Heatsink temperature may ruin resistors?

[drillz]

I have thick film planar resistors dissipating 2KW.

They are mounted on a heatsink with thermal resistance = 0.144 degree C/watt.

The resistor datasheet says that if the heatsink gets above 210 degrees C then the resistors can handle zero percent of their rated power. (Each resistor can handle 900W continuous power. -there are eight 1000 ohm resistors which are in four blocks of parallel pairs to give 2Kohm.-They are across 2000V. Each resistor dissipates I^2R = 500W.)

This is a problem as the 2KW will make the heatsink go to 288 degrees C. (0.144 * 2000 = 288 degrees C)

I am wondering if i have made some mistake here?

 

[Skogsgurra]

Yes. And to make it worse, your 288 degrees are not temperature but temperature rise. So your heat sink probably ends up at 300+ degrees.

More heat sink or a blower is needed. Plus a temperature switch that switches off if temperature gets too high.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[ScottyUK]

2kW needs a big heatsink to get rid of it - definitely force cooled, and preferably with a route out of the immediate environment.

Depending on the application, liquid cooling has some merits for getting massive amounts of heat away from small areas where a big fin-fan cooler can dissipate the heat. Lots of negatives with liquid cooling too - things have to get pretty challenging before it becomes viable, which usually means very high power semiconductors which on the whole can't operate much above 150C die temperature for regular silicon. Difficult to imagine a reason for using a liquid cooled resistance bank. Personally I would look for a different resistor type which can operate at much higher temperature. In this instance wire-wound or a punched plate type would probably suit best. Why did you choose thick film types for this application?


----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[electricpete]

Can you explain the application? I'm having a hard time imagining the need to dissipate so much power in resistors.

Is this a steady state condition? If not, of course the picture becomes a little different (less cooling capacity needed depending on the thermal capacitance and the duty cycle).

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[electricpete]

I guess maybe this is a load bank? In that case it seems the others are right you need to bring forced cooling into the picture.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[drillz]

Hello

Thankyou for your replies.
Please may I now be specific as you raise interesting points of concern.
I’m certain I divulge nothing confidential here as its just a resistive load.

The real set-up is 300R of resistance.
This comprises four parallel pairs of 150R resistors.
These resistors are from Welwyn (Manufacturer no = WDBR3-150RKT)
Resistor datasheet is

http://www.welwyn-tt.com/pdf/application_notes/WDBR.pdf

From the power derating curve at the bottom of page 2 of the above dtasheet, it clearly shows that above 210 degree C, the resistors can handle zero watts.

We are using the 150R, 3KW version, it can handle 900W (continuous).

This 300R sits across 2.5KV.
However, it is a pulsed load, and the duty cycle is 10%

During pulse ON, load current = 8.3 Amps. (Power = 21KW)

However, average power is 10% of this, thus = 2.1KW

(Supposing the ON pulse was 2ms.)

The heatsink is from ABL Heatsinks, it is manufacturers no. 173AB3000B

http://www.heatsinksonline.co.uk/Products/100series/page23.htm

http://www.farnell.com/datasheets/56040.pdf

….gives heatsink details.

Heatsink dims are 300mm * 119mm
….With forced air cooling, the lowest the R(th) can be made is 0.144 degree C per Watt.

I calculate that heatsink will get to 2100 * 0.144 = 302 degree C.
…from the resistor datasheet, this will destroy the resistors…..
I would be very grateful for any thoughts pointing out if I have missed something here.

 

[ScottyUK]

Your calculation gives you the temperature rise above ambient. Gunnar has already told you this.

Things you have missed...

- Your heatsink is too small. Buy a bigger one.

- Your resistors are loaded near to their maximum pulse rating. Reliability will suffer.

- This is a bad design. Start again.



----------------------------------

If we learn from our mistakes I'm getting a great education!

 

[LionelHutz]

No, the only thing you have missed is that the temperature you calculated is the temperature rise above ambient.

The heatsink should be ~170C - 40C ambient divided by 2100W or 0.06 degreeC/watt

But, I personally would recommend you run the resistors at about 1/2 of any of their manufacturers listed maximum ratings.

You don't necessarily need a larger heatsink. You could just use more than one of that heatsink you already have.

 

[drillz]

Hi,
Thankyou for getting back. I'll investigate shortly ...but for now, i've again been told that this set-up is going to be OK, despite having raised my concerns.

 

[itsmoked]

Great... Stand way back and use safety glasses.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[Skogsgurra]

drillz,

Who told you it would be OK?

Your accountant?

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[itsmoked]

Skoggs; That was my exact thought also!

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[sreid]

When the resistors melt down, you'll be needing a heatsink like the ones below.

http://www.r-theta.com/products_standards_fabfin_bonded.asp

 

[drillz]

Hi
LionelHutz we have taken your advise and now have two heatsinks.
Thankyou for all.

 

[Skogsgurra]

Glad you did!

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[LionelHutz]

Well, I personally think 2 heatsinks is still marginal but it's certainly better than 1 and may work. Monitor the temperature when you put this load into service and see what you get.

 

[sreid]

From a thermal point of view, it's best to separate the resistors on the heat sink (rather than have them, say, close together at the center of the heat sink). The thermal resistance of the heat sink is usually specified for a distributed thermal load.

This might be in conflict with your electrical requirements since you have selected low inductance resistors.