Mounting Sandstone Resistors

[ghurd]

We need to mount 50 or 60 5W sandstone wire wound power resistors.

Each resistor will be dissipating approximately 2.9W max.  Maximum duty cycle will be about 60 to 70%, but not often.  Ambient temperature stays between 60 and 82F, or 16 to 28C.

The little buggers get 'warm'.  We would like to keep them cooler.

Also if someone knows hot is too hot for the standard mystery surplus wire wound sandstone resistor, please tell.

My first and only thought is epoxy them on an aluminum plate.  Maybe with a couple `C' or `E' channel pieces welded on for more surface area.  Certainly a muffin fan or 2.

How bad is that idea?  Expansion / contraction issues?

Any better ideas?

G-

[Nando]

What is sandstone resistor ?.

I have been in the electronic industry for many decades and it is an unknown "terminology" to me.

If the resistor is just getting warm NO need to heat sink them at all, you start to heat sink if the temperature goes above 60 degrees Celsius (can you calculate the Fahrenheit temperature).

Nando

[stephent]

Well, ghurd, you could try the epoxy, but it might get a bit soft. Try finding a high temp epoxy and it should do. make a small lake of it around each resistor on the sides for extra heat transfer. (make it trail up the sides of the resistors, etc)

but use very little beneath the things so they make as full a contact as possible.

Why not use nichrome wire though?

How many amps ya looking to dump/dissapate and voltage levels?

[stephent]

Sandstone=sandcast body types?

[RP]

I agree if you're below the rated dissipation then you really shouldn't need a heatsink, just some airspace.  That said however, if you're concerned about over load then:

The 5 watt resistors are usually about 1/2" (12mm) square and an inch or so long.  I'd suggest a trip to the hardware store to find some aluminum U channel that you could drop them into.  If it's a snug fit, I'd skip the epoxy and just use some heat sink grease.  You could line up several per channel.

If you can't find a snug fitting channel then use angle aluminum so you can clamp it into a corner.

[ghurd]

Hi Nando,

It's something I always called them.

Cast concrete may be a better term.

You know them I'm sure.

Picture...

http://www.goldmine-elec-products.com/products.asp?dept=1124

[maker of toys]

"ceramic power resistor."  never hit that 'sandstone' term before.

it's a nichrome wire in a ceramic housing. for the home audience, the nichrome is basically the same material used in hair driers, stoves and most other electric heating applications. if kept away from oxygen (like with a ceramic coating) the wire will operate indefinitely at red heat.  the thing that kills these is rapid thermal cycling and the associated dimensional changes.  the ceramic coating helps with this by supporting the joint between the nichrome and the solderable material that forms the leads, and creating a sort of 'thermal inertia' to reduce the rate of expansion and contraction.

the rating of a resistor in watts is how many steady-state watts they can dissipate in free, still air without a damaging temperature rise. if you're operating 5 watt resistors at 3 watts continuous, leave 'em be unless their surface temperature is a big hazard to bystanders.

the muffin fan/s will help quite a bit all by it/themselves.

the epoxy will actually limit the power disipation, and act to overheat the resistors.

(it also will tend to age rapidly from the heat and fail in the not to distant future; some sort of heat-conductive filler is indicated. . . .)

if you just gotta heat-sink them, use heat-sink grease and sandwich them between two Aluminum plates (or heat sinks screwed loosely togeather. (spring load the plates? 10-20 lb per resistor is a reasonable number to shoot at.)  that way the thermal expansion will not be seriously constrained.  your U-channel (with the 'fins' away from the resistors) will probably be your best bet.

[maker of toys]

I forgot to mention:

the ceramic is baked onto the nichrome in a pottery kiln. a 'little warm' isn't going to hurt them.

-Dan

[ghurd]

RP,  I've had them come loose or crack with mechanical fasteners like screws and clamps.  I briefly considered some kind of leaf spring retainer, then decided it was too much work.

