Dummy load box heatsink Q

[Chuck H]

I'm building a load box, 4/8/16, good for 150W. I plan to put it in a 5x7x3 aluminum project box. I'm using those aluminum heat sink cased resistors, the gold ones. There will be six of them mounted with heat sink paste right on the aluminum.

So my question is, do I need an additional heat sink? My original plan was just to mount the box on feet and drill some holes in the top and bottom so the heat rising out the top would draw cooler air into the bottom.

Any advice appreciated.

Chuck

[Robert M. Martinelli]

Hi Chuck,
Those resistors are capable of dissipating without burning out their rated power, so six 25 W resistors surely won' t disintegrate themselves, but when driven with full power they' ll be hotter than hell, and, being aluminum a very good heat conductor, the box will be at the same temp, so........just for your peace of mind and safety, how about active cooling?

1 - Peltier cell(s) plus heatsink

Or

2 - PC fan on the top to extract heat from the can through some holes

Or

3 - 2 PC fans both pushing air up, one at the bottom taking fresh air from below ( provided you leave some spacing from the floor ), the other on the top extracting the heat

Or

A combination of all the above solutions

Uhmmm... another idea.....A very smart thing to do would be to make this dummy load an "intelligent" one, adding a thermostat ( in physical contact with the box, using a NC contact, open when in alarm ) to switch off a relay when the temperature reaches a dangerous level, ( should cooling fail or resistors burn ) - the relays NO contacts should cut mains voltage on the amp, making it an active safety device.

Hope this helps

Best regards

Bob

[loudthud]

I find a heat sink really isn't needed unless you want to run for long periods of time like 20 or 30 minutes at say 1/2 the power rating. If you look for the spec sheet on the resistors you should find a derating chart of power vs temperature. They can run at half power at something like 100 degrees C.

[cbarrow7625]

Yeah, but 100 degrees C is not really safe for the user. The general rule of thumb for heat sink temperatures is to keep it below 140 degrees F and/or 70 degrees rise above ambient (if ambient is higher than 70 degrees F).

Any higher than that and no regulatory agency would let it pass (UL or others). I know this device is not going to be submitted to UL but you still want the people using it to be safe as well as limit your own liability (people can sue you for anything) by showing you took reasonable precautions in the design (that is, if you intend to sell it - probably you don't).

You can convert the Audio signal to 12 VDC by using a bridge rectifier & some isolation resistors to power a small 12 Volt fan (I think Marshall did this on the PowerBrake). The isolation resistors also have to be large enough to dissipate the voltage dropped across them. Using the isolation resistors will help keep any interference or loading from the fan ending up back in your audio signal. By the way, you can run the fan off of voltages lower than 12 Volts. It will make it run quieter & allow you to use a larger dropping resistor for better isolation from the audio.

[Chuck H]

Vox, This is an item for my test bench. I don't expect it to get heavy continuous use. And active cooling seems too pinache for a test bench dummy load. But I may yet be building one as you describe if I blow mine up

Mr. thud, Good advice on looking up the derate curve for these resistors. I think I did once and these resistors were really impressive. I can't imagine any time when it would be used for 30 min at 1/2 the power rating. I'm putting a low level speaker out on it, so I may use it for practicing quiet at home with my 20 watt amp. But thats not continuous power.


cbarrow, If you read above you know that I'm not worried about anyone stressing, abusing or improperly using this box. It's just going to sit on my test bench. But I really like your idea for powering a fan. It would somehow seem silly (to me anyway) to have a dummy load you have to "plug in and/or turn on.

Thanks guys. I think I'm good to go. I may use 4 rows of 1/8" thick angle bar aluminum for the feet just to add more heat sink. I'll leave a space inside for a fan so if it seems sketch after I've used it awhile I can add one powered by the box itself.

Chuck

[defaced]

If you're going to try to expand the volume of metal used for moving heat, it needs to be in imitate contact with the resistor. The easiest way to do that is to use thermal paste between the resistor and the extra heat sink.

