Question about heatsinks

[Noah]

Some heatsinks say the are 5 C/W or 1 C/W degrees and other. What does
this mean?
How do I calculate how much power heatsinks can handle until they are far
too hot? I bought a 5 C/W heatsink and a TO-3 transistor and it was
dissapating about 65 watts and it was far too hot and had to use another
transistor and heatsink.

Thank you.

--

Noah
Linux. The leading OS of the future.

 

[Mike Harding]

On Fri, 29 Aug 2003 21:40:07 GMT, Noah <noah@susie.linux.au> wrote:

Some heatsinks say the are 5 C/W or 1 C/W degrees and other. What does
this mean?

It's a measure of the heatsink's "resistance" to heat. A 5 C/W will
not cool as well a 1 C/W

How do I calculate how much power heatsinks can handle until they are far
too hot?

The heatsink won't become too hot - but the device attached
to it which you are trying to cool may well do so!

I bought a 5 C/W heatsink and a TO-3 transistor and it was
dissapating about 65 watts and it was far too hot and had to use another
transistor and heatsink.

65W is a _lot_ of power to dissipate from a single T03 device,
you may need to look at a fan and a BIG heatsink.

I posted an account of my issues with a high power design on
this NG a few weeks ago - search Google for a thread:

"TO220 heatsink - conductive/insulating pads"

which may give you some clues and tips, also check
out the following links:

http://www.heatsinksonline.co.uk/home.htm
http://www.aavidthermalloy.com/
http://www.imiheatsinks.co.uk/

What are you making?

Mike Harding

 

[Noah]

On Sat, 30 Aug 2003 08:23:59 +1000, Mike Harding wrote:

On Fri, 29 Aug 2003 21:40:07 GMT, Noah <noah@susie.linux.au> wrote:

snip

What are you making?

Thanks Mike. I made a 5A 13.8V power supply using MJ2955 transistors. I
added the second transistor and heatsink then the temps dropped a lot. I
tested the PS pulling about 4.5 amps and it worked very well. If I didn't
add the second transistor the thing would have cooked. I still should get
a bigger heatsink for the bridge rectifier because it is running very hot.

--

Noah
Linux. The leading OS of the future.

 

[Noah]

On Sat, 30 Aug 2003 09:04:40 +1000, Phil Allison wrote:

** How come there was 65 watts dissipation ???

Should be no more than 35.



........... Phil

Dont P = VxA ?

--

Noah
Linux. The leading OS of the future.

 

[Noah]

On Sat, 30 Aug 2003 12:25:53 +1000, Phil Allison wrote:

"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.08.29.23.24.01.356328@susie.linux.au...
On Sat, 30 Aug 2003 09:04:40 +1000, Phil Allison wrote:


** How come there was 65 watts dissipation ???

Should be no more than 35.



........... Phil

Dont P = VxA ?


** Dissipation in a regulator is: I out x ( V in -Vout )

Thanks for that info Phil, but I was actually talking about using the
bypass transistors and not a 5A regulator. Yes I guess using a 5A regulator
would dissapate around 35W if used. How is it I got 65W with the power
transistors? I assume using 2 power transistors each would dissapate
around 30W each at 5A?

--

Noah
Linux. The leading OS of the future.

 

[Brian Goldsmith]

"Noah" <noah@susie.linux.au> wrote

"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.08.29.23.24.01.356328@susie.linux.au...
On Sat, 30 Aug 2003 09:04:40 +1000, Phil Allison wrote:


** How come there was 65 watts dissipation ???

Should be no more than 35.



........... Phil

Dont P = VxA ?


** Dissipation in a regulator is: I out x ( V in -Vout )

Thanks for that info Phil, but I was actually talking about using the
bypass transistors and not a 5A regulator.


** The "bypass" transistor(s) IS the regulating element and is driven by
the Voltage regulator.The transistor is called a pass device.Phils rule
of I out x (Vin -V out) still applies.

Brian Goldsmith.

 

[Noah]

On Sat, 30 Aug 2003 08:37:39 +0000, Brian Goldsmith wrote:

** The "bypass" transistor(s) IS the regulating element and is driven by
the Voltage regulator.The transistor is called a pass device.Phils rule
of I out x (Vin -V out) still applies.

So using P=VxA doesn't?

--

Noah
Linux. The leading OS of the future.

