Questions regarding battery type and what voltage moving forward

[georgia088]

Relays. Possibly. However you MUST research them carefully. The batteries are more than capable of 100A+ under load. You would ideally want a dual pole "break before make" relay. At 100A that's more likely going to be a "contactor" which itself is adding complexity.

They have some pretty heavy ssr dc-dc relays. I was thinking maybe one of those, but I haven’t checked much in to it. As for the outputs. A lot of what I run is 5v usb plug. The charge controllers have those built in regardless of voltage i assume so I could use those for that. I think. The 24poe for the u inquiry may be capable of running on 12v I’m not sure, but I do have 12v lights that I don’t think would be happy at 24v.

thabks!

[venquessa]

I was thinking about bi directional chargers set to auto engage, one or other. Which ever battery is high, charges the other. It's, I think just too complex, would need to DIY and if you get it wrong and run both together you just flatten both batteries slow running current around in a loop all day.

Ignore this. It will just oscillate back and forth until the batteries are flat.

[venquessa]

They have some pretty heavy ssr dc-dc relays. I was thinking maybe one of those, but I haven’t checked much in to it. As for the outputs. A lot of what I run is 5v usb plug. The charge controllers have those built in regardless of voltage i assume so I could use those for that. I think. The 24poe for the u inquiry may be capable of running on 12v I’m not sure, but I do have 12v lights that I don’t think would be happy at 24v.

thabks!

SSRs with inverse polarity on there "gate" input. Such that turning one on, turns the other off. Might work. One risk point is that both connect at the same time. I'm a software engineer and personally, I'd want that done in hardware as a "can't happen lock out" type thing.

USB is my primary load too!

I installed one of these on the end of the 12V battery incoming power. Im upgrading to 24V and I'm not at all worried. Granted, it only about 80% efficient overall. I use the regulated 12V from it to supply power USB hubs which contains multiple high power charge ports including QC3.0 19V ports.

https://www.aliexpress.com/item/32840324731.html

In fact, setting this to 12V, it will output 12V, regardless if you give it 5V or 48V, regardless if you chop and change that voltage.

The same device, just it's big brother, is what I use to charge LFP packs at 20Amps. It however runs off a 48V 10A DC Brick.

EDIT: The internal USB charge ports on ... well... everything these days... will be 5V 2 Amp. However if htey actually do the full 2 Amps and don't overheat you are doing well.

[georgia088]

SSRs with inverse polarity on there "gate" input. Such that turning one on, turns the other off. Might work. One risk point is that both connect at the same time. I'm a software engineer and personally, I'd want that done in hardware as a "can't happen lock out" type thing.

I haven't worked it out yet in my mind, but I was thinking have a relay for the output side of both the 12 and 24v sets of batteries. Although not ideal, I could have one turn off and have a delay before the other turns on. This would require all the networking to have to reset(really not a big deal i don't think), but I'm not sure it would be possible to do it without this happening

[georgia088]

SSRs with inverse polarity on there "gate" input. Such that turning one on, turns the other off. Might work. One risk point is that both connect at the same time. I'm a software engineer and personally, I'd want that done in hardware as a "can't happen lock out" type thing.
.

I've quickly drawn out another sketch of maybe a possibility? Both the 24v battery and 12v will have supply lines to a "panel box" that will feed all the devices. Inside this panel, anything that needs to change voltage from 12/24 to the needed voltage to run will need to be done. There will be two relays. One relay on the + going from the 12v battery to the panel and one on the + going from the 24v to the panel. A microcontroller will sense the voltage of both. If one becomes too low, it will turn that relay OFF. Delay and Turn the Other relay ON. This would prevent having them both powered at the same time. It atleast makes some sense in my mind. Am I losing it?

Thanks!

[venquessa]

The weak link is the software. As a rules of thumb, if it can happen it will and software fails.

Both relays can become active and thus "will" at some point, by malice, ignorance, accident or negligence.

A "normally closed" relay on one battery with a "normally open" relay on the other will provide a reasonable amount of protection for that.

If you put 12-24V on the coil of a NC relay it opens. However the NO relays closes. One switch input, dual pole output. MCU, one GPIO, 2 switches inverted.

EDIT: Note an MCU pin cannot drive a relay coil directly, it might drive an SSR, "datasheet details, devil etc."