I'm using a 150w or 200w resistor in my amp as a built in dummy load and I picked up a chunk of scrap aluminum from work that will be extra mass for cooling. If you have access to junk computers, CPU heat sinks would be another good source of cheap, passive cooling.

[Chuck H]

If you're going to try to expand the volume of metal used for moving heat, it needs to be in imitate contact with the resistor. The easiest way to do that is to use thermal paste between the resistor and the extra heat sink.

So It wouldn't be OK to paste the resistors to the inside bottom of the aluminum box and paste the extra heatsink to the outside bottom? I'm pretty sure I've seen it done that way.?.

I'm using a 150w or 200w resistor in my amp as a built in dummy load and I picked up a chunk of scrap aluminum from work that will be extra mass for cooling. If you have access to junk computers, CPU heat sinks would be another good source of cheap, passive cooling.

Yea, I didn't realize until I had to price them...But big heatsinks are a bit spendy. I've probably run across a hundred of them over the years that I could have salvaged. But I didn't know I should. I certainly will from now on.

[Enzo]

Drop by the local computer repair joints and ask if they have any scrapped CPU heatsinks they might give you or sell cheap.

If you have a real heatsink and an aluminum chassis, when bolted together they form one piece of metal thermally, for the most part. Yes the fins can go outside.

If you buy actual "heatsinks" from an electrical supplier, they will be expensive. if you go to the local metal seller or even Home Depot, you can get some heavy aluminum U-channel and cut lengths of it to serve as heat sinks. You could use a section of wide U and then cut some narrower U and stack them one inside the other.

You could bolt a motorcycle engine head to the thing as a heat sink. Mount a couple red-jewelled pilot lights on the spark plug holes.

[Chuck H]

Drop by the local computer repair joints and ask if they have any scrapped CPU heatsinks they might give you or sell cheap.

Well, I don't get to "drop by" many places anymore. Ice Station Zebra. The nearest computer repair is Geek Squad in Mount Vernon, 70 miles, one bridge, small towns and lots of stoplights. I do get out that way and I'll check with them next time I do. Or I'll just use some aluminum channel from Ace Hardware. I can "drop by" there.

I assembled it last night and didn't even drill any holes yet. I played my 20 watt build through it for about a half hour and it just got warm. So I just ponded it with heavy metal for about 10 min and then it got very warm, but not too hot to handle at all. Still, I'm pretty sure that if I ran a 50 watter on a square wave through it for even a little while that it would get very hot. The amp I'll be working on for while will be 20 watts. So I'll just babysit it until I can salvage or devise a heat sink and I think that'll be good.

Thanks guys

Chuck

[defaced]

So It wouldn't be OK to paste the resistors to the inside bottom of the aluminum box and paste the extra heatsink to the outside bottom? I'm pretty sure I've seen it done that way.?.

Heat transfer is directly analogous to ohms law. You have a hot resistor and a cold heat sink, that acts just like a voltage. The heat sink, because it's metal, moves heat easily, so that's like a low resistance; air does not move heat easily, so that's like a high resistance. Heat is like current. So if you want to move alot of heat from your hot resistor to your heat sink, you need something to replace the tiny bits of air between the two, which is what the thermal paste is for.

Do you need the paste, no; would it make things work to their full potential, yes. You surly won't hurt anything if you try it without so see if it does what you think it should do, if it doesn't, add some paste. A tube of the good stuff (arctic silver 5) costs like 5 to 10 bucks from a real computer store.

[martin manning]

How important is it that the load resistors be non-inductive types?

Seems like if you want the highest accuracy measuring power output you would want a purely resistive load.

MPM

[dai h.]

my understanding was that the inductance of regular Rs is too small to matter at audio frequencies (needs to get up to the mH range to matter). Tried measuring an 8 ohm 50W Dale aluminum housed and the L was pretty small. Forget exact value but was in the microhenry range. Also (since Ls in parallel is less L), you could put regular ones in parallel to get less L.

[Enzo]

And whatever self-inductance they have is also paralleld by their own low resistance. The effect of all that is minimal.