I want to make my own BMS

Factory400 said:

Interrupting the power switch is exactly what I was referring to earlier.
On many modern devices, however, the power switch is just supplying a little voltage to the input of a microcontroller that in turn runs some code that shots down the inverter. Most of the time that is good - but I would not rely on it to save my batteries from destruction. Way too many ways for that to fail IMO.

Click to expand...

I don't remember the last time I had a power switch fail to work on a piece of equipment .... There are always possible points of failure, but I try to figure out what is most reliable and also try to decide what mode I want to the fail mode.
I don't know the circuit on my inverter, but if the power control internally failed I think it would most likely fail to turn on ... not fail to turn off .... Otherwise, some EE guy didn't do a proper job.

Bob B said:

I don't remember the last time I had a power switch fail to work on a piece of equipment .... There are always possible points of failure, but I try to figure out what is most reliable and also try to decide what mode I want to the fail mode.
I don't know the circuit on my inverter, but if the power control internally failed I think it would most likely fail to turn on ... not fail to turn off .... Otherwise, some EE guy didn't do a proper job.

Click to expand...

When the consequences are severe - I trust no one.
That is why I design in fail safes when a single point of failure can destroy my battery pack while I am in the middle of nowhere.

I feel like that is a critical role of a BMS - shut everything down if things go off the rails. No questions asked, just shut it down.

Factory400 said:

When the consequences are severe - I trust no one.
That is why I design in fail safes when a single point of failure can destroy my battery pack while I am in the middle of nowhere.

I feel like that is a critical role of a BMS - shut everything down if things go off the rails. No questions asked, just shut it down.

Click to expand...

But .... you want that design to be such that other devices aren't damaged during that process. Shutting down the input power to a device would be the a good way to do that the way I am seeing it.

Bob B said:

In my mind, I need to differentiate between an actual communications protocol being used for control ... and what I have always called hard wired control.
I think you guys are splitting some kind of hairs to decide that 12VDC on / off is a communications signal. To me, that is "hard wired" and the most bullet proof type of control that can be achieved.

Click to expand...

Yes, it's probably the most bullet proof one but it's only a control signal relying on something else to cut the high power path.

Bob B said:

For my inverter, I am interrupting the power switch. That seems to me to be the most reliable way to control it that is possible other than disconnecting the AC power input.

Click to expand...

Yes, but you added the relay yourself, as said earlier most inputs on SCC, inverters, etc. are directly managed by software.

And even your solution actually doesn't interrupt the high power path. I wonder what that switch actually controls...

The problem isn't how complex is the com, but the fact you need to rely on a bunch of other things to cut the power instead of cutting the power yourself.

Bob B said:

I think you EE guys are making clarity impossible ...LOL

Click to expand...

It's quite the opposite actually, by using random definitions for specific terms you usually end up with big mis-understandings.

Bob B said:

I don't know the circuit on my inverter, but if the power control internally failed I think it would most likely fail to turn on ... not fail to turn off .... Otherwise, some EE guy didn't do a proper job.

Click to expand...

That's only a guess, and I'm personally not ready to make that bet when safety is at play.

Factory400 said:

When the consequences are severe - I trust no one.
That is why I design in fail safes when a single point of failure can destroy my battery pack while I am in the middle of nowhere.

I feel like that is a critical role of a BMS - shut everything down if things go off the rails. No questions asked, just shut it down.

Click to expand...

Yep, exactly

Bob B said:

But .... you want that design to be such that other devices aren't damaged during that process. Shutting down the input power to a device would be the a good way to do that the way I am seeing it.

Click to expand...

Ideally, yes, but when safety is the priority we prefer to let something potentially destroy itself to keep other more important things safe. A good example is a MRI quench.

Also lots of people reported MPPT SCCs are actually very resilient regarding interrupting the battery side before the PV side.

And for me, if interrupting the power to a SCC or an inverter kills it, then someone didn't do its job correctly...

