Power issue

Hi,

I'm facing a problem with a project that involves multiple temperature readings, motor control, and several relays. I'm using an Arduino mega 2560, powered by a computer alimentation (300 W), in 5V.

The problem is, that when I connect the VCC pin (of the Arduino) to my circuit, the alim block shuts down immediatly. If the Vcc pin is not connected, the alim runs normally, and the arduino is powered correctly. And if I do the opposit, if I disconnect the Gnd and connect the Vcc, the alim runs normally as well.

I tested my circuit with a multimeter, did a continuity test between its Gnd and its Vcc, to check for short cut, and got no continuity. I checked the resistance betwen those, and obtained about 3 kOhms, which seems normal considering that all resistors are put in parallel (altough I didn't do the maths ^^).

The motors I'm using are servo (7 of them), they are "detached" when not used, I'm also using 12 relays (dispached on 2 plates), 9 thermistors, and a DHT 22 sensor.

So the question is, why does the alim shut down ? Could there be a short cut else place ? Could it be a flaw in the chip ? Or is there another problem ?

Any help would be appreciated :slight_smile:

Try putting a big(ish) resistor in series between the supply & Vcc pins, to restrict the current flow (47K might be a good start). If it doesn't shut down, see if you can get an amps reading in series with the resistor. Be careful though... I don't know what the max wattage a 300W computer PSU can put out on its 5v rail, but it may well be substantial.

It may be that when your circuit is quiescent (not powered) then it's fine, but as soon as it powers up some active component is creating a short circuit.

If it still powers down, try bigger resistors. Also try a large resistor (1M+) directly across the power supply 5v rails, if it still shuts down then either you've got the wrong leads (you're not hooking into the standby switch are you?) or the PSU is shot.

PS: If there is a chip or some other device going to dead short when powered up, a 47K might not be enough to stop the magic smoke coming out. So maybe start with a bigger resistance, if you feel that might be the case.

Thanks for the reply !

I tried to put a 1M resistor in series, the alim didn't power down ! I read 3.7 µA. I did the same with a 100K resistor, same thing, the alim still doesn't power down, and I got 36.4 µA. (Indeed, the PSU can deliver up to 20 A in 5V, that's why I'm using it, I need the power for some of the component ^^)

So according to you, there is a component that creates a short cut when I power up the circuit ? How do I know which one it is ? ^^(It's gonna be hard to test them all (but not impossible) since there are about 20 different components connected to the circuit, and they are all welded... I tested them before welding and there was no problem) Is there another possibility ?

It does sound like there's either something trying to draw more power than the PSU is capable of at start-up (any large capacitors across the +5v/GND rails?).

Since your measured current multipled by 10 as the resistor divided by 10, it's obviously trying to draw more power yet. Try going for a 1k resistor, that should give your circuit 3mA, leave it powered for a little while & go exploring with the multimeter, or an oscilliscope if you have one. Check each IC for excessive heat, it should take them a while to warm up on such little current.

Do you have flyback/freewheel diodes on each of your motors & relays?

Nope, no capacitor at all in my circuit...

Indeed, with a 1k resistor I get 2.18 mA, and I tried with a 220 ohms, I get 8 mA. Isn't it normal that it tries to draw more power ? I mean, the Vcc pin is a DC voltage source, so if I reduce the resistance, automatically, the current will increase (to a certain point), right ? Following Ohm's law (to simplify "a little"), if I have (ideally) a 5V source, the resistance of the circuit for 8 mA would be 625 ohms, knowing that I put the circuit in series with a 220 ohms, I still don't have zero resistance, so no short cut ? Am I reasoning correctly ?

I didn't put any diode for the motors, nor for the relays... Since I have never been taught to do so (I have only been practicing for a year), I guess the ones I use already includes them ? (I use SG90 and SG5010 servos and the relays are on a chip)

Hi,

Please read the first post in any forum entitled how to use this forum.
http://forum.arduino.cc/index.php/topic,148850.0.html then look down to item #7 about how to post your code.
It will be formatted in a scrolling window that makes it easier to read.

Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?

We need to see your circuit.

Thanks.. Tom.. :slight_smile:

Hi,

The problem is, that when I connect the VCC pin (of the Arduino) to my circuit, the alim block shuts down immediatly. If the Vcc pin is not connected, the alim runs normally, and the arduino is powered correctly. And if I do the opposit, if I disconnect the Gnd and connect the Vcc, the alim runs normally as well.

How are you powering the Arduino if you have the alim (powersupply??) disconnected?

Tom… :slight_smile:

It is unclear from you posts what "the alim shuts down" means.

What color wires are you connecting to on the power supply? A schematic would be a huge help in understanding what you're trying to do. It sounds as though you're connecting to the auxiliary supply which cannot supply the required current.

Some older computer power supplies must have a minimum load on one or more outputs to maintain output voltage. The typical load is a 10 ohm, 10 watt resistor from VCC/+5V to GND.

Here is a reference for the connections and a plug-in board design:

and

http://dangerousprototypes.com/docs/ATX_Brakout_Board_design_overview

TomGeorge:
Hi,

Please read the first post in any forum entitled how to use this forum.
http://forum.arduino.cc/index.php/topic,148850.0.html then look down to item #7 about how to post your code.
It will be formatted in a scrolling window that makes it easier to read.

I’m not sure I understand, since there is no code in my post ^^
I put a simplified schematics of the circuit in attachment (fritzing made). I’m not using breadboards, I’m using Eurocard full lign pattern, and there are 12 relays, not just one.

TomGeorge:
How are you powering the Arduino if you have the alim (powersupply??) disconnected?

