Hi Guys,
Does anyone know how I could step down from 12v car battery to arduino 5v/3.3v operation. I will be installing my arduino to my car. I was thinking about a voltage regulator?
Thanks Alot for the help.
Ayman
Hi Guys,
Does anyone know how I could step down from 12v car battery to arduino 5v/3.3v operation. I will be installing my arduino to my car. I was thinking about a voltage regulator?
Thanks Alot for the help.
Ayman
Yes, a voltage regulator.
Which one depends on how plan to use it.
Switching regulator like this so you're not just creating a ton of heat.
Be careful with powering with the +5v pin. You should read this topic
http://arduino.cc/forum/index.php/topic,82046.0.html
and especially this response.
Hmm. Reminds me of the discussions regarding the need for a voltage selector jumper or switch on the boards.
I'd suggest a USB charger that fits in the cigarette lighter socket. Cheap, efficient... just need a USB cable between it and the Arduino to power it up.
Yeah, that'd be a cleaner way to go. Example:
http://www.eforcity.com/81407.html?efprcggbadtf120904=sc+gb+81407&utm_source=google&utm_medium=product_search&utm_campaign=google_base&aid=1192&gclid=COfdlM2SoLICFUfd4AodSRQAtw
Kind of low power
Output Voltage: Vout: 5.0V, Range: 5.0 - 5.6V, Vpp: 200mA max
Pretty sure higher output current units can be found.
Here's a 500mA unit
http://www.smarthome.com/6410/Sima-Home-Car-USB-Power-Adapters/p.aspx
Search for "car usb power adapter" vs cell phone charger.
Why all the talk about this charger and that charger? And a power selector switch? Why not put a voltage regulator on the damn thing that will handle automotive voltage levels? Is that so difficult? Is there a cleaner way than that?
edit: Here is one model (not Arduino) that already has that capability. And you guys can't??
I use the usb port on my routers to get power for mine if I use automotive power, but at $160 per unit, it is a bit expensive just for power.
I thought all Arduino boards already have a regulator... What Arduino are you using? My Uno runs just fine in the car with 12V plugged into the barrel jack. (I do have a DC-DC converter generating the -12V supply for an op-amp.)
They do have a regulator. A linear (not switching) regulator. An Uno may be ok with a linear on 12 volts, but automotive can get up to 15 volts. My Mega (or an Uno) with an ethernet shield onboard is a whole different thing.
SurferTim:
They do have a regulator. A linear (not switching) regulator. An Uno may be ok with a linear on 12 volts, but automotive can get up to 15 volts.
The Uno's voltage regulator has a maximum input voltage of 20 VDC, 12 VDC is only the top of the recommended range so it can handle any possible voltage level from an automotive battery. Of course, the higher voltage will mean more heat generated by the voltage regulator and unnecessary heat is generally something to avoid. However, depending on other aspects of a given project this could be acceptable; especially if the board is not fully enclosed or if the enclosure has sufficient ventilation holes.
That is not correct. With a linear regulator, the limit is not current, but power. If the device draws 200ma at 5 volts, that is 1 watt. If you use a linear regulator and an input voltage of 15 volts, the regulator must dissipate 2 watts to regulate that load of 1 watt. I guarantee looking at that hardware setup, that is way over its limit.
SurferTim:
That is not correct. With a linear regulator, the limit is not current, but power. If the device draws 200ma at 5 volts, that is 1 watt. If you use a linear regulator and an input voltage of 15 volts, the regulator must dissipate 2 watts to regulate that load of 1 watt. I guarantee looking at that hardware setup, that is way over its limit.
If by "hardware setup" you mean the copper pad used as a heat sink on a stock Uno board, than you should realize that it's much larger than just the cut-out in the soldermask. I took a caliper to an Uno R3 board I had handy and estimated the surface area to be about 132 mm2, also I'll point out that there are several vias that physically connect this pad to a somewhat larger copper pad on the other side of the board. The effect wouldn't be quite as good as doubling the total heatsink area but it will be close. I'll also note that my estimates are intentionally conservative, and aren't taking into consideration that these two pads are connect to traces and one or two other pads.
Next let's look at the datasheet for the regulator used on the Uno R2, R3, & SMD; the NCP1117ST50T3G. From the datasheet the power dissipation is internally limited, once the die reaches 175 °C the regulator shuts off, but there is a formula for estimating power dissipation (PD) for a passively cooled SOT?223 packaged version of this IC is: PD = (150 °C - TA)/ R?JA. On page 8, "Figure 21. SOT?223 Thermal Resistance and Maximum Power Dissipation vs. P.C.B. Copper Length" the relationship between minimum pad size and expected thermal resistance (R?JA) . The measurements are for a square copper pad with sides L in length, but 132 mm2 is the surface area of a square with 11.489 mm sides and as stated before the second pad increases the effective size. So for the sake of arguement let's say the total effective heatsink is equal to a square pad L = 22.5 mm, and going by the lower curve an R?JA of approximately 71 °C/W, maybe less. So to get a PD of 2 W, 150 °C - TA needs to be at least 142 °C, therefore TA would be 8 °C.
