I’m aware that it will get hot if it steps down voltage (which it will have to in my project), and so that rules out hot glue or any other thermally-insulating glue. I’m mounting it to the inside of an Altoids tin, so I figure I should use the tin as a heatsink–but I’m a programmer/electrical guy and have no idea what my options are.
epoxy? i’m not sure…or else you could use “the last glue” www.thelastglue.com . this stuff is awesome for everything. i have it, and it is perfect. it honestly glues everything (except food grade polypropelene or cotton).
Are you using a breadboard or just a board with holes in it? Either way, you can simply bolt the heatsync to it and have it stand straight up, like this
[PoorASCIart]
]|
|
-----|---------
[/PoorASCIart]
If that made sense. I can try to find pics.
EDIT… Just read your post… properly… I guess you could drill a hole in the tin and use the bolt hole for what it’s intended (a bolt), and simply secure it this way. You may have a problem with the contacts touching though… not sure of exactly how it looks.
What voltage are you regulating from and what load are you using the regulator with? Low-dropout regulators can get hot, but as long as you’re not putting too much strain on them and they have enough air circulation you can possibly get away without a heatsink at all which makes mounting a great deal easier. If you’re not stepping down the voltage too much and the load is sufficiently low you may not be drawing too much power through the regulator. Be careful though, the regulator will still get hot and could burn you if you’re not cautious. As for mounting, the regulator has a standard TO-220 package so the legs are .1 inches apart. You can simply place it on a protoboard of some sort (also available at RadioShack) and wire to it as necessary.
Jbot,
The 7805 series regulators have the center pin and the tab at the same potential, i.e. power supply common. It the Altoids can is going to be grounded, you can just bolt the tab directly to the can. A little heatsink goop will help in the tranfer. The way this works is a simple calculation on the drop and current for the regulator. If you are dropping 4 volts (9 volts in to 5 volts out) at 100 ma you are going to have .4 watts dissipated by the regulator. Not too bad. If you are dropping 7 volts (12 in and 5 out)at 1 amp then you have to get rid of 7 watts and that will cause you some problems with just the little can to get rid of the heat. You can always add a good heatsink to the bolt on the outside of the can to increase the amount of heat you can transfer to the air outside of the can. That way both the can and the heatsink will dissipate the heat. Remember that this family of regulators does have temperature compensation that will lower the output if they start to run hot.
Also: The hole is what you use to fasten these things to heat sinks and altoids cases. Nobody uses adhesives.
Use a small (#6) nut and bolt to fasten it. White heat sink goop will help a lot. An altoids tin is probably good for 1.5 or 2 watts, max, assuming free airflow. whytheheckme’s suggestion to bolt a heat sink to the outside is a good one, with heat sinks, bigger is always better.
Put some tape under the metal leads coming out of the regulator so even if bent a bit they won’t short to the case.
I might as well say what this is for…I’m planning on copying someone else’s idea, wiring up the 5v from the regulator to a USB port. That way I can charge my iPod and at least power (not necessarily charge, due to some slightly more complex circuitry in the charger) my Motorola Razr on the road from a 9v battery (or 2). I don’t honestly know what the draw is like to charge/power either device. I suppose I should do my homework before I tinker around any more…or I could do it by trial and error.
The more practical/serious use for this is stepping the 12v on the robot to 5v for my custom camera board. Again, I have no ideas as to how much this draws. I can pull out the meter and measure it, if need be.
Thanks (for taking me seriously after I said flat-out that I was mounting something in an Altoids tin…)
JBot
EDIT: [offtopic] Al, did you guys get your faulty pot wiring figured out yet? It looked like someone had cannibalized those PWM y-cables for pot wiring–every single one of those I’ve used has turned out defective, so I don’t use them anymore. Had you seen that before Atlanta? And was that the problem you were having at Atl?[/offtopic]
Sorry if I seem a bit incoherent/rambling…I really should get some sleep. And yet I continue to type…
you know Ive used the 7805 for lots of projects(usb car charger, loads of ttl circuits off of 9/6v), and Ive never had a problem with over heating(not to say it doesn’t get hot just that it doesn’t seam to be a problem). so you can put a heat sync on it if you’d like but I think that may be over complicating things.
if it is for charging a USB device it should not pull more than 500 ma. I made one of these devices a while ago and just soldered a 9V battery snap to the input side of the LM7805 and a usb connector on the output side and wrapped it all in a blob of electrical tape after bolting the LM7805 to a piece of metal. It worked. Since than more efficient designs have been released on to the internet such as the MintyBoost
Regulating 12V to 5V with that regulator and an unknown load might produce a good deal of heat. For that case you’re going to want to have a good heat sink at least. Is it not possible to power it off the RC directly? If you’re worried about it you could use a switching regulator for the same purpose. They’re much more efficient and shouldn’t get very hot at all.
I actually had a friend who made an iPod charger out of an Altoids tin a number of years back. It worked pretty well, and always smelled curiously strong.
I believe because of poor wording that this is not legal under the 2007 rules. However, this is irrelevent (until next year, anyhow) because there is no tech inspection at IRI…I’ll just make sure I insulate it well.
I will measure the load of that camera board, possibly tonight, to see if it is going to work for my purposes.
And I’ll probably go through with that curiously strong charger. I’m taking a look at MintyBoost to see what I can borrow (steal) from there.
Now before any body jumps at me, if it gets hot enough to melt the hotglue he has bigger problems then how he is mounting it. I have use a 7805 to regulate down form 9 volts to run a servo that had a big load on it. The regulator got now where nere hot. It got “warm” if you can call it that. it was barely enough for me to feel.
Oh, I know that I had a discussion for awhile during the season here on CD, but I didn’t know there had been an official response. Thanks for the link.
As Al points out, the 7805 case (and mounting tab) is tied to the center pin (power supply common). For some applications, you might need to isolate the regulator from the chassis. (In FIRST, for example, the robot chassis is required to be isolated from the +12VDC AND return/ground). There are thin mica insulators available for the TO-220 package that will provide electrical isolation with minimal thermal resistance. If you use your chassis as a heat sink and wish to keep it isolated from the power supply, you should use one of these mica insulators - with some thermal compound to improve the thermal conductivity between the regulator and heat sink.
Not to de-rail this topic again too much, but I wanted to add an update. After further investigation, the wiring was not at fault at all. We do not use ‘cannibalized PWM y-cables’, but actually create our own out of bulk PWM wire. The wiring was stressed a bit, but not defective or faulty. The digital pots, however, were where the problem was. Both the elbow and the shoulder pots are the MA2s from usdigital.com, and both died, either due to static shock or strong magnetic field too near them (as they work off change in magnetic fields to determine rotation). We are still investigating the root cause.
Either way, we are fixed, and ready to rumble at IRI now