Bob B said:

But .... you want that design to be such that other devices aren't damaged during that process. Shutting down the input power to a device would be the a good way to do that the way I am seeing it.

Click to expand...

Go back a few posts.....

Factory400 said:

I think there is no question that the ideal operation is to communicate with the various devices in the system so that they can gracefully turn themselves on/off or adjust some parameter on their own terms.

In the event that communication fails or the device does not respond as desired - the BMS should have the ability to disconnect to avoid permanent damage. In a perfect world, the BMS would never need to do a hard disconnect - it is the last line of defense to protect the cells.

Click to expand...

The hard disconnect is: "the last line of defense to protect the cells"
Ensuring the survival of your battery pack is worth the unlikely possibility damage from a hard disconnect. Ask your devices to shut down - if they do not comply too bad the electricity will be cutoff.

Why is this concept seeming so complicated. It is a simple backup protocol when the system enters dangerous behavior.

I don't think we are going to come to any sort of agreement ... and, the terminology I am using ARE standard terminology in the control industry. What we are talking about here is the BMS emergency CONTOL of charge and load.

In the case of charge controllers, I am advocating a hard disconnect of the input power to those controllers .... not the output power. Both are guaranteed to shut down the charge, but will not defy manufacturers safety warnings.

Shutting off the power switch to an inverter will remove 99.99% of it's current draw and there won't be any concern about doing a pre-charge to get things back into operation. I see the likelihood of the inverter power switch failing to shut off the inverter at the same time I have an under voltage condition on my pack to be something so remote that it isn't worth considering.

What about the MOSFETs shorting out in the inverter/SCC for example? because no control line will stop that and the current might stay under the fuse limit (or at least be low enough that the fuse will take minutes to react).

BiduleOhm said:

What about the MOSFETs shorting out in the inverter/SCC for example? because no control line will stop that and the current might stay under the fuse limit (or at least be low enough that the fuse will take minutes to react).

Click to expand...

Is a short on those MOSFETs going to result in a dead battery even if the solar input is removed? If so, I will rely on me operating a manual disconnect.

Depends on the topology of the converter but yes, it can, it's basically 50/50.

Problem in that example isn't the battery going dead (well, it is, but it's secondary), but things like fire for example.

Well ... in that case, the last line of defense will be me and a fire extinguisher ..... or insurance ....LOL

It seems to me defining "communication" as one wire control is so broad that it's a useless discussion point. By that definition, ever gate drive is communication. Every logic level change in a trace is communication. Some just enjoy to argue.

Cal said:

By that definition, ever gate drive is communication. Every logic level change in a trace is communication.

Click to expand...

Yes, true. Thank you for pointing that out.

Kinda sounds like they are tending more toward the information exchange .... Not just a binary control.

Bob B said:

Kinda sounds like they are tending more toward the information exchange .... Not just a binary control.

Click to expand...

"transmitted"
"conveyed"

When you use the word "control" that implies that you have transmitted or conveyed some sort of signal from one thing to another in order to control it. You don't send a 'hardwire', you transmit information as a current, voltage, smoke, hand wave, sound waves, etc.

The wires don't do anything other than provide a path for the actual communication which is transmitted via some electrical signal. Who cares if it is a dead simple voltage/current activating a relay coil or some 10G Ethernet. The concept is the same.

The only reason this is important is because a BMS can communicate in a variety of different ways from simple to complex. Trying to differentiate injects the confusion. In a planning discussion there is no conceptual difference between a relay and an Ethernet packet that ultimately activates a function. It is all the same and should be treated as such until the actual circuits are being designed and built.

Factory400 said:

"transmitted"
"conveyed"

When you use the word "control" that implies that you have transmitted or conveyed some sort of signal from one thing to another in order to control it. You don't send a 'hardwire', you transmit information as a current, voltage, smoke, hand wave, sound waves, etc.

The wires don't do anything other than provide a path for the actual communication which is transmitted via some electrical signal. Who cares if it is a dead simple voltage/current activating a relay coil or some 10G Ethernet. The concept is the same.

