What is the actual, real meaning of "Peak" ?

[Dui ni shuo de dui]

Hello everyone,

Usually, when we talk about batteries, controllers and so on, the two ratings "peak current" and "continuous current" rapidly appear. I understand the general principle of what is a peak and what is continuous, but is there an actual standard who scientifically defines what is a peak?

It is obvious to me that one of the most important point would be the time component. So how long is supposed to be a peak exactly? 1 second? 2 minutes? half a nanosecond? Where is the limit between peak and continuous?
It really matters a lot because 30KW for half a second is definitely not the same as 30KW for 5 seconds.


Some manufacturers will sometimes tell you how long those peaks migh be, this essentially applies in datasheets for controllers, I've more rarely seen it on batteries. For instance, one manufacturer will tell you that this controller can handle a peak of current of XXX Amps for 10 seconds, this other controller can handle a peak of YYY Amps for 1 minute, etc. So it seems to me that there is no consistency in what a peak might be.

An other thing I'm wondering is what voltage sag is associated to these current peaks? You can have a peak of 3000 Amps, but if your voltage then sags to 0.1V it is just pointless.

So my question would be, is the "peak current" figure just some kind of marketing stuff manufacturers communicate to unaware customers, or is there an actual, real life standard who defines it clearly. And if yes, what is it? How long should a standard peak be? What % of voltage sag is allowed?

Thanks in advance for anyone having an answer to that

 

[jonescg]

Constant is what you can run it at, all day (or until it runs flat, in the case of a battery) without causing lasting damage.

Peak. I believe, is what it is capable of under ideal conditions, for a period long enough to not damage it irreparably. So for a cell, 2C constant might be what it will deliver without excessive sag and temperature rise (under 45'C or something). 5 C might be what you can get for 30 seconds before the cell begins to heat up past 50'C and have it sag to half of it's resting voltage. You might even get 7C for 10 seconds before these bad things happen. So peak could be either of these ratings...

So yeah, it is all somewhat arbitrary, but think of 'peak' in terms of seconds, and constant in terms of minutes I guess.

 

[Dui ni shuo de dui]

Peak. I believe, is what it is capable of under ideal conditions, for a period long enough to not damage it irreparably. So for a cell, 2C constant might be what it will deliver without excessive sag and temperature rise (under 45'C or something). 5 C might be what you can get for 30 seconds before the cell begins to heat up past 50'C and have it sag to half of it's resting voltage. You might even get 7C for 10 seconds before these bad things happen. So peak could be either of these ratings...

Yeah, that's more or less how I understand it too.

Problem is:
"to not damage it"
This seems like a subjective thing to me, cells wont go from "no damage at all" to "damaged" at a specific defined point. They are always somehow damaged even during normal operation, so there must be some kind of threshold.

" without excessive sag"
What is an excessive sag? Seems very subjective to me too


"for a period long enough "
That's in my opinion one of the main point. what is a "period long enough"?

All of this seem pretty random to me

 

[billvon]

So my question would be, is the "peak current" figure just some kind of marketing stuff manufacturers communicate to unaware customers, or is there an actual, real life standard who defines it clearly. And if yes, what is it? How long should a standard peak be? What % of voltage sag is allowed?

Continuous current is the amount of current you can get all day long (or as long as the battery lasts) without exceeding thermal limits.
Peak current is the amount of current you can get without damage or tripping anything - thermal limits are not considered.

Some manufacturers break them out further than this. For example, Outback inverters call out surge, 5 second and 30 minute power ratings as well as continuous power. "Surge" is generally defined as long enough to cover a motor starting surge, usually about 200-300 milliseconds or so.

 

[Punx0r]

Of all of them, "Continuous" is probably the best defined/most standardised. It's usually >60 mins and is intended to be sufficient to allow the system to reach thermal equilibrium.

 

[999zip999]

Wow yes . You can feel the cost of sag and the effect of load as you are doing on your battery load or source. Sag equals heat. You do feel it you do know it as you ddoin it. Real time. The same as when you're riding you feel if you cut the side wall of your tire and it gets loose or swooshy. It's the same when you break a spoke you did feel something's off. All this you should be able to feel in your ass as your ass is on the seat of your bike . just like in your car when you know your stomping on it too hard and there's not enough gasoline to air and you're going to burn your motor. I really think you should be able to feel this stuff. Or have sense of what's going on. That is time having fun and so much fun things melt and things melt. As we've all melted many things in our lifetime. Headroom only last so long or seconds. Melt. I know when I got my A123 pack all my time was full load up the till I melted 3 motors 2 controllers samenon melted batteries yet . Stop trying ? Stored energy used on demand . Ok.