 

[Mike Harding]

On Sat, 30 Aug 2003 08:50:35 GMT, Noah <noah@susie.linux.au> wrote:

On Sat, 30 Aug 2003 08:37:39 +0000, Brian Goldsmith wrote:


** The "bypass" transistor(s) IS the regulating element and is driven by
the Voltage regulator.The transistor is called a pass device.Phils rule
of I out x (Vin -V out) still applies.

So using P=VxA doesn't?

Yes it does (for resistive loads). I'm not sure where Phil
gets his maximum of 35W from?

The power dissipated in your supply will be dependent
upon your supply voltage. So if you have (say) an 18V
transformer it will give you about 25V DC with a decent
size reservoir capacitor. Therefore if your output is 13V8
the pass transistor will have to drop about 11V so at 5A
current it will be dissipating 55W.

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan

Mike Harding

 

[Phil Allison]

"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.08.30.08.52.19.750257@susie.linux.au...

On Sat, 30 Aug 2003 08:37:39 +0000, Brian Goldsmith wrote:


** The "bypass" transistor(s) IS the regulating element and is driven by
the Voltage regulator.The transistor is called a pass device.Phils rule
of I out x (Vin -V out) still applies.


So using P=VxA doesn't?

** Never did - never could.



............ Phil

 

[Phil Allison]

"Mike Harding" <mike_harding1@hotmail.com> wrote in message
news:l0r0lv48qijrovm6llr7h7e7oj6rsq7vgd@4ax.com...

On Sat, 30 Aug 2003 08:50:35 GMT, Noah <noah@susie.linux.au> wrote:

On Sat, 30 Aug 2003 08:37:39 +0000, Brian Goldsmith wrote:


** The "bypass" transistor(s) IS the regulating element and is driven
by
the Voltage regulator.The transistor is called a pass device.Phils rule
of I out x (Vin -V out) still applies.

So using P=VxA doesn't?

Yes it does (for resistive loads). I'm not sure where Phil
gets his maximum of 35W from?

** WRONG.

Output power is not dissipated power in the heatsink.

The power dissipated in your supply will be dependent
upon your supply voltage. So if you have (say) an 18V
transformer it will give you about 25V DC with a decent
size reservoir capacitor.

** Huh - what a bad design.



Therefore if your output is 13V8

the pass transistor will have to drop about 11V so at 5A
current it will be dissipating 55W.

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan

** Yep, a 13 volt AC tranny is more like it.

A 6 volt DC in / out differential at full load is enough - gives 30
watts dissipation.



.............. Phil

 

[Kevin Ettery]

"Phil Allison" <philallison@optusnet.com.au> wrote in message
news:3f508fe1$0$6526$afc38c87@news.optusnet.com.au...

"Mike Harding" <mike_harding1@hotmail.com> wrote in message
news:l0r0lv48qijrovm6llr7h7e7oj6rsq7vgd@4ax.com...
On Sat, 30 Aug 2003 08:50:35 GMT, Noah <noah@susie.linux.au> wrote:

On Sat, 30 Aug 2003 08:37:39 +0000, Brian Goldsmith wrote:


** The "bypass" transistor(s) IS the regulating element and is driven
by
the Voltage regulator.The transistor is called a pass device.Phils
rule
of I out x (Vin -V out) still applies.

So using P=VxA doesn't?

Yes it does (for resistive loads). I'm not sure where Phil
gets his maximum of 35W from?


** WRONG.

Output power is not dissipated power in the heatsink.

Phil, it isn't output power they're talking about, its the power that needs
to be disipated by the regulating device - P always equals Volts x Current
(for DC anyway - extra terms are included in AC calcs). So volts drop
accross regulator x current through regulator will give power the heatsink
needs to be able to dissipate.

The power dissipated in your supply will be dependent
upon your supply voltage. So if you have (say) an 18V
transformer it will give you about 25V DC with a decent
size reservoir capacitor.


** Huh - what a bad design.



Therefore if your output is 13V8
the pass transistor will have to drop about 11V so at 5A
current it will be dissipating 55W.

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan



** Yep, a 13 volt AC tranny is more like it.

Agreed. Should always size the tranny so that, at full load, it can supply
slightly more than the required voltage.

A 6 volt DC in / out differential at full load is enough - gives
30
watts dissipation.



............. Phil

 

[Phil Allison]

"Kevin Ettery" <kpettery@dcsi.net.au> wrote in message
news:3f515af9@news.leadingedgeinternet.net.au...