[venquessa]

Single DPDT relay would work. It has 2 contacts which switch together. Disconnects 2 and connects 2. It's "make then break." It's just a component you need to pick carefully and study datasheets to pick the correct one.

[georgia088]

Single DPDT relay would work. It has 2 contacts which switch together. Disconnects 2 and connects 2. It's "make then break." It's just a component you need to pick carefully and study datasheets to pick the correct one.

So you are thinking to use the DPDT for both the + and - of the 12v supply and 24v supply? Could they share a common ground and use a SPDT relay? I haven't found any DPDT relays that will handle the amperage needed that don't cost a fortune.

The weak link is the software.

Maybe I can find a software engineer to look over my code if I go this route....

Thanks!

[venquessa]

I'd imagine switching the positive is fine, as long as you have another way to isolate the batteries for maintenance etc.

Using MCBs or physical fuses sized just above your maximum expected load should help add reassurance such that if the batteries ever become connected in parallel and the 24V LA surges the 12V lithium, those breakers will pop. Similar for short circuit.

[venquessa]

On the software. It's not that it would be hard to write. Software lock outs aren't hard to write.

The problem is the risk. As it involves high amperage batteries which is a fire risk, there is real risk to property or life. I professionally would not want to accept that liability and would insist on hardware safeties.

[Ampster]

.........

The problem is the risk. As it involves high amperage batteries which is a fire risk, there is real risk to property or life. I professionally would not want to accept that liability and would insist on hardware safeties.

I think it is good that you are concerned about risk. How are you assessing the risk of LFP or Pb batteries? At 12 or 24 volts what is the risk to life?

[georgia088]

On the software. It's not that it would be hard to write. Software lock outs aren't hard to write.

The problem is the risk. As it involves high amperage batteries which is a fire risk, there is real risk to property or life. I professionally would not want to accept that liability and would insist on hardware safeties.

Yea, I will definitely add fuses on the positive side of both battery packs. Would they blow if for some reason something went wrong and the 24v started feeding in to the 12v LFP?

thanks!

[venquessa]

I think it is good that you are concerned about risk. How are you assessing the risk of LFP or Pb batteries? At 12 or 24 volts what is the risk to life?

Fire.

LFP 12V is 4 cells in series. Fully charged is around 14.40V, absolutely maximum permitted is 4.20V per cell.
LA 24V fully charged is about 28.8V or more likely rest at 25V.

It's over 6 volts per cell. Supplied by a big fat current source like a 24V LA marine battery

It might not produce fire immediately, you might get lucky and the cells bulge, fail and vent. However, they don't always contain the flames and hot gases which can set fire to near by combustibles.

It's not something I would like to play with.... well, actually, it IS something I'd like to play with, but not with MY batteries and outside in a safe place with somewhere to hide while you watch the show.

On fusing. There is a whole post on where the fuses should go in DC systems. As it involves the dreaded topic of grounding let's leave it

[venquessa]

That said. If you had a BMS on the LFPs (as you really should), it is very likely the BMS disconnects the pack - quick enough to save the LFPs. BMSs have a maximum breaking current, this is probably one reason to not skimp on that value. Even if you pack normally produces < 50Amp. Having a BMS cable of breaking the FULL short circuit current of the pack in question is essential.

[georgia088]

Assuming I get a bms for lifepo4, and fuse correctly, now I’ve got to come up with how to tie outputs together from 12v source and 24v source to provide 5v and 12v outputs that will only be supplied by one or the other (12v or 24v).

id like to make this as concise as possible. I’ve considered creating a PCB to handle the mcu and the dc to dc converting input/output. The relays would only allow either 12 or 24v to the input side of PCB. Once the 12v OR 24v is present on the input side, I would need to convert it to the desired output voltages. For 24v this would mean down to 12v and 5v. And for 12v this would mean down to 5v.

Could I use something like this
5pcs JacobsParts MP1584 3A Adjustable DC Step Down Converter Buck Module 3.3V 5V 9V 12V 24V https://a.co/d/iSW15Bw

Assuming it provides half the amperage it claims, each of these could handle approximately 3 usb ports (.5-1A per usb port). I also am assuming I could wire these in parallel to provide 6-9 usb ports for output devices (each dc/dc adapter handling 3 usb ports?

The idea being that the outputs are tied together from 12 and 24v sources, but only one of them one be “on” at a time (either 12v or 24v)
Any of this sound feasible?

Thanks!