Maker Pro
Maker Pro

Extending the life of a Tek 2465?

D

DaveC

Jan 1, 1970
0
Where did you get this heatsink? I'd like to get one to add to my U800.

Surplus electronics shop here in Sillycon Valley.
I spoke with a Tek engineer who told me that excessive heat was the
failure mode for U800 and that heatsinking it is a good idea, so I think
you're on the right track.

Intuitively, it just seemed like the right decision.
I know someone else who heatsinked his U800 and hasn't had any problems
so far. The only caveat he said is the studs are at -5V so one has to be
sure that the heatsink doesn't touch the case...

There are little vinyl "nipples" that slip over the studs to insulate them.
Just be sure to put them back on when you've finished the job. Might be
prudent to trim those nipples to accomodate the additional height of the nuts
and heat sink. Not much, just a few mm.

There is a flat cable connector that plugs onto header J102; you can see the
header pins in this photo:

http://home.covad.net/~peninsula/Before.jpg

It's easy to pull the connector off. I didn't notice, and got Counter/Timer
option errors on startup. (If you don't have C/T option, might not have this
connector; don't know...). Also was told that it is common to pull off one of
the coax cables running through this area, which gives similar power-on
diagnostics errors.

This heat sink fits nicely over the J102 connector when properly installed.
Was like it was made for this application :)

Don't overtighten the nuts. You could stress the studs and possibly pull them
from the PCB, which means they might come out and bounce around the circuit
area (!). Not a good thing... :)

Good luck,
 
D

DaveC

Jan 1, 1970
0
On further consideration, I haven't done all that much to properly heat sink
this IC.

Most of the heat is coming from the metal "substrate", the built-in heat-sink
which is visible as a tab under the nut on the right end, and a notch under
the nut on the left end:

http://home.covad.net/~peninsula/Before.jpg

My heat sink contacts the full surface of the IC's plastic body, which is
radiating a minimal amount of heat. There is a #4-40 nut between the tab and
the black heat sink on the right end only. This is all that is conducting
heat from the tab to the sink. On the left end, there is no contact at all
between the metal substrate and the black heat sink, except whatever might be
coincidentally conducted through the stud (which probaly 'aint much):

http://home.covad.net/~peninsula/After.jpg

To do this project right would require some kind of machined spacer for both
ends, one to fit the tab's square area, and the other the notch's. Both would
have to be no greater than the distance from the body's surface to the metal
substrate's surface. Pretty small requirements for machining such parts.

Might cobble something together to do this. Have to see how resources become
available...

In the mean time, I'll keep the scope in Normal sweep mode when it's powered
on, which keeps the display blank. This, as pointed out by someone in this
ng, keeps U400 working less, and hence, cooler, when it's not needed.

Enjoy,
 
J

Jim Yanik

Jan 1, 1970
0
Surplus electronics shop here in Sillycon Valley.


Intuitively, it just seemed like the right decision.


There are little vinyl "nipples" that slip over the studs to insulate
them. Just be sure to put them back on when you've finished the job.
Might be prudent to trim those nipples to accomodate the additional
height of the nuts and heat sink. Not much, just a few mm.

There is a flat cable connector that plugs onto header J102; you can
see the header pins in this photo:

http://home.covad.net/~peninsula/Before.jpg

It's easy to pull the connector off. I didn't notice, and got
Counter/Timer option errors on startup. (If you don't have C/T option,
might not have this connector; don't know...). Also was told that it
is common to pull off one of the coax cables running through this
area, which gives similar power-on diagnostics errors.

If the option(s) are not installed,or does not require a signal
pickoff,there are jumpers that are installed;see your schematic for the
vertical section.
 
J

John Woodgate

Jan 1, 1970
0
To do this project right would require some kind of machined spacer for
both ends,

Cut a strip of thin sheet metal (beer can is good) whose width is the
same as the height of spacer you require, and roll it round the shaft of
a suitable screwdriver. Cut off the surplus. Instant spacer with no
lathe in sight. For a thicker wall, roll more than one turn.
 
D

DaveC

Jan 1, 1970
0
Cut a strip of thin sheet metal (beer can is good) whose width is the
same as the height of spacer you require, and roll it round the shaft of
a suitable screwdriver. Cut off the surplus. Instant spacer with no
lathe in sight. For a thicker wall, roll more than one turn.

Thanks, John. Simple solution. But that's a mighty thin spacer... the
distance from the top of the IC to the top of the tab:

http://home.covad.net/~peninsula/Before.jpg

Maybe just several layers of mashed beer can, w/o the wrapping exercise?
 