"Phil Allison" <philallison@optusnet.com.au> wrote in message
news:3f508fe1$0$6526$afc38c87@news.optusnet.com.au...

"Mike Harding" <mike_harding1@hotmail.com> wrote in message
news:l0r0lv48qijrovm6llr7h7e7oj6rsq7vgd@4ax.com...
On Sat, 30 Aug 2003 08:50:35 GMT, Noah <noah@susie.linux.au> wrote:

On Sat, 30 Aug 2003 08:37:39 +0000, Brian Goldsmith wrote:


** The "bypass" transistor(s) IS the regulating element and is
driven
by
the Voltage regulator.The transistor is called a pass device.Phils
rule
of I out x (Vin -V out) still applies.

So using P=VxA doesn't?

Yes it does (for resistive loads). I'm not sure where Phil
gets his maximum of 35W from?


** WRONG.

Output power is not dissipated power in the heatsink.

Phil, it isn't output power they're talking about,

** You don't say - dickhead.



.............. Phil

 

[Noah]

On Mon, 01 Sep 2003 18:41:56 +1000, Phil Allison wrote:

In any case, Phil's estimate of 35W is based on an *assumption* about the
characteristics
of your transformer and should therefore be ignored.


** Well fuck you Mr Wog Shithead.

While I appreciate your help Phil, I don't think that is nessesary.

--

Noah
Linux. The leading OS of the future.

 

[Phil Allison]

"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.09.01.09.38.50.330706@susie.linux.au...

On Mon, 01 Sep 2003 18:41:56 +1000, Phil Allison wrote:

In any case, Phil's estimate of 35W is based on an *assumption* about
the
characteristics of your transformer and should therefore be ignored.


** Well fuck you Mr Wog Shithead.


While I appreciate your help Phil, I don't think that is nessesary.

** How the ***HELL *** would you know ??????????

Zabkar ( a moronic VCR tech) has been on my case for ages.

He reaps what he sows.




............ Phil

 

[Noah]

On Sat, 30 Aug 2003 21:52:18 +1000, Phil Allison wrote:

<snipped>

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan



** Yep, a 13 volt AC tranny is more like it.

A 6 volt DC in / out differential at full load is enough - gives 30
watts dissipation.

Here is what I have. I have a multitapped DSE transformer which is 5A @
12V and I have it wired to give 15V AC at the moment which gives around 21
volts DC.
Should I wire it to give 12V AC and then that will give about 16.9V DC?
Would this be alright for my 13.8V supply?
Thank you

--

Noah
Linux. The leading OS of the future.

 

[Phil Allison]

"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.09.02.06.40.43.983377@susie.linux.au...

On Sat, 30 Aug 2003 21:52:18 +1000, Phil Allison wrote:

snipped

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan



** Yep, a 13 volt AC tranny is more like it.

A 6 volt DC in / out differential at full load is enough - gives
30
watts dissipation.


Here is what I have. I have a multitapped DSE transformer which is 5A @
12V and I have it wired to give 15V AC at the moment which gives around 21
volts DC.

** With NO load - right ??

Should I wire it to give 12V AC and then that will give about 16.9V DC?
Would this be alright for my 13.8V supply?

** Under load the DC voltage will drop at least 25% plus there is a large
ripple voltage at 5 amps unless a LOT of filtering capacitance is used (
like about 20,000 uF) . I suggest you try using an LM338 K regulator IC -
Jaycar have them. That 60 VA DSE tranny is not big enough for continuous
useage with 80 - 90 watts of DC pre- regulator.




.............. Phil

 

[Mike Harding]

On Tue, 02 Sep 2003 06:39:34 GMT, Noah <noah@susie.linux.au> wrote:

On Sat, 30 Aug 2003 21:52:18 +1000, Phil Allison wrote:

snipped

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan



** Yep, a 13 volt AC tranny is more like it.

A 6 volt DC in / out differential at full load is enough - gives 30
watts dissipation.

Here is what I have. I have a multitapped DSE transformer which is 5A @
12V and I have it wired to give 15V AC at the moment which gives around 21
volts DC.
Should I wire it to give 12V AC and then that will give about 16.9V DC?
Would this be alright for my 13.8V supply?

Give it a try you can't do any damage. The lower input
voltage you can produce the less heat produced. Your
problem may be that at full load the transformer drops
off too much to maintain regulation at the output - or it
may not depends on the transformer specs and even
if your 13V8 falls to 12V5 it may not matter anyway -
depends on your load.

