I have a proto shield project that needs to have 15V supplied to it (from a connector on the board) (long story short: I want to adapt the 5V outputs of the arudino to drive synth circuits which require higher voltages)...and I also want to be able to run this beastie without having it hooked to the USB on my computer. So I wanted to power it from the same 15V supply.
I was concerned that connecting the 15V supply straight to Vin would make the regulator on my mega 2560 work too hard/run too hot, so I thought I would drop the voltage to 9V with a zener.
To be more explicit, my 15V supply connects to the other 15V points on the board, and also to a 500R limiting resistor which then goes to the cathode of a 1N4739 zener. The anode is connected to the 15V ground, which is also connected to the proto shield ground. I also have a 10uF cap connected across the zener to help stabilize any fluctuations.
On my breadboard if all I have is the 500R and the zener, I get the expected 9.1V at their junction. On the proto shield, if it's not plugged in to the Mega, I also get the expected 9.1V. But when I plug the shield into the Mega, the voltage at that point drops to just under 5V. I expect the 5V pin near there to go to 5V, but not the top of the Zener. And the 5V supply (which I do also use on the proto shield) drops to about 3.6V.
All of this occurs even if I remove all of the (socketed) transistors from the circuit. And since I get 9V when I'm not plugged into the Mega, my presumption is that I do not have a short in my circuit.
Am I drawing enough current to the Mega to need to reduce that limiting resistor down to 200R or lower? Am I really likely to be drawing 20mA between the zener and the regulator of the Mega (that's what it would take to drop 10V+ across the 500R)? Or am I doing something more stupid than that?
You can use Ohm's law to calculate the current through that resistor from the voltage drop. From your description there is no doubt where the voltage drop is and no doubt why it is occurring. The resistance is excessive for the current and voltage drop required. This setup does seem like a nasty way to generate 5V. Can you install a switch mode power supply? That would be a much nicer solution. Otherwise at least use an external voltage regulator to drop the supply to a defined level. With a substantial part of the drop occurring at a resistor, the supply to the Arduino's regulator will be all over the place and who knows how close you will be to brown-outs.
I have to have 15V on the proto shield, to be able to interface outward, regardless of a switch mode or other type of power supply?? It is already coming (at 15V) from the external supply, which does happen to be a switched mode supply. Not sure what that has to do with needing 5V for the Arduino itself, other than I'm trying to avoid multiple supplies.
Is it going to be a good idea to put a regulator on the proto shield when there's one on the mega as it is? The only reason I'm doing the drop from 15V to 9.1V is so that I feed that regulator something that's not going to make it so stinking hot (especially with a shield over it); I'm not trying to use a zener instead of a regulator. And I'm only using a resistor there as a current limiter; I'm not trying to feed the regulator from just a resistive voltage divider.
It just seemed weird to me that the mega would be drawing 20 mA, and I'm trying to be sure that makes sense before I change the circuit, in case there's something else I'm doing wrong causing that kind of current draw.
So maybe it would be better if I'd just asked "is it normal for the regulator on Vin for an Arduino 2560 to draw 20mA", so we don't get off in the weeds of telling me to use a different power supply? I don't see anything telling me the expected current DRAW from the board, only its ability to supply 50mA.
I haven't memorised the values, but there have been numerous studies into the power consumption of different Arduinos, mainly focused on ways to minimise it to maximise battery life. I feel sure that a Google for Arduino current consumption would locate the information you're asking for.
For a Duemilanove (just one...mine...there might be unit-to-unit variations):
With a 9V power supply only (no USB), current consumption is 26mA.
With USB only (no power supply), current consumption is 35 mA. This was measured with a DMM after fuse F1 was removed (the DMM measures in series with the fuse).
With both the 9V power supply and USB plugged in the current consumption is 35mA from the power supply only.
So it looks like the FT232 does enter into a partial shutdown mode when there is no USB data, but still consumes current.
which would lead me to believe it is normal. Sorry I didn't think of googling it that way, I was originally just thinking about the voltage and current together which would be why I didn't see any of that sooner.
By the time I found a resistor small enough to keep the voltage at something the regulator could work with (100R), I found that my actual current draw from the Mega is on the order of 70mA. Power = I^2 * R == .49W ... which is bad juju for a 1/4 W resistor; heck I'd probably be burning myself on it if it were a 1/2W resistor.
Lower current draws from before were because the current would rise, drop the voltage below nominal, reset, lather, rinse, repeat. Even if I went with a 50R 1/2 W I have around here somewhere, I don't think that would be good either.
So yes, the zener was a bad idea.
I'm just going to hook the 15V to Vin and have done with it. At least it's within the rated spec.
