Difference between "Current Source", limiting, OC protection

I don't understand the EE-level abstract physics, and in fact for this discussion, please emphasize actual products out in the marketplace, use the theory only as needed to explain their behaviour as much as possible.

Context is a large low-SoC LI bank that wants to pull huge amps, fed by a charge source with limited power.

There are plenty of charge sources that set a maximum cap on output current, as well as your CV cap setpoint. Some even better, let you adjust that maximum cap on output current on a per-session basis.

These I've been classifying as "current limiting", and "adjustable current" respectively.

But they are not strictly speaking a "current source" are they, in the ideal theoretical POV, as opposed to a "voltage source"?

This has come up as I've discovered the law "you may not connect two current sources in series", looking at @doctorbass brilliant charging array idea using looped-back DC-DC converters.

Flatpack2, the big Huawei supply etc, apparently can and do get connected in series, but also feature adjustable current output.

Then you have power supplies like Mean Well, some models have explicit current limiting, others just OCP, of type "current limiting" as opposed to "hiccup" or "latching." These usually kick in at a level far above their continuous current output rating.

Would that be usable in this LI charging scenario? I would think not, if the feature results in the voltage dropping below what's required to get charging current into the battery.

Any and all guidance on these topics would be greatly appreciated, including links to 101-level learning resources.

Consider that basically all lithium batteries want to draw more current than you are supplying during the first part of the charge (hence, cc stage) and reevaluate. You want a cc/cv charger. Any of the known suspects will work with correct settings.

john61ct said:

This has come up as I've discovered the law "you may not connect two current sources in series",

Click to expand...

That is because in general electronics theory terms, a current source "absolutely" determines the current in a series circuit. Two of them in series doesn't work because if one is different than the other, the circuit can't exist because current in a simple series circuit is always the same at all points in the circuit.


Also, in reality, most things you'd first consider to be a current source are actually just a self-adjusting voltage sourse that has a set current maximum, so at that maximum current the voltage adjusts itself to produce that current based on the resistance of the circuit. IF the resistance rises above the point at which maximum current can flow, then the voltage rises to whatever the maximum the device is set to output. That's not precisely CC/CV, but it is a first approximation of the process.


So you can series these because tehy don't operate as a pure current source.


FWIW, there have been PSUs that don't work very well when connected in series, because their designs are more like a current source, so they "fight" each other trying to regulate the current, and never stabilize. Sometimes they swing wildly enough to damage themselves or each other. Don't see this very often; usually it's in the cheapest stuff, or in purpose-made stuff for a very specific system, whose designers didn't consider or intend anyone would ever want to use them outside that single specific application, much less in ways they weren't used in taht system.

So in practice are these two scenarios equally acceptable?

A. bunch of isolated DC-DC buck converters (voltage sources all) in series with my main rectifier, say a Flatpack2 in order to boost voltage.

The rectifier is controlled by CAN to keep current output well below the converters' max amps rating


B. main rectifier is a big Mean Well, voltage source only, say 48V. DC-DC buck converters added as above, also voltage sources, each adding 12V.

Then the last, lowest voltage buck converter (say 5V) includes an adjustable current limit feature, its maximum current output is lower than all the other source's max amps rating.

And it's adjusted well below its max rating.

Just for giggles, assume all the above devices in both scenarios have solid OCP features, whether latch shutting down, hiccup style or whatever.

I am no electronics expert, but I collect information from many sources that seem to be competent. EV cars have a large battery pack, and after that large pack has been depleted, many potential customers express some concern over the amount of time needed to charge up the pack.

To address these concerns, many products make sure to emphasize that "in just 30 minutes" that perhaps 100 miles' worth of range can be added by a smart charger.

That caught my eye. Apparently, when a pack is empty (*3.0V per cell) it can absorb a higher rate of charge than when it nears a full state of charge state.

So then, it appears to me that the fastest possible charging would incorporate a sliding amount of amps, that is tied to the heat of the pack. And...if that is true, then pack cooling can increase the rate of the sliding charge.

I thought that issue deserved its own thread, addressed here https://endless-sphere.com/forums/viewtopic.php?t=101288

Assume my two alternative scenarios are at the same overall voltage and amps output, the distinction between them is the relative safety of how the current gets limited when using the doctorbass "charging array" idea when a single huge power supply isn't affordable.

In order to do the current limiting, the unit that does it has to have enough voltage range to be able to drop *it's* voltage so that it plus all the other seriesed voltage sources total up to only as much as is needed to cause the current flow specified, thru the resistance of the load it's applied to.

So if you had a battery that when discharged is at say, 42v, and a charging setup meant to bring that to 52v, the unit that does the current limiting has to be able to change by as much as 10v.

Aha, that makes good sense.

So the relative power or current output does not matter - much? at all?

So, an example

A 154V nominal bank, charging goal somewhere between 166V and 172.8V, or let's say going to top balance at the latter.

Since the bank is never allowed to go below 144V, the maximum delta is 30V.

Wow, that does present a challenge, other than non-portable, lab-style PSUs that's a wider range than I've seen.