 

[john61ct]

In general public product literature, Peak spec on loads is just a warning that the source must have lots more capacity than what it's rated at continuous.

For sources, yes it may be for such sort times, could not handle even a second may be a long time.

Great precision is too confusing for the non-tech public.

So IRL talking here, specify both the power source and the load(s) proposed, and likely broad general discussion on recommendations will be more productive than getting into detailed oscilloscope analysis.

 

[Dui ni shuo de dui]

Ok so that seems to confirm what I thought, which is that peak figures are basically not reliable/mostly marketing. They can give a ballpark idea of what might be the limits when they are accompagnied by a time figure (like 200A for 2 sec), but they should not be used for actual power calculations most of the time. Just some kind of indicative thing.
They can't really help you to get a fair comparison between different hardware either...

That kinda sucks, I was hoping to be wrong and that there would be something more scientific.

 

[billvon]

Ok so that seems to confirm what I thought, which is that peak figures are basically not reliable/mostly marketing.

In my experience they are pretty reliable. The thing will deliver the peak power for a short time. It will deliver continuous power forever. Those are useful data points.

They can give a ballpark idea of what might be the limits when they are accompagnied by a time figure (like 200A for 2 sec), but they should not be used for actual power calculations most of the time.

Why not?

When designing something electrical, you need to know peak draw to size breakers, fuses and wiring. You need to know continuous rating for thermal calculations. Having both allows you to make those calculations, which is good.

 

[Chalo]

When I match a battery, a controller, and a motor, I assume that whatever maximum power is possible to feed through the system will in time be demanded for a prolonged period. So I only use a motor and a battery that are both capable of whatever continuous maximum power the controller can carry.

 

[john61ct]

Necessary but not sufficient.

A huge mains aircon unit, I want to run off a given inverter.

Unless I find a trusted person saying they know from experience X will work for sure with Y, maybe needing Z soft-start capacitor

ideally buy from places with good returns policies, or that let you bring in the aircon unit to test off their inverter, or

it will be a crapshoot until I actual get the gear and test it.

 

[Dui ni shuo de dui]

In my experience they are pretty reliable. The thing will deliver the peak power for a short time. It will deliver continuous power forever. Those are useful data points.

Useful if you know what "short time" actually is.
Don't get me wrong, I don't mean that it is totally useless. I just mean that there is nothing scientific about it.

Actually why am I talking about that?
When you get to select a controller or a battery, you'll have to browse the different manufacturers.
Let's just focus on batteries for the example. (figures are random, just to illustrate)
A LG cell might be given by the manufacturer for 3C continuous and 6C peak
A A123 cell might be given at 3C continuous and 10C peak.
Those figures might be true, but there is just no way to actually compare those cells and pick the best one based on that.
Let's say your goal is to power a motorcycle for a 0-60mph in about 5 secs.
You'll need your cell to handle the peak power for arount 4-5 seconds then.

But nothing tells you which of the previous cells will be able to handle it. Maybe the LG cell will be able to sustain peak power for 10 secs and the A123 only for 1 sec.
That's the reason why I was asking this question in the first place, I don't see any way to accurately/objectively compare these cells based on these figures, so that makes hardware choice and comparison a bit difficult. I was wondering if it was just difficult for me due to my ignorance or if it was difficult due to the lack of standard.

When designing something electrical, you need to know peak draw to size breakers, fuses and wiring. You need to know continuous rating for thermal calculations. Having both allows you to make those calculations, which is good.

A fuse, a breaker also have a time constant, you can put a lot of amps through a fuse, way more than it is rated for, if the time period is short enough. So time matters, always. That's why I'm surprised I cannot find any actual definition of what is this "short time".

Anyway, I think I've got my answer by now, so thanks everyone for your inputs.

 

[billvon]

Useful if you know what "short time" actually is.

Useful anyway.

Device is rated at 10 amps continuous, 25 amp surge. So you use a 30 amp breaker, wire for 30 amp faults and make all your loss calculations at 10 amps.

Lets just focus on batteries for the example. (figures are random, just to illustrate)
A LG cell might be given by the manufacturer for 3C continuous and 6C peak
A A123 cell might be given at 3C continuous and 10C peak.
Those figures might be true, but there is just no way to actually compare those cells and pick the best one based on that.

I agree. And let's say the packs are rated at 3C continuous and 6C peak for 10 seconds. There is STILL no way to compare the two. What is the sag at 6C? The ESR of the pack? Is that at 25 degrees C or 40?

If your point is that it's hard to compare batteries - I definitely agree with that.