Conversation started May 11, 2017 at 17:53.
jrh
jrh
May 11, 2017 17:53
When calculating the current requirements for a circuit involving complex devices like a camera's CCD or an ethernet card, is it reasonable to treat it like a resistor (with constant resistance)? It seems counter-intuitive because it seems like a camera's power usage would vary depending on how it was used.
(seems like something like that would be hardly even linear, let alone constant resistance/load -- DC circuit if it matters)
KyranF
KyranF
are you trying to size an appropriate power supply @jrh?
jrh
jrh
yes
Asmyldof
Asmyldof
@jrh When calculating the current requirement you just use specifications in whatever way you have them. When modelling the device you may use a resistor for simple simulations
KyranF
KyranF
the ethernet card and CCD datasheets will say typical and/or max power consumption
jrh
jrh
it sort of came up because the datasheet gave me watts @ 12 V, but we're supplying 24V (which is within the limits of what the device can handle)
It seems a little weird to me that the device literally consumes more power at 24V, when it can run comfortably at 12V
Asmyldof
Asmyldof
May 11, 2017 17:58
@jrh That would depend on the device
KyranF
KyranF
assume there is enough decoupling capacitors for those 'complex' modules and that average current (but use the max possible to ensure you are covered) can just be assumed to be all they need. If an ethernet card says it uses 1A at 12V when at max usage, that is 12W. Would consume 500mA at 24V assuming no huge differences in the power conversion circuitry onboard. Efficiency at 12V might be similiar (or it might not be) at 24V as well, so the total Wattage would be close in any case
Asmyldof
Asmyldof
A computer card almost always has internal regulators that are somewhere around 90% to 95% efficient at their typical supply.
jrh
jrh
hmm... interesting. thanks!
Asmyldof
Asmyldof
My points were going to lead to possibly it being only 80% or at worst 75% at 24V, so if you take a margin it's likely to be fone
But if you are designing for production / quantity you likely want to make some verifications and measurements yourself when powered by a much overdimensioned supply
KyranF
KyranF
Once you have the full "budget" of your devices, add on a 25% margin (or more..) to cover possible shitty/false ratings on the power supply you buy, or strange peak current loads that a beefier supply can handle better than one that is running very close to your requirements. ^^ basically what Asmyldof said
nothing worse than a peak load causing a brownout
Asmyldof
Asmyldof
May 11, 2017 18:02
The last part of what I said is really "get something at least twice what you think you need for your first or first two set-ups", so that you can verify these peaks and actual usage characteristics, knowing the supply won't limit them
jrh
jrh
definitely. I've been on some under-spec'd lines, it wasn't fun
Asmyldof
Asmyldof
But if you only build one for private use, that's of low importance
jrh
jrh
@KyranF If the card consumed 1A at 12V, wouldn't it consume 2A at 24V based on Ohm's law?
that would be a "12 ohm" ethernet card
That's 4x the power from 12V, I'm guessing the excess power at 24V would be dissipated as heat?
KyranF
KyranF
@jrh no, because the device only uses what it needs. it has active conversion circuitry on it. Think of it as a power equation, not just a voltage and current because that is only valid at that point in time
jrh
jrh
So it can't be modeled as a resistor, pretty much?
KyranF
KyranF
May 11, 2017 18:17
if it needs 12W to get its job done, and that happens to be 1A at 12V, then if you use the Power equation ( P = IV) then at 24V, 12 = I * 24, making I= 0.5A
it can, for a given set of parameters
jrh
jrh
oh
KyranF
KyranF
if you keep the voltage constant at 24V, along with all your other devices, then yes they can all be considered static resistors for that situation ONLY.
thinking of them as resistors might be unnecessary anyway, look at it as a sum of power equation
If you find the 'power' (specifically at 24V if you can, otherwise do the conversion like we showed for the different current usage between 12V and 24V ) for each device, then the sum of power consumption gives you a clear requirement for your power supply.
if you need 120W to power your system, then going for a 200W supply might be a good start, for example. A 480W supply would be overkill.
jrh
jrh
(my power supply is spec'd in amps but I'm guessing that would just be P = IV)
KyranF
KyranF
well yeah, if it's a 24V fixed output supply, and it gives 3A output, then you have approx 72W supply
 
Conversation ended May 11, 2017 at 18:20.