Elecraft › [HAM]

OT: heat sinking resistors

Classic

List

Threaded

12 messages Options

Vic K2VCO

OT: heat sinking resistors

I apologize in advance, but the smartest guys I know are here.

I just received a few Ohmite thick film power resistors for an attenuator I'm making,
rated 20 watts. They are 15mm x 10mm x 3mm. Obviously they have to be heat-sunk
(heat-sinked?) if they are going to dissipate that much power.

How do you do this? There's no hole in the middle...I can just clamp them between a couple
of pieces of aluminum, but is there a right way to accomplish this?
--
Vic
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Craig Smith

Re: OT: heat sinking resistors

Hi Vic ...

My first inclination would be to go to the Ohmite website and look at the
datasheet for the resistors and search for any application notes. One or
both of these sources should address, at least in relative terms, how much
heat is typically removed via conduction through the leads vs. from the body
of the resistor. At these power levels, thermal considerations must have
been part of the mechanical package design of the device. If you are lucky,
you will find something like a graph of thermal resistance through the leads
vs. PCB trace area, airflow or some other variable under your control. Good
luck!

... Craig AC0DS

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Craig Smith

Re: OT: heat sinking resistors

In reply to this post by Vic K2VCO

Another thought, Vic ...

If the approach I suggested earlier results in a large, cumbersome or
expensive packaging job for you, the easiest thing may be to redesign the
attenuator using a larger number of lower wattage resistors. This would
distribute the heat better and could result in a smaller final package with
a lot less effort on your part.

73 Craig AC0DS

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Dave KQ3T

Re: OT: heat sinking resistors

In reply to this post by Craig Smith

Here's the Ohmite resistor guide web page:
http://www.ohmite.com/resguide.html

73,
Dave KQ3T

Craig D. Smith wrote:

> Hi Vic ...
>
> My first inclination would be to go to the Ohmite website and look at the
> datasheet for the resistors and search for any application notes. One or
> both of these sources should address, at least in relative terms, how much
> heat is typically removed via conduction through the leads vs. from the body
> of the resistor. At these power levels, thermal considerations must have
> been part of the mechanical package design of the device. If you are lucky,
> you will find something like a graph of thermal resistance through the leads
> vs. PCB trace area, airflow or some other variable under your control. Good
> luck!
>
> ... Craig AC0DS
>
>

______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Mark Bayern

Re: OT: heat sinking resistors

In reply to this post by Vic K2VCO

> How do you do this? There's no hole in the middle...I can just clamp them between a couple

... lets see ... a 10x15x3mm Ohmite thick film rated at 20 watts ...
with a little web research you get:

<http://www.ohmite.com/catalog/pdf/tah.pdf>

According to that data sheet, "Mounting: Requires the use of a snap-on
style heat sink. A thermal compound should be properly applied."

Mark AD5SS

On Fri, May 21, 2010 at 3:27 PM, Vic Rosenthal <[hidden email]> wrote:

> I apologize in advance, but the smartest guys I know are here.
>
> I just received a few Ohmite thick film power resistors for an attenuator I'm making,
> rated 20 watts. They are 15mm x 10mm x 3mm. Obviously they have to be heat-sunk
> (heat-sinked?) if they are going to dissipate that much power.
>
> How do you do this? There's no hole in the middle...I can just clamp them between a couple
> of pieces of aluminum, but is there a right way to accomplish this?
> --
> Vic
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>

Phil Salas

Re: OT: heat sinking resistors

In reply to this post by Vic K2VCO

"I just received a few Ohmite thick film power resistors for an attenuator I'm making,
rated 20 watts. They are 15mm x 10mm x 3mm. Obviously they have to be heat-sunk
(heat-sinked?) if they are going to dissipate that much power."

FYI - Both Ohmite and Caddock make versions of these resistors that have screw holes for mounting. The lower power parts (15-20 watts) are pretty good thru 2-meters. However, the higher power versions (especially the 100-watt versions) can start to look bad at 50 MHz due to the large package capacitance-to-ground.

Phil - AD5X
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html

Don Wilhelm-4

Re: OT: heat sinking resistors

Phil,

Is this recent data? Has something changed? I have had plots of the
Caddock 50 ohm 100 watt resistors taken to 250 MHz and the return loss
indicated that they made very good attenuators and dummy loads at that
frequency - assuming they were mounted with zero length leads directly
to a BNC or SO-239 connector. My data is 2 years old.

BTW - the facts I have quoted are good only for the 50 ohm Caddock
resistors - I have found that other values were "as stated"
non-inductive, but they certainly were capacitive. A capacitive dummy
load is just as bad as an inductive load.