Hmm just checked and Grin's Satiator is well isolated and goes 20-60V, but just that piece is $300+ and only contributes 360W max.

Other suggestions anyone?

Meantime, strategy tweak idea.

Make the array easily reconfigurable, and bring the bank up in staged increments of under say, 12V at a time.

So first from 144 to 154, then kick in another 12V DC-DC voltage source and go from 154 to 164V, then from 164 to the final target 172.8V

That would allow the use of say Flatpack2, the main rectifiers, for doing the CL, with a 44 - 57Vdc range, delta over 12V available.

Good?

Every device in series has to support the maximum current the system will see.

Power rating doesn't "matter" because that's determined by current * voltage, so if the unit can supply the current you need at the voltage you need, it automatically has enough power. (assuming the device is rated for both it's max voltage *and* it's max current at the same time, continously).

Using a "staged" setup would be awfully annoying to do every time you charge your battery.

It also adds the risk of doing something wrong at every stage at every charge of the battery--the more steps there are to doing something, the more likely it is that eventually a mistake will be made. Consequences of the mistake range from annoyance (charge not completeing because something was left disconnected) to fire and destruction of system and battery and building and possible loss of life, if something was hooked backwards or somehow across a battery connection, etc.

Unless there's simply physically or financially no way to do what you want any other way, I wouldn't want to do that.

amberwolf said:

Every device in series has to support the maximum current the system will see.

Click to expand...

Yes I may not have stated clearly enough, the CL unit's maximum current output is lower than all the other source's max amps rating.

And it's adjusted well below its max rating.

What I meant by "current doesn't matter?" is, whether or not the CL unit contributes 20% of the total current, or 80%, does not bear on its effectiveness or safety concerns?

The wide-enough voltage range is what's important.

And now I just realized, it isn't the unit's **user-adjustability** V range that counts, but how low its internal components can drop the voltage in handling load over-current.

I've only come across that spec on very high quality supplies, but seem to recall its usually very low.

amberwolf said:

Using a "staged" setup would be awfully annoying to do every time you charge your battery.
…
Unless there's simply physically or financially no way to do what you want any other way, I wouldn't want to do that.

Click to expand...

Completely agree, more a thought experiment to test my limited understanding of the principles.

john61ct said:

What I meant by "current doesn't matter?" is, whether or not the CL unit contributes 20% of the total current, or 80%, does not bear on its effectiveness or safety concerns?

Click to expand...

Every unit in series contributes 100% of the current.

Each one only contributes some percentage of the *voltage*, however.

Hmm, well I guess that's why I used Power before.

I'm thinking "pass through" must be involved?

Scenario: main isolated output AC-DC rectifier is capable of producing say 10kW at 48V call it 180A.

put in series with the output of one 48-48V isolated DC-DC converter, and a bunch of smaller ones say 5V & 24V to get to 167V

One of those components - the question of this thread being "which one"?

keeps the whole-array current down to say, 25A.

So from a lay POV, really only the main AC-DC rectifier is "contributing" current (/ power) to the array.

All the other devices are just converting voltages, have to **tolerate** the "passing through" of the overall 25A?

And since we want the main AC-DC rectifier to output its @48V current at ~75A, it needs to be another device, a DC-DC converter, that does the current limiting, is that correct? Can't be that main rectifier unit doing it?

Really appreciate your sticking with me helping me grok these concepts.

Vicor AppNote
Constant Current Control for DC-DC Converters
http://www.vicorpower.com/documents/application_notes/an_ConstantCurrent.pdf

OK, having stalled on the other aspects for now

I'd like to clarify one small issue, from my OP:

The term "current limiting"

Seems obvious, if a charger, converter or rectifier/PSU features "adjustable current" then that is a subset of current limiting, right?

But that does not imply strictly speaking they are a "current source" does it, as opposed to a "voltage source"? Just that they can operate in either "mode"?

And it seems that "current limiting", in the sense of over current **protection**, too often means voltage drops so low charging is in effect stopped, so SoC will never rise to the point that CV stage ever kicks in.

When discussing / looking for cheaper power supplies and DC-DC converters, people start saying things like "the only way to control current is to drop voltage".

Which may be true in theoretical EE circuit design terms, but IRL, there are plenty of actual "shore power chargers", fancy alternator VRs solar controllers etc that have this functionality.

What is the EE-compliant terminology when seeking charge sources that both limit current to a maximum (adjustable or just pre-set) **and** keep voltage as high as possible well over battery resting voltage, striving for the Absorb/CV stage voltage setpoint **and also** delivering that intended maximum current?

john61ct said:

What is the EE-compliant terminology when seeking charge sources that both limit current to a maximum (adjustable or just pre-set) **and** keep voltage as high as possible well over battery resting voltage, striving for the Absorb/CV stage voltage setpoint **and also** delivering that intended maximum current?

Click to expand...

