Current limiting - voltage drop

[rcheshire]

I would like to explicitly limit current to a few components while keeping voltage within their operating range.

The closest I can get is with a high power mosfet. Supply power 12v 12.5a output 11v ~3a.
Is this a good way of doing this or can I expect lots of heat from the mosfet.

Is their a better way without dropping voltage too low.

[bojan]

Depends on current and voltage drop.

P=U*I

where
U = voltage drop on control element (MOSFET or resistor)
I = current.

In your case, the dissipation will be ~6W, quite a lot - you definitely need heatsink to prevent MOSFET thermal destruction.

However, even simple resistor (capable of required dissipation, say 10W) will do, no need for MOSFET. It will get quite hot, though.


But, I am not quite clear what you really want to do here. Is it just a general question or you have some application in mind?
If you want regulated supply voltage, while the current is limited in case of short circuit, then you should use some of dedicated voltage regulator IC's (7812, for example) with built in current limiting.

The better way (lower dissipation, even when regulator is shorted at output) is to use SMPS IC's (http://www.maximintegrated.com/products/power/switching-regulators/) but for this approach, you have to have some experience with electronics... and the circuit will be more complicated.

http://www.jaycar.com.au/images_uploaded/LM2576.PDF
http://www.dimensionengineering.com/info/switching-regulators
http://www.analog.com/static/imported-files/product_highlights/177228276DCtoDC_Regulator_E.pdf

[rcheshire]

Thanks Bojan for all the info. Very much appreciated.

The application is specific, in that I want to supply a camera, among other things, from a bus with regulated 12v, 3A maximum. Total system load, if everything were operating at maximum is ~ 7A. The SMPS is rated at 12v 12.5A.

The whole idea is to have one power source and distribute all power needs from a bus, instead of my current arrangement of several plugin adapters. The bus will be mounted on the dovetail - DEC axis.

The attached image is the basic idea. The loads are represented by resistors limiting to max current for the service, camera etc. That is, the 4, 12, 12, 12 and 17.9R resistors. There's also a variable brightness LED work light. The 5v supply switches the dew heater between max and approx half power through the mosfet - could be a pin from a micro controller.

The LM2576 looks like the ticket, although I will need several.

[bojan]

Well, talking from my experience, it is always better to have separate regulator for each load (appliance) next to it.

The simplest way is to have ~240V AC to whatever DC next to appliance.
If you really want low voltage bus, then the best is to go for, say, 48VDC or AC (suitable transformers may be found, they were used for soldering irons or for power distribution in industrial environment for power tools, like electric screwdrivers and so on) or 15 VDC, and have individual SMPS's (or linear regulators, fol low power consuming appliances, like camera).
This way, you will prevent noise from spreading (from "dirty" loads such are motors or similar) to noise sensitive appliances (camera, for example).
Also, in case of fault on particular appliance, the rest of the system will still be functional.
My 2 cents..

[MrB]

Rowland, is the current limiting a form of protection incase there is a short circuit?
You do understand that although the SMPS you have states 12V @ 12.5A, it will not actually try to push 12.5A thru your camera?
For example, if the camera spec states it has a current draw of 3A at 12V, that is the max it will draw from the supply, 3A.
So the SMPS will still have ~8.5A overhead for other consumers.

Think of it this way, supplies 'push' voltage on consumers.... push too much(high voltage) and the consumer pops.
Consumers 'take' current from supplies, if they take too much the supply pops. Simplistic but hoping it makes sense?

And my apologies if this is already known to you, it just reads as though you are a little confused about this but then that could just be my misinterpretation of your post.

[rcheshire]

Hi Simon. Thanks. I understand the current thing. I want to place a limit of 3 amps on one of the loads in particular, the camera, which comes with a 3 amp adapter and is running a peltier device. It occured to me that this was an avenue for current draw in excess of the limitation imposed by the manufacturer by way of its adapter - cant be sure of my facts, so playing it safe.

I have a handful of L317's on hand, which should improve the characteristics of the supply to each appliance as well. The SMPS I'm using is a little noisy and affects the Arduino, in particular. Trying also to avoid interference with the camera.

I've attempted to provide current/short circuit protection based on Brendan's (wasyoungonce) suggestion in another post, with an N channel mosfet.

[rcheshire]

A little more work with an L317 calculator, power dissipation 75mw at 3 amp and 25mw at 1 amp. This seems very low to me.

12v in 11.975v out R1 40kohm R2 132kohm.

[MrB]

No worries :thumbsup:

[DavidTrap]

Isn't a fuse a "current limiting device"?

DT

[bojan]

It is.. with foldback characteristic, and cheapest of all :P :thumbsup:

[rcheshire]

Dang! So it is!

[MrB]

:thumbsup:

[tonybarry]

Hi Rowland,

You need a minimum of about 2.5V drop across the LM317 to get it working. Otherwise it won't regulate. Vin = 15V, Vout = 12V is good. Again, check your currents and power dissipations.

The 317 is able to shut itself down if thermal excesses occur, but this is not a good idea unless you actually have a fault.

Regards,
Tony Barry

[rcheshire]

Thanks Tony. That detail I missed. I will have to reconsider my approach to all of this:question:

[alistairsam]

Hi Rowland,

I'd suggest a fuse as well. slowblow would work well.
your appliance should draw within its rating. if it draws more and is working normally, then you might have to increase your allocation or planned allocation in your bus. so if its to be within 4A, use a slowblow fuse for 4A.

have you considered the switching regulators that are pin compatible with the 78xx?
they are LDO and usuallly need only 1.2v higher than the desired output.
here is a good one
http://www.dimensionengineering.com/products/de-swadj3
the 5V version is available here
http://www.robotgear.com.au/Product.aspx/Details/464
Cheers

[rcheshire]

Thanks Alistair. A fuse is failsafe, although I was looking for a more elegant solution, such as current regulation with ripple reduction. Working around a less than ideal power supply. I have used cheaper versions of the 78xx pin compatible regulators.