73,
Don W3FPR

Phil & Debbie Salas wrote:

> "I just received a few Ohmite thick film power resistors for an attenuator I'm making,
> rated 20 watts. They are 15mm x 10mm x 3mm. Obviously they have to be heat-sunk
> (heat-sinked?) if they are going to dissipate that much power."
>
> FYI - Both Ohmite and Caddock make versions of these resistors that have screw holes for mounting. The lower power parts (15-20 watts) are pretty good thru 2-meters. However, the higher power versions (especially the 100-watt versions) can start to look bad at 50 MHz due to the large package capacitance-to-ground.
>
> Phil - AD5X
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>

Phil Salas

Re: OT: heat sinking resistors

If they are mounted against a ground-plane (which is often the case for
heat-sinking), I've found that the 100-watt resistors are not too good above
50 MHz. The 30 watt resistors are better, but the 15 watt resistors are
best thru 2-meters. I built a series of 50 ohm terminations where the
resistors were bolted to a brass plate, and the brass plate was attached to
a N-connector. The resistor leads were about 0.2" long (one pin attached to
the connector center pin, the other soldered to ground). Measured with an
AIM4170 up to 180 MHz.

Phil - AD5X
----- Original Message -----
From: "Don Wilhelm" <[hidden email]>
To: "Phil & Debbie Salas" <[hidden email]>
Cc: <[hidden email]>
Sent: Friday, May 21, 2010 5:50 PM
Subject: Re: [Elecraft] OT: heat sinking resistors

> Phil,
>
> Is this recent data? Has something changed? I have had plots of the
> Caddock 50 ohm 100 watt resistors taken to 250 MHz and the return loss
> indicated that they made very good attenuators and dummy loads at that
> frequency - assuming they were mounted with zero length leads directly to
> a BNC or SO-239 connector. My data is 2 years old.
>
> BTW - the facts I have quoted are good only for the 50 ohm Caddock
> resistors - I have found that other values were "as stated" non-inductive,
> but they certainly were capacitive. A capacitive dummy load is just as
> bad as an inductive load.
>
> 73,
> Don W3FPR
>
> Phil & Debbie Salas wrote:
>> "I just received a few Ohmite thick film power resistors for an
>> attenuator I'm making, rated 20 watts. They are 15mm x 10mm x 3mm.
>> Obviously they have to be heat-sunk (heat-sinked?) if they are going to
>> dissipate that much power."
>>
>> FYI - Both Ohmite and Caddock make versions of these resistors that have
>> screw holes for mounting. The lower power parts (15-20 watts) are pretty
>> good thru 2-meters. However, the higher power versions (especially the
>> 100-watt versions) can start to look bad at 50 MHz due to the large
>> package capacitance-to-ground.
>>
>> Phil - AD5X
>> ______________________________________________________________
>> Elecraft mailing list
>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>> Help: http://mailman.qth.net/mmfaq.htm
>> Post: mailto:[hidden email]
>>
>> This list hosted by: http://www.qsl.net
>> Please help support this email list: http://www.qsl.net/donate.html
>>
>>
>

Don Wilhelm-4

Re: OT: heat sinking resistors

Phil,

Thanks for that information. Mine are mounted to a heat-sink culled
from a CPU cooler, and the connector is mounted to a piece of aluminum
angle also affixed to the heat sink. This assembly is standalone and
does not connect to any other ground plane. Although the heat sink may
constitute a small ground plane, the fact that it is isolated may
account for the difference in the measurement results.

BTW, I have found Caddock to be more consistent than Ohmite. Caddock
advertizes their 50 ohm thick film 1% resistors as non-reactive, and the
last time I looked, Ohmite did not make that statement.

73,
Don W3FPR

Phil & Debbie Salas wrote:

> If they are mounted against a ground-plane (which is often the case for
> heat-sinking), I've found that the 100-watt resistors are not too good above
> 50 MHz. The 30 watt resistors are better, but the 15 watt resistors are
> best thru 2-meters. I built a series of 50 ohm terminations where the
> resistors were bolted to a brass plate, and the brass plate was attached to
> a N-connector. The resistor leads were about 0.2" long (one pin attached to
> the connector center pin, the other soldered to ground). Measured with an
> AIM4170 up to 180 MHz.
>
> Phil - AD5X
>
>

Geoffrey Mackenzie-Kennedy

Re: OT: heat sinking resistors

In reply to this post by Vic K2VCO

Vic,

I would suggest that you do not clamp each resistor between a couple of
pieces of aluminum, because some types of uncapsulated thick film power
resistors are quite fragile. They may not appear to break if stressed, but
there is the risk of creating a hairline fracture across the resistor
"element".

For the two power attenuators that I use in my Rx IMD test setup before the
combiner (precision attenuators not required in this part) I used 20 watt
thick film power resistors in TO-220 packages, with their individual
heatsinks sitting upright, well separated from any grounded metalwork.

