EV to P-. It can all be run from P- or C- and still work...just less optimally from C-. Of course, charging from C- also requires flowing current through 2 sets of mosfets in series. IMHO, this too is less than optimal.
These BMS look to support charge current limiting via the PC app only, but I haven't messed with it at all since I deliberately avoid the C- mosfets. My suspicion is that the charge mosfets simply shut off if the charge current exceeds the set limit. This is typically how current limiting on a BMS works. They don't regulate anything. They don't provide any form of constant current control. Consider them to be an on/off switch and nothing more. If you want to limit charge current, do that at the charger. All of my chargers have a bank of resistors on them for this purpose. My method is rather "brute force" in nature, but very simple to use and implement. The best way to limit current is to drop the output voltage of the charger to closer to the current pack voltage. There are after market boards that can integrate with most PSU's for this purpose. I forget the details, but there's a thread on ES for a board that does this. If you buy a well designed charger, they usually have this function built into them in some way. That is the way to limit current when charging. The BMS charge mosfets are just on/off switches. I seriously doubt they do more than that.
For anything 100 volts or less and you want a robust and well designed, highly flexable charger, consider buying the Chargery C10325 charger. It does real CC control, voltage control and lots more and it's good for 1500 watts. Good luck trying to blow it up. It's fully protected. Of all the manufactured chargers I've tried out, it's my favorite.