Mike Harding

 

[Noah]

On Tue, 02 Sep 2003 17:25:08 +1000, Phil Allison wrote:

"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.09.02.06.40.43.983377@susie.linux.au...
On Sat, 30 Aug 2003 21:52:18 +1000, Phil Allison wrote:

snipped

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan



** Yep, a 13 volt AC tranny is more like it.

A 6 volt DC in / out differential at full load is enough - gives
30
watts dissipation.


Here is what I have. I have a multitapped DSE transformer which is 5A @
12V and I have it wired to give 15V AC at the moment which gives around 21
volts DC.


** With NO load - right ??


Should I wire it to give 12V AC and then that will give about 16.9V DC?
Would this be alright for my 13.8V supply?


** Under load the DC voltage will drop at least 25% plus there is a large
ripple voltage at 5 amps unless a LOT of filtering capacitance is used (
like about 20,000 uF) .

I am using a total of 17,600 in caps.

I suggest you try using an LM338 K regulator IC -
Jaycar have them.

I didn't use that because it costs a fair bit of dough.

That 60 VA DSE tranny is not big enough for continuous
useage with 80 - 90 watts of DC pre- regulator.

Ahh. When they say it can output 5A at 12V like it says, this should mean
MAX continuous right? Or if I used a LM338 5A regulator, how would this
make a great difference? Maybe I should have but would it need a huge
heatsink?

--

Noah
Linux. The leading OS of the future.

 

[Noah]

On Tue, 02 Sep 2003 17:25:30 +1000, Mike Harding wrote:

On Tue, 02 Sep 2003 06:39:34 GMT, Noah <noah@susie.linux.au> wrote:

On Sat, 30 Aug 2003 21:52:18 +1000, Phil Allison wrote:

snipped

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan



** Yep, a 13 volt AC tranny is more like it.

A 6 volt DC in / out differential at full load is enough - gives 30
watts dissipation.

Here is what I have. I have a multitapped DSE transformer which is 5A @
12V and I have it wired to give 15V AC at the moment which gives around 21
volts DC.
Should I wire it to give 12V AC and then that will give about 16.9V DC?
Would this be alright for my 13.8V supply?

Give it a try you can't do any damage. The lower input
voltage you can produce the less heat produced. Your
problem may be that at full load the transformer drops
off too much to maintain regulation at the output

That's what I worry about.

- or it

may not depends on the transformer specs and even
if your 13V8 falls to 12V5 it may not matter anyway -
depends on your load.

Could that damage the regulator?

--

Noah
Linux. The leading OS of the future.

 

[Phil Allison]

"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.09.02.07.43.04.775315@susie.linux.au...

On Tue, 02 Sep 2003 17:25:08 +1000, Phil Allison wrote:


"Noah" <noah@susie.linux.au> wrote in message
newsan.2003.09.02.06.40.43.983377@susie.linux.au...
On Sat, 30 Aug 2003 21:52:18 +1000, Phil Allison wrote:

snipped

Choose a transformer which (when all tolerances are
taken into account) will produce DC of (say) 16V for
a 13V8 supply. Or use a big heatsink and fan



** Yep, a 13 volt AC tranny is more like it.

A 6 volt DC in / out differential at full load is enough -
gives
30
watts dissipation.


Here is what I have. I have a multitapped DSE transformer which is 5A @
12V and I have it wired to give 15V AC at the moment which gives around
21
volts DC.


** With NO load - right ??


Should I wire it to give 12V AC and then that will give about 16.9V DC?
Would this be alright for my 13.8V supply?


** Under load the DC voltage will drop at least 25% plus there is a
large
ripple voltage at 5 amps unless a LOT of filtering capacitance is used (
like about 20,000 uF) .

I am using a total of 17,600 in caps.

I suggest you try using an LM338 K regulator IC -
Jaycar have them.

I didn't use that because it costs a fair bit of dough.

That 60 VA DSE tranny is not big enough for continuous
useage with 80 - 90 watts of DC pre- regulator.


Ahh. When they say it can output 5A at 12V like it says, this should mean
MAX continuous right?

** The DC output ( rectifier and cap) is always less than the VA rating -
by 20 % or more.

So you can get 50 watts DC continuous - not 80 or 90 from a 60 VA
tranny.



Or if I used a LM338 5A regulator, how would this

make a great difference?

** Low drop out voltage = 2.7 volts at 5 amps, typically.




............ Phil