The PSU stays connected to the Arduino. It’s when I link the 5V pin of the arduino to my circuit that the PSU stops functionning; the PSU’s fan stops running, the Arduino receives no more power.

avr_fred:
What color wires are you connecting to on the power supply? A schematic would be a huge help in understanding what you’re trying to do. It sounds as though you’re connecting to the auxiliary supply which cannot supply the required current.

I’m using a red wire of the PSU to supply power to the Arduino (5V and up to 20 A), and it is connected to the jack plug. Should I use a “12V wire” ? (Altough it worked with 5V in a previous version of the project)

avr_fred:
Some older computer power supplies must have a minimum load on one or more outputs to maintain output voltage. The typical load is a 10 ohm, 10 watt resistor from VCC/+5V to GND.

I already put a power resistor, 10 W (connected to a red wire as well), to keep the PSU on. (Indeed, without it, the PSU stopped functionning after half a second, but the problem isn’t the same here. The PSU runs normally, the power resistor heats and the LEDs on the Arduino are on, until I connect the circuit’s Gnd and Vcc to the arduino)

That Fritzing is not at all helpful since it shows breadboards and many servos…you haven’t mentioned servos before and what is the second breadboard that looks like it contains just resistors and capacitors? And what is a Wire 514? And where is the power supply?

Anyway if you are using 5V from the PSU it should not be connected to the Arduino jack plug. The minimum voltage for that is 6V and 7V+ is better.

It would also make a lot more sense if your circuit was connected directly to the 5V power supply rather trying to run all that current for your sensors and servos (or whatever they are) through the Arduino. That is a really bad idea.

Steve

I'm using a red wire of the PSU to supply power to the Arduino (5V and up to 20 A), and it is connected to the jack plug.

So the center pin of the power barrel jack is what you've been calling VCC? I was going to ask about that because nowhere on the schematic or board will you find the letters VCC. The input of the barrel jack is PWRIN in the schematic and is spec'ed (recommended) at 7 to 12 vdc.

Based on the Fritzing, I would not recommend powering anything up until you post a real schematic, hand drawn is fine and perhaps a photo of the bits and pieces you have. I have no idea of what that thing is in the Fritzing marked Wire 514.

As mentioned, you cannot power the servos using the +5 supply regulator of the Mega. Connecting the +12V supply to the barrel jack will let the smoke out due to the current requirements of the servos.

Sorry if the fritzing is useless, I will draw a more complete draft asap.
The "wire 514" stands for the relays.

As I have said, I have been practicing electronic for a (small) year only, and I am on my own with this project, so I probably made other mistakes, I am sorry if I don't seem precise or "serious" enough... :-\ (I did as much research as I could given the little time I have, and the size of the project)

slipstick:
It would also make a lot more sense if your circuit was connected directly to the 5V power supply rather trying to run all that current for your sensors and servos (or whatever they are) through the Arduino. That is a really bad idea.

Indeed it would :smiley: I will change that as well !

avr_fred:
So the center pin of the power barrel jack is what you've been calling VCC?

No, I was refering to the 5V pin on the arduino (on which the breadboards are connected in the fritzing).

I think there are +7V wires comming out of the PSU, I will check that as well, to supply power to the Arduino through the barrel jack.

Do you think my "main" problem is linked to those mistakes ? Could you explain it ?

Thanks for your answers anyway !

Hi,
Can you post a picture of you project please?

Thanks.. Tom.. :slight_smile:

There is the hand draft of the “present situation” (with the arduino still powered in 5V through the barrel jack, and the hole circuit powered by the arduino and not directly by the PSU). I hope this one will be more usefull. :grin:

I also put in attachment pictures of the project, but I am not sure those will be helpfull, it’s a bit messy in there, since I haven’t arrange the wires yet… :roll_eyes:

And the remaining pictures.

So, I connected the circuit directly to the PSU, and the problem remains : now the PSU stops running instantly after being turned on (the fan turns half a second), which is logic if I have a short somewhere in the circuit.

So the question is : how do I find the short (if there is one)? I used a 10 Ohms resistor (and got 24 mA) and checked each IC for exessive heat like AdeV said. I waited 15 minutes and found no IC heating… Should I wait more ? Is there something else to do ?

You've not been very clear but I am assuming that when you say "the circuit" you mean all the servos, relays and sensors. Is that correct? Have you tried connecting just "the circuit" to the power supply with the Arduino completely disconnected? What happens then?

According to your diagram if any one of your relays is ever switched it will connect the 5V power line directly to ground. That's a serious short circuit. You probably don't really mean that, but that is what it looks like. And if just one of them is actually wired that way then that will be the cause of the short. And because it depends on a relay being powered you won't be able to measure it without the power on.

Steve

slipstick:
You've not been very clear but I am assuming that when you say "the circuit" you mean all the servos, relays and sensors. Is that correct?

Yes, that is correct :slight_smile:

slipstick:
Have you tried connecting just "the circuit" to the power supply with the Arduino completely disconnected? What happens then?

Yes, I tried that, the same thing happens : the PSU stops running.

slipstick:
According to your diagram if any one of your relays is ever switched it will connect the 5V power line directly to ground. That's a serious short circuit.

Are you sure ? What I meant on the diagram was that the "devices" (which is either Peltier cells, a heating lamp, a fan, or 12 V leds) are connected to the relays on one side, and to the ground on the other side, but the current has to pass through those devices to reach the ground, so there shouldn't be any short circuit, right ? (and the relays are in "normally open" configuration, so if they don't receive instruction from the arduino, the devices never receive current since the "switch" is open)
The top and bottom wires of the relays are there to power the relay's chip, and the wires on the right are those connected in the relays, they are there to power the device... That was probably not very clear on my draft :confused: