Series connection break Idea - please comment

[whereswally606]

Series connection break Idea

So in my moped I have currently 16s lifepo4 eventually 24s

I wish to use bc168 chargers x 4 (3 initially)
to charger 6s sub packs to 3.6v

Here I have an issue of needing to have series disconnects after each 6s.

The problem is having high current wires fed to somewhere accessible to put a loop key (or 2 or 3)
also I don't want to increase the length of the cell connections if at all possible.

It would be great if I could use relays to do this so I don't have to make up long high current wires however I have little experience with relays and need to be absolutely sure that the cells are disconnected when charging and wont disconnect when running the moped scooter. Obviously they also have to be man/woman/relay enough for the job i.e. be able to take the max battery amps that the controller with draw or receive via regenerative braking.

Any help would be greatly appreciated.

The bike is already running and will get a BMS shortly so all this may just be only musing the idea over but I like the idea of tighter control of charging at the cell level and to a lower specific voltage hence improving cycle life.

 

[Eteck]

Here I have an issue of needing to have series disconnects after each 6s....and need to be absolutely sure that the cells are disconnected when charging

Are you sure that is so critical if you are going to use 3-4 independent chargers ?
The device info says: "There is no conflict between each circuit. The power can maintain 8A output, all voltage , current will be display accordingly."
You might contact the seller/manufacturer and see what they say for your intended scheme.

 

[amberwolf]

Assuming the chargers are not themselves isolated (totally independent input power and output power), then:

As long as the "12v" supply for each charger is isolated and independent, there is no need to further isolate pack sections.

If you use a single supply for all chargers, then you must isolate the pack sections.


If all chargers are themselves isolated, then nothing else need be done.



problem with relays is if they get enough mechanical shock they'll bounce, and you get arcing on the contacts if current is flowing when this happens durring a ride. can weld contacts together, and then you think the pack is isolated when you engage the relays but it isn't, and then you get kapowie when you hook up the non-isolated chargers.

 

[Alan B]

BC168's are excellent chargers. They have been found to interfere with one another if the balance wiring shares current, even though supplies are isolated. Operating them isolated may also cause voltage differentials between their casings.

Using LED supplies is an alternative that has worked well for me. They are isolated and their cases remain grounded, even though their outputs are connected to points in the pack. I used three of them to make an 18S 1KW charger. I brought out 14 gauge wires from the three 6S sections in the pack so the current was separate and the three supplies didn't have to be interconnected at all. I have a thread around here somewhere on it.

 

[Alan B]

Here is the thread: https://endless-sphere.com/forums/viewtopic.php?f=14&t=42169

 

[whereswally606]

AW, thanks for the advice about relays, found 12v solenoids at £12 each which have both normally open and normally closed connections but I think that higher gauge wire out of the pack is better so going with your advice and not down the relay route.

Alan B, Lifepo4 is a funny beast. It appears the best way of charging it is to count energy in and energy out. currently I'm bulk charging the lot using 3 12v HP server supplies and a cheap 24v Meanwell clone with a fetcher v3 limiter board on it all in series. Just need to finish the Fechter Goodrum bms bought from Zenid but wiring it in is something I've not yet worked out. where to put it and how to cool it.

for the sake of getting out and using it I've been charging monitoring at cell level with Flytron's Icells 16 cells 2x8. not ideal but accurate enough. and while discharging using a bvm8 with a low cut off beeper at 2.85v

I blew up 2 of my bc168 chargers about two years ago when I forgot to remove the link between my two 6s subpacks on a 12s lipo. However I found a chap in Thailand who is going to have a go at repairing them. A friend is doing volunteering out there in June so postage is low for me too. If they are fixed I will perhaps buy a couple more but if not carry on down the bulk charging route.

apparently I read somewhere its possible to use the Lead acid chargers which the zephyr BMS which would be good but not sure I trust it (the advice).

 

[Alan B]

I am familiar with charging LiFePO4, lots of systems do coulomb counting for state of charge tracking, but I haven't seen chargers use that for determining end of charge. They still look at cell or pack voltages for that.

I use a BC168 on my series connected packs, but just one, and move it from section to section. First I bulk charge to a bit lower voltage, and then use the BC168 to balance top off the pack. A couple times a year is all that's needed to keep a good pack balanced.

The LED supplies are a great improvement. They have built in voltage and current limiting so all the mods and external paraphernalia are not needed. They are sealed and fanless. They are settable with a screwdriver to the voltage and current needed, and they do the job day after day for years without complaint.

 

[whereswally606]

My 900w step up dcdc came today from China. So instead. Of the cheap Meanwell clone and limiter board I stepped up the 36v of the 3 HP 400watt supplies to 54.4v which is 3.4v per cell. Works really well and voltage doesn't appear to drift up like when I have the MW doing the same job. Also it was £16 and will do up to 120v so will still be able to use it for the 26s when I upgrade my controller.

Sent from my ALE-L02 using Tapatalk

 

[Eteck]

I blew up 2 of my bc168 chargers about two years ago when I forgot to remove the link between my two 6s subpacks on a 12s lipo.

So those chargers can't handle it then.

It seems you would still have that hazard even with a fancy relay setup. You might forget to set some charging switch correctly
or a relay could hang and ZAP out go two or three chargers again.

 

[whereswally606]

Etech, the BC168s are up to it.
I plan to learn from my mistake(s). The disconnects for the sub packs on loop keys will be built into a block which will prevent access to the single cell balance taps that the bc168s charge through. This way it will be impossible to plug anything into the balance taps whilst the sub packs are in series. This is an idea I got from Tench's series parallel harness on hit bighit (which is a masterpiece of ebike building). Before I shorted them the bc168s worked flawlessly. It should also act as a bit of a weather guard do to the location of the batteries on the scooter and the wires exiting near the back wheel. Needs to have some security from folks trying to take the loop keys when the bike is not safely in a garage, (not much of the time).

I am wavering at the moment with doing all this though
a) bottom balancing is working very well, just need to be able to coulomb count reliably.
b) I haven't had the bc168s repaired yet or bought the 3rd and 4th I need for 16s and 24s
c) My wife and I are having our first child due October so I've got to start saving / reigning in the spends.

 

[Alan B]

BC168's pause the charging to make accurate measurements, if two of them are both working on the same pack they will be out of sync and can affect each other's measurements. Someone reported this effect, I don't recall who.

A mechanically blocking plug arrangement would be good, something engineered so procedural errors cannot be made.