Quote:
Originally Posted by jonyjoe101
The problem with the mosfets is they leak voltage and the battery will read about 11.7 volts when the bms activates. Some chargers seeing a "low" battery will try to force a charge into the battery and since the bms won't allow charging you might get voltage surges. I lost 12 volt fans/pumps and led lights that couldnt handle the surges (while charging with solar).
|
I think you are describing a situation where the BMS is disconnected (i.e. a series MOSFET is OFF) but still showing some terminal voltage. The MOSFET is a very high impedance when OFF. It is as if it is not there. If there is a voltage it must becoming from a solar controller or somewhere else.
If the BMS is sensing the terminal voltage of the battery that might be applying a voltage. I'm guessing somewhat as I'm not that clear on what the topology is of a BMS. My assumption is that is sensing the individual cells and if something is outside the limits it disconnects with a series MOSFET. I guess it also has some small equalization currents it can apply but dor the Daly that is only 30 mA.
This is another reason to improve the logic in the DC-DC charger. The Renogy describes something it calls "
Lithium Activation" for starting a BMS disconnected battery.
Quote:
Originally Posted by jonyjoe101
I tried using a overvoltage protection relay to protect my devices but they operate to slow prevent damage. One thing that does work is connecting all 12 volt devices to a 12 volt voltage stabilizer.
I'm using the chargery on my 4s 220ah lifepo4 batterybank, I havent lost any devices since I installed this bms 2 years ago.
|
That is good to know. I have now ordered $200 in BMS/Active balancers and whatever solution I choose it is multiplied by 2 because of my two separate batteries.
I'm not looking forward to changing my BMS but it is good to know there are fallbacks.
Quote:
Originally Posted by jonyjoe101
The problem with a mosfet bms activating is that it remains activated until manually reset. While its activated whatever is trying to charge it will keep cycling back and force trying to force a charge on the battery. The TVS has to be able to handle that constantly. With a mechanical relay, the battery will be out of the loop once the bms activates.
|
As noted above, your BMS should never disconnect. I have the Rich Solar 40Amp controller with separate solar, Battery, and Load terminals. If the batteries get low the solar controlled should disconnect the load way before the BMS disconnects. Basically, you need to watch the batteries and make sure that either through loads or self-discharge the LiFePO4 packs do not trigger their respective BMSs to disconnect. If you are going into long-term storage where there is the possibility of draining the batteries, then I would disconnect the solar ( I have a 15 amp disconnect).
So basically I do not understand the rationale of having a mechanical relay disconnect. Is this between the LiFeLP4 and the solar charge controller?
I have this 120V 10A charger. It seems to be pretty stable when I charge a lead-acid battery. And it does seem to shut down but it is a lead-acid battery charger. Maybe I need to rethink that as a long-term charger.
Quote:
Originally Posted by jonyjoe101
In my situation I only charge my lifepo4 with solar, so with the alternator might be different results. But since you mention a load dump, you will be ok if charging with the alternator only, since when the bms activates there shouldnt be any surges since the 12 volt lead acid start battery will still be accepting the alternator output, you might not even need the TVS.
|
[/QUOTE]
So if you are only charging with solar, what happens to your solar controller when there is a BMS initiated battery disconnect??
Load dump is generally related to anything inductive where you disconnect the load (i.e. V=L di/dt). This is primarily the alternator but could also be a motor or possibly long leads to a battery that is disconnected. The voltage specs are like +/- 100 volts. Generally, any series resistance helps limit the current spikes. With Lithium, there is much less battery internal resistance. The hardest thing to protect is surges at low impedance power terminals. So Lithium is among the worst.
One thing that I have come much more aware of is the danger of a BMS disconnecting a battery from a solar charger with a high solar panel input. With the disconnected battery input, almost all of the solar controllers say DO NOT DO THIS! Because of this, I'm setting g all of my LiFePO4 battery fusing based on a catastrophic failure level (way above active control limits).
I just came up with another problem scenario.
For example, the solar charge controller should only push 25 amp for my 350-watt panel to the battery. I will also have a 10 amp 120 VAC shore power charger. The Renogy will be capable of 40 amps so this is 75 amps maximum. If I disconnect the battery because of the over current I will blow the solar controller.
This is an unlikely scenario but it is possible if I am sitting outside, onshore power, and start the engine.
OK I just decided I need to look at the house battery voltage before deciding to push 40 amps into it based on alternator voltage.
So I searched for "voltage stabilizer' and found these. They all seem to be DC to DC converters. The active power supply control is going to be in the low kHz range. If there is any capacitance then DC to DC will do something so I can not know how to quantify the difference between a "stabilizer' DC-to-DC to the other plain DC-to-DC converters. I have a 20 amp breaker on my DC loads (excluding invertor) so I could pair up a couple of the 10 amp DC-to-DC units for $40. I don't know how to compare that to spending $100 for two of the official "stabilizers".
Secondly, you have to ask what is this doing that the Rich solar controller is not doing? (The Rich has separate load terminals) If a spike is coming down the battery terminals does it just propagate through the solar controller and into the load? If so what is the point of the separate load terminals? The solar controller will try and keep the battery terminals stabilized depending upon the charge mode. So I'm not sure what the difference would be. Now if you did not have separate load terminals then yet the stabilized will isolate battery/charger junk from the loads.
So enough guessing, I drew up what is a likely solar charge controller topology where there are separate terminals for each collar, battery, and loads.
https://www.amazon.com/-/es/gp/produ...2L3PSG09&psc=1
https://www.amazon.com/dp/B082GD23W3...2s9dHJ1ZQ&th=1
I think the HSS (e.g. VNQ660SP) will go a long way to remove transient if not stabilize like a DC to DC convertor. It is not a DC to DC so it cant do a voltage translation like Buck-Boost. But what it does is active controls the gate on the MOSFET to limit current or open the circuit completely in the case of overcurrent/thermal. That little PowerISO-10 will provide 10 am per 4 channels. You have to keep it cool but it works very well.
I designed this product as a sorta hobby, sorta see what it takes to produce a commercial product. It worked very well and is primarily possible because of the VNQ660 type tech .
You can see it is pretty fast with turn-on and turn-off slopes of 300-400 Volts/mSec. That means a 10-volt swing happens in 0.03 mSec This is approaching 15 Khz but doesn't even consider the transient surge suppression. It also has real transit performance with respect to load dump (see Table 12). It doesn't regulate but does real well with junk!
I will have to heat sink them but will probably fabricate a small device to pass the power leads through. i plan to make something for some bulkhead connectors on the battery box.
These devices are just a smart FET with additional functions so once they are ON they are bidirectional. I don't think it will matter if a battery is charging or discharging.
https://www.amazon.com/gp/product/B0...?ie=UTF8&psc=1
https://www.thegsresources.com/_foru...State+Powerbox
looks promising
Linear Mode