The only reason this is important is because a BMS can communicate in a variety of different ways from simple to complex. Trying to differentiate injects the confusion. In a planning discussion there is no conceptual difference between a relay and an Ethernet packet that ultimately activates a function. It is all the same and should be treated as such until the actual circuits are being designed and built.

Click to expand...

I totally disagree ..... In the control industry that is not the definitions we use .... but I give up. ?️?️

You've worn me down... and that isn't easy to do ... LOL

Bob B said:

You've worn me down... and that isn't easy to do ... LOL

Click to expand...

Imagine what my wife has to put up with, lol.

Movin' on......

Back to pure BMS chat: The last major discussion point was the merits of a master disconnect as part of the BMS, especially if the system is capable of triggering an indirect shutdown.

My view is that it is not a good idea to trust external devices to behave a certain way when a simple malfunction can destroy the heart of the system - the cells. Sure it is not likely, but then again it is not all that likely I get into a car crash yet seat belts and air bags are mandated by law because the consequences are so severe.

Most of my professional experiences have convinced me to have backup systems for anything that is important to me.
Especially when they are cheap - MOSFETs are cheap compared to tanking a substantial LiFePO4 pack in the middle of nowhere.

At the end of the day - it is subjective, of course. I personally put a huge priority on redundancy. In my RV rig, there is a lot of redundancy on things that can cause big problems.

Bob B said:

I totally disagree ..... In the control industry that is not the definitions we use .... but I give up. ?️?️

You've worn me down... and that isn't easy to do ... LOL

Click to expand...

So .... I was just hungry.

I won't be continuing the discussion beyond this post .... but here is the dictionary definition of hardwire. The number 1 interpretation is how I was using the term. If I go to any person who works with a BMS (Building Management System ...LOL) and say something is hardwired, they will know exactly what I am talking about.

Factory400 said:

Back to pure BMS chat: The last major discussion point was the merits of a master disconnect as part of the BMS, especially if the system is capable of triggering an indirect shutdown.

Click to expand...

Factory400 said:

My view is that it is not a good idea to trust external devices to behave a certain way when a simple malfunction can destroy the heart of the system - the cells.

Click to expand...

This discussion is rather goofy. "external devices" is a bogus concept. ICs soldered onto a PCB were "external", but once soldered they are not? Why? Many different companies are involved in supplying the components that are brought together. I see no reason to trust the IC maker of some FETs but not a producer of a completed module. It does not make much sense to sell N different sized BMSs for N different amp ratings just for the goal of saying it is "internal". Just like inverters and chargers are sold separately and sized for the Amps that will be handled, it makes sense to sell a power cut off switch independent of the BMS. This is especially true given the mechanical demands a high amp connector requires.

@Pidjey I made my own BMS, and like the Electrodacus no power flows through it. I used the ESP32 for the microcontroller. I have developed a bunch of software and am still working on it. The software can easily exceed the efforts to make the hardware. I'm willing to collaborate if you like.

johntaves said:

This discussion is rather goofy. "external devices" is a bogus concept. ICs soldered onto a PCB were "external", but once soldered they are not? Why? Many different companies are involved in supplying the components that are brought together. I see no reason to trust the IC maker of some FETs but not a producer of a completed module. It does not make much sense to sell N different sized BMSs for N different amp ratings just for the goal of saying it is "internal". Just like inverters and chargers are sold separately and sized for the Amps that will be handled, it makes sense to sell a power cut off switch independent of the BMS. This is especially true given the mechanical demands a high amp connector requires.

Click to expand...

Internal devices are ones that I have vetted and designed as a coherent system that I understand in great detail.
External systems are totally and completely out of my control or understanding.

This is pretty normal.

Factory400 said:

Internal devices are ones that I have vetted and designed as a coherent system that I understand in great detail.
External systems are totally and completely out of my control or understanding.

Click to expand...

I don't see how this definition affects anything related to this thread.