73,
Geoff
GM4ESD

Vic Rosenthal wrote on Friday, May 21, 2010 at 9:27 PM:

> I just received a few Ohmite thick film power resistors for an attenuator
> I'm making,
> rated 20 watts. They are 15mm x 10mm x 3mm. Obviously they have to be
> heat-sunk
> (heat-sinked?) if they are going to dissipate that much power.
>
> How do you do this? There's no hole in the middle...I can just clamp them
> between a couple
> of pieces of aluminum, but is there a right way to accomplish this?
> --
> Vic

W8JI

Re: OT: heat sinking resistors

I've used those resistors here and they are fine up to at
least 30 MHz sandwiched.

They will not crack if he uses common sense on materials.
The heatsink should be rigid and very flat, and the clamping
metal preloaded with some sort of tensioning or spring
washers. A couple split ring washers in series at opposite
ends (one under head, one under nut) tightened to just
collapsing would be good. If someone is really worried about
it they could put fish paper or a layer of thin Teflon
between the holding plate and the resistor, so long as the
tension washer stack can keep pressure steady. There are
always Belleville washers which are best, but simple spring
washers (split ring) would work in a non-critical system
like this.

The only time anyone would get into trouble is if they use
improper hardware, like stainless screws and a rigid or very
high pressure lock washer like a star or internal tooth
type.

There will be very little difference in performance with a
heatsink floating in air, or with the resistor attached to a
heatsink that is massive and grounded. Whatever is
mechanically best will work OK.

Anyone without good mechanical sense needs to stay away from
clamping things.

73 Tom

----- Original Message -----
From: "Geoffrey Mackenzie-Kennedy" <[hidden email]>
To: "Vic Rosenthal" <[hidden email]>
Cc: "Elecraft Discussion List" <[hidden email]>
Sent: Friday, May 21, 2010 8:21 PM
Subject: Re: [Elecraft] OT: heat sinking resistors

> Vic,
>
> I would suggest that you do not clamp each resistor
> between a couple of
> pieces of aluminum, because some types of uncapsulated
> thick film power
> resistors are quite fragile. They may not appear to break
> if stressed, but
> there is the risk of creating a hairline fracture across
> the resistor
> "element".
>
> For the two power attenuators that I use in my Rx IMD test
> setup before the
> combiner (precision attenuators not required in this part)
> I used 20 watt
> thick film power resistors in TO-220 packages, with their
> individual
> heatsinks sitting upright, well separated from any
> grounded metalwork.
>
> 73,
> Geoff
> GM4ESD
>
>
> Vic Rosenthal wrote on Friday, May 21, 2010 at 9:27 PM:
>
>> I just received a few Ohmite thick film power resistors
>> for an attenuator
>> I'm making,
>> rated 20 watts. They are 15mm x 10mm x 3mm. Obviously
>> they have to be
>> heat-sunk
>> (heat-sinked?) if they are going to dissipate that much
>> power.
>>
>> How do you do this? There's no hole in the middle...I can
>> just clamp them
>> between a couple
>> of pieces of aluminum, but is there a right way to
>> accomplish this?
>> --
>> Vic
>
>
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:[hidden email]
>
> This list hosted by: http://www.qsl.net
> Please help support this email list:
> http://www.qsl.net/donate.html

gm3sek

Re: OT: heat sinking resistors

In reply to this post by Don Wilhelm-4

Don Wilhelm wrote:

>Phil & Debbie Salas wrote:
>> "I just received a few Ohmite thick film power resistors for an
>>attenuator I'm making,
>> rated 20 watts. They are 15mm x 10mm x 3mm. Obviously they have to be
>>heat-sunk
>> (heat-sinked?) if they are going to dissipate that much power."
>>
>> FYI - Both Ohmite and Caddock make versions of these resistors that
>>have screw holes for mounting. The lower power parts (15-20 watts)
>>are pretty good thru 2-meters. However, the higher power versions
>>(especially the 100-watt versions) can start to look bad at 50 MHz due
>>to the large package capacitance-to-ground.
>>
>> Phil - AD5X

The two of you may be applying different standards for VSWR performance.

If you wish the dummy load to be an accurate 50-ohm reference for
measurement purposes, then 1-2pF of shunt capacitance can be quite
serious. At 145MHz, 2pF of shunt capacitance would increase the VSWR to
about 1.1; as a 50-ohm reference, that would be "quite poor".

On the other hand, if you only need somewhere to dump unwanted RF power,
that same VSWR of 1.1 will be "just fine".

There are some examples on my 'In Practice' reference page:
http://www.ifwtech.co.uk/g3sek/in-prac#0908

Whatever your requirements, you can estimate the minimum VSWR quite
accurately by assuming 1-1.5pF of shunt capacitance for each TO-220
package. In practical layouts, the VSWR at VHF will almost always be
dominated by the series inductance of your own connecting wires.

--

73 from Ian GM3SEK
http://www.ifwtech.co.uk/g3sek
______________________________________________________________
Elecraft mailing list
Home: http://mailman.qth.net/mailman/listinfo/elecraft
Help: http://mailman.qth.net/mmfaq.htm
Post: mailto:[hidden email]

This list hosted by: http://www.qsl.net
Please help support this email list: http://www.qsl.net/donate.html