Another way to do it might be like this...

Q1 needs to be a power transistor, and heatsink. It will drop around ½V at .5A. You need to have CR1 thermally connected to Q1, glue the diode to the transistor case.

Took longer than usual for me, its been a long day.

.

Your transistor needs to be a PNP.

 

Why does your load draw as much as 44A?
The battery in my car provides 600A to the starter engine when it is cold. But the clock in the car does not draw 600A! It draws only 0.001A.

 

Wayneh, you could get a lower Vce if you attached the collector of Q1 to the +12v supply via a current limiting resistor; say 100 to 150 Ohms, rated at 2W or more. Otherwise, you won't get a Vce below around 0.8v due to the Darlington configuration.

This circuit is similar to what Wayneh posted, but using an N-ch power MOSFET instead of transistors.

R1/D1 establish a reference of ~700mV; it's just an inexpensive way to get a moderately stable low voltage reference using commonly available components. The voltage across D1 will vary over temperature and with battery voltage somewhat, but should be plenty accurate for what you're doing. R1 supplies ~10mA to D1 to develop the voltage across D1.

R2 is a 10k pot that allows you to adjust the maximum current setting. I have it set to roughly 2/3, so the node labeled "Ref", shown as V(ref) on the right, is about 700mV*2/3 = 466mV.

As current flows through Rsense, an 0.8 Ohm power resistor, a voltage develops across Rsense, shown as V(Sense) on the plot. E=IR, and V(ref) ~=466mV, so the opamp will adjust the MOSFET's gate voltage to try to keep the noninverting (+) and inverting (-) inputs to the opamp the same. 466mV/0.8 Ohms ~= 583mA, which is approximately your desired limit.

C1 is in the circuit again to show how things act when there is a dead short on start-up. Note that the power dissipation in the MOSFET is quite high at the beginning; around 6.5 Watts. As the capacitor charges via the current-limited current sink provided by the MOSFET, power begins to be dissipated in the load (Rload) and less in the MOSFET. If the MOSFET does not have sufficient heat sinking for periods when the output is shorted, it will overheat and burn up.

Note that after the capacitor becomes near fully charged [I(c1) falls near zero] that V(gate) increases sharply, turning the MOSFET on fully as V(sense) falls below V(ref). At this point, there is very little voltage drop across the MOSFET, and very little power dissipation in it; virtually all of the power is being dissipated in the load.

The parts shown can be found in any Radio Shack store.

An LM358 or LM2904 would be a better opamp to use as they are dual opamps rather than quad. You'll need to make certain that the other opamps in the package won't cause problems by wiring each unused inverting (-) input to it's output, and all of the non-inverting (+) inputs to ground.

 

I can't find the LM324 in the list of components in LTSpice.

 

It's not supplied with LTSpice; you have to add it.

 

Wayneh, you could get a lower Vce if you attached the collector of Q1 to the +12v supply via a current limiting resistor; say 100 to 150 Ohms, rated at 2W or more.

Slap-in-the-forehead simple. Great point.

I had a 2N3055 looking for a home, so I used it here. Would a MOSFET such as IRF540 (which I also have a bunch of) provide any advantage here? I guess I wouldn't need to worry about the current load on Q1, except to limit it.

 

I threw a smattering of opamps in the attached .zip file.

Close LTSpice before you copy the files into your SwitcherCad subdirectories.

The .asy files go in \Program Files\LTC\SwitcherCad\lib\sym\Opamps
The other files go in \Program Files\LTC\SwitcherCad\lib\sub

Then start LTSpice again. You should be able to add those opamps to a schematic using the Component icon (looks like an AND gate) from the Opamps subdirectory.

Also attached is pot.zip
Put pot.sub in the \lib\sub directory, and pot.asy in the lib\sym directory.

 

Slap-in-the-forehead simple. Great point.

Make sure to use that current limiting resistor, or you'll likely burn up Q1.

I had a 2N3055 looking for a home, so I used it here. Would a MOSFET such as IRF540 (which I also have a bunch of) provide any advantage here? I guess I wouldn't need to worry about the current load on Q1, except to limit it.

One advantage is that once the MOSFET gate is charged/discharged, there is virtually no current flow from the opamps' output. Ib is eliminated, which makes the circuit somewhat more efficient. Also, you eliminate Vbe(sat) from the equation entirely; with a 2N3055 you might get down to perhaps 0.2v Vce if you're supplying lots of base current, but with a power MOSFET, you just have Rds(on). An IRF540 has an Rds(on) of 77m Ohms; with Id ~=2.6A, you'll get a Vds of around 0.2v. I suppose it depends on what your desired current will be and what MOSFET or transistor you use.

 

Can someone explain to me why I am getting this small current spike at the front?

It's hard to see in the picture, but the current spikes to ~630mA at first and then falls to about 570mA right away.

 

Can someone explain to me why I am getting this small current spike at the front?

It's hard to see in the picture, but the current spikes to ~630mA at first and then falls to about 570mA right away.

When I used your model of the LM324, I didn't have this problem. Interesting...

 

Try changing the stop time to 0.1mS, or just zoom in on the spike by clicking and dragging.

I don't get that in my simulation, but some overshoot would actually be expected. The transients would be short-lived though; as quickly as the LM324 could respond it would adjust the gate voltage.

 

The LM324 model I used came from ST Microelectronics.
The LM324/NS is a National Semiconductor model.

 

Is there a singular website to find add-on models for LTSpice? I am looking for the 7812 linear regulator now and can't find it with a google search.

 

Did you join the Yahoo! group, "LTSpice Users Group"?

Lots of models/symbols available on that site.

 

Did you join the Yahoo! group, "LTSpice Users Group"?

Lots of models/symbols available on that site.

Yeah, I did that when I first installed the program. Maybe I need to search around more on that website. I was looking under the \lib section and didn't see much.

 

Yeah, I did that when I first installed the program. Maybe I need to search around more on that website. I was looking under the \lib section and didn't see much.

It's under "Linear Regulators"

http://tech.groups.yahoo.com/group/LTspice/files/ Lib/Linear Regulators/

 

Well, I think I will probably pick up the components for that circuit and start building. I'll let you guys know how it works out.

One last question. Is is possible to regulate the voltage from my battery (12.6V - 10.0V) to keep it at a constant 12V? I have never hear of any such thing but thought I would double check just to make sure. I could use a 7812, but then I would need a 4 cell LiPo instead.

 

Your transistor needs to be a PNP.

So it does. Thank you.

 

Is is possible to regulate the voltage from my battery (12.6V - 10.0V) to keep it at a constant 12V? I have never hear of any such thing but thought I would double check just to make sure. I could use a 7812, but then I would need a 4 cell LiPo instead.

The 7812 needs at least about 13.6 volts in to hold at 12v out, so no. The only way you can get more out than you put in is to use a DC-DC converter, but you don't want to bother with that.

 

The 7812 needs at least about 13.6 volts in to hold at 12v out, so no. The only way you can get more out than you put in is to use a DC-DC converter, but you don't want to bother with that.

Yes, I realize that the 7812 won't work, but are there any other relative simple devices/circuits which will?