Stephent,  About 15A at "12V" dump, about 10~14A peak charging input.   Still testing a dump controller design, so I would like the circuit to operate for a time, and cycle often.

May even add a couple switches to remove part of the resistor bank as the solar input falls.

I have not had the best luck working with nichrome, and there isn't really space for coils where they need to fit.  

And they could (meaning WILL) be subject to physical abuse while long term testing.  I don't want shorted coils that no one notices.  It is in a home.

[willib]

Ok G i'm curious.

parallel or series? or a combination of the two?

what ohms total?

what application needs 50 or 60 5W resistors?

[ghurd]

Hi Dan,

Why has nobody heard "Sandsone Resistor" before? (kidding)

I must look at far too many surplus catalogs.

"Sandstone Power Resistors" is actually the the sub-heading of the above link, which by the way is Not where I bought them. (the disclaimer thing)

These are not the tube wound ceramic type. These are little white or sand blocks with the coils placed into one side.

G-

[ghurd]

Hi Willib,

Still testing that dumpload controller design.  So far, so good.  Except the charging amps vary SO much here this time of year, it will only cycle a couple times a day at best.  Plus we need to be using the batteries.  Summer may take the I(in) to +15A, so I'd like to be able to add more, or remove part of the resistor bank.

Looking at 50 or 60, 50 ohm resistors in parallel for a bit under 1 ohm, at ~14.2V.

The resistors were nearly free.

[Nando]

Ok, cast ceramic body.

Nando

[willib]

another use for varible current control

Ed needs one too , to do his boost controller experiments

you could simulate the dc comming from the panels , and not have to wait for daylight.

you could also use a variac and rectify the output like Jerry is doing.

i had one , it wasnt mine exactly,but dad would have given it to me, if i had known it was going to go in the garbage i could have rescued it.

sigh.

[RobC]

I still think it would be easier to find some old electric furnace elements and use them. I run across dead furnaces several times a year. RobC

[zap]

That must be very hard on your vehicle's tires and suspension.  ; )

[ghurd]

"OHIO" and "Snowbelt".  It's like 8F here now with no sun. Again.

Nobody here uses electric heat.  Thats like a 1909 VDB penny, or alien abductions.  

I've heard of it, but never actually seen one.

It still leaves the issues of nichrome wire.

[ghurd]

I have considered multipule stage dumping.

It is cheap and easy with this design.

(cheaper and easier than variable current, for me)

The system is a real live off-grid system.  Nothing is simulated.  Daylight here sucks.

A variac costs more than everthing combined, except the PVs and batteries.

(maybe I don't read Enough surplus catalogs?)

[DanG]

A nice aluminum alloy sheet (not too soft) and some precision cuts & bends to make a spring finger to trap ceramic against the sheet, friction fit; just strips with one small bend near the end to trap the resistor then nut and bolt to apply tension. Use of HS compound recommended but not necessary...

[maker of toys]

yeah, I looked at the link.

they're still "ceramic power resitors" in the generic sense.  maybe I've not  been exposed to enough catalogs?  (I only have about 7 on my desk at work)

the tube type are generically "wirewound power resistors"  This sort also comes in an adjustable version.

there are also 'carborundum power resistors' which use monolithic carborundum on a ceramic former.  (they are REALLY expensive, but have no significant inductance.) these are sometimes "non-inductive flameproof power resistors"

-Dan

[johnlm]

Thanks for the laugh.

Johnlm

[alancorey]

I mounted some once by milling slots into a block of aluminum that fit with about 0.010 clearance, then mounting the block to a heat sink.  Heat sink compound on all joints.  That wasn't 50 or 60 though.

Maybe make a strap of aluminum the width of the body length, hitch them down to a flat area of the heat sink with the strap on top and a screw beween each pair?  That way you don't have to try precise bends to get a tight fit.  Heat sink compound of course.  Plain old silcon grease if you want cheaper.  You could bend the leads into hooks on the ends and solder the wire into them along the line of resistors, then if you need to test them later just melt the joint and pull the wire back out.

  Alan