Usually just referred to as CC-CV power supply.
Below the voltage set point, it operated constant current. Once the pack reaches the voltage set point, current will drop below the CC set point and voltage will stay at the CV set point. Most battery chargers are some variation of this.

fechter said:

john61ct said:

What is the EE-compliant terminology when seeking charge sources that both limit current to a maximum (adjustable or just pre-set) **and** keep voltage as high as possible well over battery resting voltage, striving for the Absorb/CV stage voltage setpoint **and also** delivering that intended maximum current?

Click to expand...

Usually just referred to as CC-CV power supply.
Below the voltage set point, it operated constant current. Once the pack reaches the voltage set point, current will drop below the CC set point and voltage will stay at the CV set point. Most battery chargers are some variation of this.

Click to expand...

Completely separate topic.

The CC-CV transition point is not set by the regulator, happens with a plain PSU, SoC at transition determined by

(besides the V setpoint, which is all most "chargers" are regulating)

C-rate, chemistry, acceptance, resistance.
_______

Examples:

Context is an LFP bank capable of pulling 4-5C rates from the source. I do not want it fed faster than 0.3C

Say a tiny battery, 12V 20Ah and a large 3kW source. The source won't have OCP issues, but it needs to have (ideally) **current adjustment** so I can set a 7A limit for that session.

Or at least it has a rougher de-rating feature, so the max current output can be lowered to 80% or 30% as needed.

With large enough setups, such derating may be required to work with low-capacity upstream circuits, say an 800W genset rather than an EU2200.

With a larger bank, say 800Ah, it may be required to prevent the OCP feature of the rectifier / PSU from getting triggered.

Ordinary chargers designed for (low CAR) lead may not even have decent OCP.

So **that's** what I'm talking about.

_______
And it seems the two rough terms "current limiting", and "adjustable current"

are as good as it gets? To distinguish from power sources

chargers
rectifiers
power supplies
DC-DC converters

That may just have (good / bad / indifferent) OCP.

_______.
From my OP question:
But they are not strictly speaking a "current source" are they, in the ideal theoretical POV, as opposed to a "voltage source"?

A current source will supply or attempt to supply a constant current to the load regardless of voltage. Of course there are limitations in the "compliance" with practical circuits.

Things get much more foggy when comparing marketing terms. A simple fuse can be considered OCP. Some supplies will shut down if the OCP limit is reached. Some will limit the current, but not survive long at that limit. For battery charging you need current limiting that allows both the charger and battery to stay below the level where damage or overheating happens.

fechter said:

A current source will supply or attempt to supply a constant current to the load regardless of voltage. Of course there are limitations in the "compliance" with practical circuits.

Things get much more foggy when comparing marketing terms. A simple fuse can be considered OCP. Some supplies will shut down if the OCP limit is reached. Some will limit the current, but not survive long at that limit.

Click to expand...

Yes, the functionality I'm talking about really has little to do with OCP at all, except I guess that supplies that let the user adjust the max current rate actually output, against a load that's "pulling" 100x more, is likely to already have good OCP as well.

Mean Well HLG-xxx-xxA are a good example

The largest is only 600W, but they can apparently be stacked in parallel or even series'd no problem. Voltage adjustable 50-57V (I like 56V for 16S LFP)

Output current can be adjusted between 5.6-11.2A, so at one unit per 30-40Ah of bank can be charged using these without fear of damaging the bank, nor the PSUs.

I really doubt these qualify as "a constant current source" in the theoretical sense, since you can't have two such devices connected in series, right?


> For battery charging you need current limiting that allows both the charger and battery to stay below the level where damage or overheating happens

No, I need current limiting that will maintain "effective charging", as high as possible a voltage while "striving" for the CV setpoint, and continuing to cap current while the voltage is also being capped there for a decent Absorb cycle.

Many apparently smart people keep telling me that's just not possible, I guess because of theoretical constraints in abstract circuit design where things are perfect?

But in fact I see fancy charging gear doing so all the time, alternator VRs, shore power chargers, DC-DC chargers.

So I'm using "adjustable current output" as the label for now, also "current-limiting" when it's not just OCP related.

Dont overthink it.
Just try and understand the different suppies.
A CC only supply is usually used in led drivers where the voltage does not matter and only current matters. Hook this up to a battery and it will keep pumping in power/amps until the battery explodes due to massive overcharging above the rated voltage.

A CV supply is like a normal supply like you find i a computer or most hardware. They supply a fixed voltage and once you exeed the rated amps it will simply shut down.

A CC/CV does both and is what you want. It is set for a specific voltage and a certain amount of amps and it will just cap out the amps until it reaches the set voltage and then it naturally goes into CV mode. This is nothing special and happens naturally with charging batteries.

A useful PDF from Texas Instruments

Know Your Limits
http://www.ti.com/lit/an/snva736/snva736.pdf

Also a good video from Great Scott
https://youtu.be/8uoo5pAeWZI

Interesting ICs
https://www.maximintegrated.com/en/products/power/protection-control/protection-ics/MAX20313.html

Appnote from Maxim
https://www.maximintegrated.com/en/app-notes/index.mvp/id/3869