The failsafe option is easier, providing SMPS ripple is very low. Virtually all adapters are SMPS with ? ripple anyway. Noise reduction has been a major goal.

I'm pretty tired of working around the limitations of the power supply. Now looking at a laboratory spec unit. Time is money as they say and I have been saving for some more gear, so a good SMPS will make a huge difference and simplify things enormously and provide more flexibility.

[alistairsam]

Hi Rowland,
As you said, time is money, so to simplify things, just have a separate adaptor for your dew heater and power the remainder with a single supply
That will reduce noise and keep it simple.
Remainder would be a camera, mount, USB hub, none of which are noisy.
You can get a 4 or 5 core wire from jay car, and run both supplies via the same cable. Just use a 5 pin connector on the box.
That's what I do.
Cheers

[Barrykgerdes]

Hi Rowland

I have been watching your efforts for some time but I don't think you understand the relationship between what you have and what you need. Your drawings look great but they really do not solve the problem of getting 7A out of 3A.

Firstly you have a basic power supply that can deliver 12 v at 12.5amps. If this is a standard off the shelf regulated power supply from Jaycar or other supplier you won't need anything else because it is designed to deliver 12v at 12.5a , and you can supply everything off it on a single bus. This is the way we do it in the field.

If it is not a standard off the shelf, eg home made and you will need to regulate it and and that will depend on the actual design as any sort of regulator will need at least 2 or 3 volts in excess of the 12 volt to handle the regulator, and a regulator capable of supplying the full rated load current.

Your stated load requirement is 12v at 7A. There is no way that you can get that from a 12v 3A regulator. I think you have envisaged limiting the current drawn by the items connected to the bus so that they won't draw their rated current. This just won't happen as the item will draw the current it requires at 12volt or not work properly. Yes you can put a current limiting device in the lead to the item but all that will actually happen is the voltage across the device will drop to the point where it will not operate.

Think of it this way. If the item draws 3A at 12v it presents a theoretical resistance of (R= E/I) of 4 ohms. If you limit the current drawn to say 1 A , the voltage available across the unit will be 4 volts (same formula). Of course as you reduce the voltage on the device it will actually draw less current and some point of equilibrium will be met with the 1A. However I doubt if you could get any usefull performance out of the device under this condition.

If you want to isolate each load to avoid interference you will need to increase your main supply bus to at least 15V and run each item off its own secondary regulator. This is hardly worth the trouble as modern commercial devices designed to operate off 12v DC will have some sort of input filtering and any item in it requiring a stable voltage will have a secondary regulator. The worst interfering device you will likely have is a dew heater with a pulse width modulated regulator. These can be noisy so this will be best served with its own simple supply.

Barry

[rcheshire]

Barry. You'll be pleased to know that I am not going to run the dew heater with PWM. I've taken your advice on that one - too noisy for this setup. The 12v power supply is moving on in favour of a better one. Variable voltage and very low output ripple. Much more flexible for future requirements as well.

Just to be clear, my intention was to power the bus with 12v 12.5A, with the appliances pulling at max, approx 7A. That is, 3A, 1, 1, 1 and 1A, plus a bit for LED lighting. I toyed with the idea of using linear regulators to each appliance, to clean up the supply and limit current, but I'm tossing that idea now and opting for the simpler option of feeding directly from the bus, using a better power supply.

I know I cant get 7amp from 3amp.

Rowland

[MrB]

You can run dew heaters on PWM with minimal noise.
Thanks to the thermal mass of the mirror's and lenses we use, you can make a start in the right direction by using a very long period, say a few seconds. That brings the frequency a long way down out of harms way. The only real problem* then would be the harmonics produced by switching the strap on and off, which you should be able to attenuate with an RC low-pass filter.
(*probably not a very big concern, considering the very low fundamental frequency and the voltage used)
The resistance you already have(dew strap) so selecting a suitably sized capacitor, you should be able to get a sensible time constant.

[rcheshire]

Hi Simon. I will look into that, although I'm doing something similar. The plan is to switch the dew heater through a mosfet and cycle at long intervals and low power, in laymans terms - HIGH and LOW from an Arduino output. Not sure how effective that will be? Just a matter of getting the current through the element at an appropriate level. Can be switched in response to humidity as well.

[MrB]

It will be effective, but you shouldn't need to manually set the current, just run the heater at 12v, let it draw what it needs(which for most heaters is minimal) and the PWM does the rest.
By setting the PWM duty cycle somewhere between 0 and 100%, you are effectively adjusting the current(averaged over time).
Say for example the heater draws 1A for 2 seconds, then is off for three seconds(40% duty cycle), that is essentially the same as running the heater constantly at 400mA. PWM is much easier(and likely cheaper) to implement than an adjustable constant current source.

[rcheshire]

This is the final iteration, minus fuses.

Starting bottom left, the 17.6R, 4 x 12R, 4R and corresponding current values, represent each appliance at max load and are not part of the circuit as such. That way, I can get an accurate reading of the total current demand on the power supply.

The dew heater element, 17.6R is switched between HIGH and LOW through a mosfet, at this stage - the Arduino digital pin voltage and setting is represented by the 5v supply and switch.