I admit that the passive cooling is not as good as I previously thought. However, it's still physically possible for the current versions of the Uno to dissipate 2 W with the on-board regulator. Although in most cases active cooling, like high speed muffin fans or blowers drawing air from outside, would be necessary.
Maybe it is just me, but where is all that heat sink you are talking about? This page has pics front and back.
SurferTim:
Maybe it is just me, but where is all that heat sink you are talking about? This page has pics front and back.
http://arduino.cc/en/Main/ArduinoBoardUno
I already stated, the copper pads under the solder mask (the thin blue and white material covering most of the board). They are the raise areas almost filling the board space between the barrel jack and USB port and on the other side a somewhat larger one. They visible one the full sized pictures on that page if you look closely. Technically any thermally conductive material will act as a heatsink if placed in physical contact with a heat source above the ambient temperature, and copper is a rather thermally conductive metal. Of course thin flat copper pads on a PCB aren't as effective as the arrays of fins one might normally think of as "heatsinks", but the chart in the datasheet is intended for just using copper pads without an additional parts.
If the heat sink is under the conformal coating, it is a suppository. Subtract 50% for the coating, and I guess 1 watt.
SurferTim:
If the heat sink is under the conformal coating, it is a suppository. Subtract 50% for the coating, and I guess 1 watt.
Solder mask is not conformal coating, at least not what is normally ment by the phrase. They normally use different materials and conformal coating is appreciably thicker because they have different purposes. As the name implies, solder mask's primary purpose is to keep solder from adhering to most of the PCB while also providing some electrical insulation to the contductive traces and pads. However, unlike conformal coatings, solder mask is easy to scrape off. Some people can even do it with just their fingernails! In contrast, conformal coating is added to components to protect them from dust, moisture, harsh chemicals, and temperature extremes (some methods of applying it may also coat the PCB as a by product, but the real reason it's applied is to protect the components). So many types of conformal coating are intentionally thermally insulative, and all have to be relatively thick and difficult to remove.
The solder mask will reduce the cooling slightly vs. bare metal, but not nearly as much as conformal coating would! Although not specified either way, I personally think the effect of the solder mask is factored into the graph in the datasheet. It would be very unlikely that the manufactures are expecting their customers to have +20 mm2 bare pads on a PCB using surface mount parts, and to work at the recommended maximum of 12 VDC pads that big often would be necessary. Even if the chart data wasn't factoring a mostly covered pad, I doubt it would make more than a couple precentage points worth of difference (at most).
We are discussing a topic that shouldn't even be here. If that were a switching power supply, how much power would that dissipate? Jeez!!
edit: My stuff works. I know it is a strange approach, but after three decades of electronics experience...
SurferTim:
We are discussing a topic that shouldn't even be here. If that were a switching power supply, how much power would that dissipate? Jeez!!edit: My stuff works. I know it is a strange approach, but after three decades of electronics experience...
I never stated that a switching power supply wouldn't work, nor was my intent to question your ability. As far as I'm concerned the thing in contention was the specific claim of 200 mA at an input voltage 15 VDC being impossible for an Uno's linear regulator.
My own experience is that it is possible to power an Uno, with minor additional current draws, beyond 12 VDC, though it was with batteries that had a peak voltage closer to 14 VDC than 15 VDC. So instead of simiply dismissing 15 VDC as impossible, I actually wanted to check it out. Even if what I initially thought was wrong, at least I would know for sure and it would be a potentially useful example to others on the forum of how to use some of the more esoteric information in datasheets. As it turns out I wasn't completely correct either, it's possible but for most conditions impractical for any reasonable length of time (without actively cooling the Uno). IMHO, that is something worth knowing and sharing.
In any case, I agree with you that going directly from an automotive battery with a voltage significantly above 12 VDC is not an advisable design choice. Using a separate switching regulator to reduce the voltage to 7 VDC to 9 VDC and then use the on-board regulator to drop it to 5 VDC, would be more energy efficient and thermal managment would likely be a non-issue. Of course, using two linear regulators in a similar fashion would also make thermal managment much easier to address, if only by spreading the heat needed to be dissipated (though it still wouldn't be any more energy efficient, probably less so). If any of these alternatives are practical for a given project they would be a significant improvment over a direct connection to an automotive battery.