Recommendations for long term thermal conductance?

SteveH

I am not sure what your question or need for more heat sink resistors.  I know you can buy heat sink resistors that have built in heat sinks to supposedly keep them in their heat range.  The resistors in you picture appear to have heat sink fins on them and if they are running over their heat range you may need to install ones with a higher wattage.  I have such a resistor to install in a vintage stereo Sansui 9090db receiver that is known have this resistor fail but have not gotten around to installing it after over 15 years!  I do not put many hours on this one, usually just Christmas time.

I have added extra heat sink to a hot running power transistor in another receiver that had too small heat sink area originally.  I used heat sink from junk computers and JB Welded and bolted the extra heat sink to the existing heat sink.  JB Weld has steel filings as the filler and will conduct heat forever without failing with age like the HS greases.  I do not know how efficient it is at conducting heat but would be much better than regular epoxy.  I also added a cooling fan as this heat problem had turned the printed circuit power board a dark brown.

Charlie

Charlie, Thank you for your reply and your recommendation.  The continuous wattage applied to the pictured resistors will be limited to about 30% of their rated capacity of 100 Watts each.  These resistors are designed to be mounted on external heat sinks even though they do have limited heat dissipation ability without doing so.

The external heat sinks are intended to reduce the resistors' internal operating temperatures to safer levels and prolong their service life by providing additional means of heat dissipation.  Heat transfer from the resistors to the sink is significantly improved by a thermal conducting medium to fill any tiny gaps in between them.

Any additional recommendations about whether thermally conductive silicone tape or thermally conductive epoxy has better long term durability are welcome.

From what I've read, the thermal pads don't have as long a life as a good thermal compound.  I've also not put a ton of thought into this, you may be over-thinking the issue for this application.

Available for < $5 on the big A or take apart an old computer.

@gunrunnerjohn posted:

From what I've read, the thermal pads don't have as long a life as a good thermal compound.  I've also not put a ton of thought into this, you may be over-thinking the issue for this application.

I’d drink to that John, 37 years in the Automotive field, and all I’ve ever used is good old thermal compound…..I’d think not having any compound or sink of some sorts would tell on you…..fixed a many of heat sinked components that had none at all that led to the demise of the component

Pat

SteveH

Another way to reduce the heat from the finned 100 watt resistors might be to half the resistance of the finned resistor and install two of them on the heat sink and wire them in parallel.

Charlie

Thanks RoyBoy    It should read "series" not parallel.

Charlie

@Choo Choo Charlie posted:

SteveH

Another way to reduce the heat from the finned 100 watt resistors might be to half the resistance of the finned resistor and install two of them on the heat sink and wire them in parallel.

Charlie

You have to double the individual resistance if you put them in parallel.

If you halve the individual resistances, you put them in series.

@Choo Choo Charlie posted:

SteveH

Another way to reduce the heat from the finned 100 watt resistors might be to half the resistance of the finned resistor and install two of them on the heat sink and wire them in parallel.

Charlie

That's essentially the plan, a series/parallel arrangement to reduce the wattage / heat in each resistor to an acceptable level.

Parallel: 1/RT = 1/R1 + 1/R2 + .... + 1/Rn

Series: RT = R1 + R2 + .... + Rn

Wattage across each resistor is calculated using Ohms law V=I x R and Watts Law W = V x I.  In parallel circuits, current (I) is divided proportionally. Voltage is the amount drop of across each resistor.

However, the total amount of heat (measured in Watts) generated at a given voltage and current across the resistor array is essentially the same.  That heat still needs to be safely dissipated with external means.

My question is about thermal conducting mediums.  I appreciate the help and suggestions.

Does anyone have product suggestions for long term thermal conducting materials: paste, etc?

@SteveH posted:

Does anyone have product suggestions for long term thermal conducting materials: paste, etc?

You'll likely have to go trolling for that answer.  I'd check with electronics sites for those kinds of answers.  I have to be honest, I haven't ever had the need for "long term" thermal paste quotes.   Stuff like computer builds I always figured after six-eight years it would be unlikely if they're still in use.

@Choo Choo Charlie posted:

SteveH

I am not sure what your question or need for more heat sink resistors.  I know you can buy heat sink resistors that have built in heat sinks to supposedly keep them in their heat range.  The resistors in you picture appear to have heat sink fins on them and if they are running over their heat range you may need to install ones with a higher wattage.  I have such a resistor to install in a vintage stereo Sansui 9090db receiver that is known have this resistor fail but have not gotten around to installing it after over 15 years!  I do not put many hours on this one, usually just Christmas time.

I have added extra heat sink to a hot running power transistor in another receiver that had too small heat sink area originally.  I used heat sink from junk computers and JB Welded and bolted the extra heat sink to the existing heat sink.  JB Weld has steel filings as the filler and will conduct heat forever without failing with age like the HS greases.  I do not know how efficient it is at conducting heat but would be much better than regular epoxy.  I also added a cooling fan as this heat problem had turned the printed circuit power board a dark brown.

Charlie

9090DB - A great amp! Repaired many - over the years. Some, of the newer compounds do not have issue - in my military electronic days - we would use those resistors for electronic load banks. I would use the large ohmite resistors

@RoyBoy posted:

You have to double the individual resistance if you put them in parallel.

If you halve the individual resistances, you put them in series.

Plus one what RoyBoy said.

Steve.  web site  "TOMS HARDWARE " did a comparison of thermal compounds a few years ago and a more recent update of the comparisons. At the bottom of the page is a jpg chart of all they tested with the best at the top.          j

https://www.tomshardware.com/b...s/best-thermal-paste

What I have seen used in recent years is a Thermal paste called Shin Microsi.  You will find it preprinted on most high end computer heat sinks.  It is Grey and at room temp it has the consistency of peanut butter.

Usually the heat sink paste is intended to fill the micro pores of each mating surface, so a very thin coat of thermal material between the surfaces is best.   

I would recommend that you first polish the contact surfaces with 0000 steel wool down to a mirror surface  then apply a thin coat of Shin Microsi or one of the "silver" based Thermal paste products.  A mechanical clamping mechanism would be necessary as well, similar to what is used on computer processor sockets.

Or could you use  an active load instead of resistors and reuse that power elsewhere.

@JohnActon and @aussteve thank you for sharing the information about thermal conducting compunds.  I appreciate the recommendations.

Just to second the recommendations from tomshardware. Been overclocking CPUs and GPUs for decades with thermal paste.

The only thermal death I experienced was a GPU that used tape instead of paste.

Some systems have gone beyond 5 years with no maintenance or replacement of heatsinks or paste. Typical temperature range was 80-85c/175-185f. Usually the paste outlasts the viability of the components to run current programs/OS.