J

John Woodgate

Jan 1, 1970
0
Thanks, John. Simple solution. But that's a mighty thin spacer... the
distance from the top of the IC to the top of the tab:

http://home.covad.net/~peninsula/Before.jpg

Maybe just several layers of mashed beer can, w/o the wrapping exercise?

I can't tell what thickness you need from the picture. But I assumed
from what you wrote that it was thicker than a normal washer.
 
D

DaveC

Jan 1, 1970
0
I can't tell what thickness you need from the picture. But I assumed
from what you wrote that it was thicker than a normal washer.

Yes, maybe 3/32 inch (1.2 mm). I put a thin #4-40 nut on top of the tab to
hold it to the PCB, then the new sink, then the original nut and
star-lock-combo you see. I'd rather replace the thin nut (its surfaces are
quite curved; not near the total surface area of the tab) with something
larger and more thermally conductive.

Is it really more helpful, thermally, to take off the anodize from the under
side of the heat sink (as long as it is smooth)? Can steel wool accomplish
this?

Thanks,
 
D

DaveC

Jan 1, 1970
0
Cut a strip of thin sheet metal (beer can is good) whose width is the
same as the height of spacer you require, and roll it round the shaft of
a suitable screwdriver. Cut off the surplus. Instant spacer with no
lathe in sight. For a thicker wall, roll more than one turn.

Great idea. But because of the small area, I'll have to use it in a "z"
shaped form, folded several times, rather than rolled.

Anyone know of a non-food product that comes in aluminum cans? I'd like to
use thin aluminum sheet, but without the need to remove that inner coating
that protects food and drinks from the aluminum.

I could sand off this layer from beer can or such, but then I'm looking at
roughening up the surface, which would 1) reduce the thermal transfer
properties, and 2) make the aluminum thinner.

Ideas?

Thanks,
 
J

John Woodgate

Jan 1, 1970
0
Is it really more helpful, thermally, to take off the anodize from the
under side of the heat sink (as long as it is smooth)? Can steel wool
accomplish this?

No. The anodizing is very thin and doesn't have a particularly poor
thermal conductivity anyway.

If steel wool will get it off, it's paint, not anodizing. Anodizing is a
layer of dyed aluminium oxide, chemically bonded to the aluminium mass.
 
J

John Woodgate

Jan 1, 1970
0
Great idea. But because of the small area, I'll have to use it in a "z"
shaped form, folded several times, rather than rolled.

Now I don't understand what you are trying to do. I thought you just
wanted a spacer. Beer can material isn't good thermally; the aluminium
is in a 'hard' metallurgical state, which reduces its thermal
conductivity. The same applied to kitchen foil.
Anyone know of a non-food product that comes in aluminum cans? I'd like
to use thin aluminum sheet, but without the need to remove that inner
coating that protects food and drinks from the aluminum.

If the thermal conductivity matters, scrounge a tiny bit of soft or
quarter-hard thin sheet from your local metal converter. Or you could
make a ring out of copper wire a bit thicker than you need finally and
file off top and bottom surfaces to give you a better thermal contact.
I could sand off this layer from beer can or such, but then I'm looking
at roughening up the surface, which would 1) reduce the thermal transfer
properties, and 2) make the aluminum thinner.

Forget it.

Incidentally, I make heat sinks for my own stuff and
repairs/improvements out of bits of scrap copper water pipe.
 
D

DaveC

Jan 1, 1970
0
Yes, I want to get the heat from the tab to the bottom of the heat sink, a
distance of about 1.2 mm. Thermal properties count.
Incidentally, I make heat sinks for my own stuff and
repairs/improvements out of bits of scrap copper water pipe.

Copper sounds ideal. I'm sure I could find a piece of copper that thickness.

Thanks,
 
J

Jim Yanik

Jan 1, 1970
0
Yes, I want to get the heat from the tab to the bottom of the heat
sink, a distance of about 1.2 mm. Thermal properties count.


Copper sounds ideal. I'm sure I could find a piece of copper that
thickness.

Thanks,

Looking at the pix on your site,it seems to me you have adequate contact
already.Maybe if you can find a surplus HS a tad longer,sure,go ahead,but I
doubt you'll gain much in dissipation over what you have now.
IMO,a different style HS with fins closed to the center of the IC substrate
would be best,as the HS cross-section(thickness) limits what heat gets
dissipated.
But hey,it's your scope.
 
M

mike

Jan 1, 1970
0
DaveC said:
It is a Thermalloy product:

< http://www.thermalloy.com/bin/stdisp.pl?Pnum=507222b00000>

Although I've now decided it's not long enough. I want one to extend the
entire length of the IC, including the maximum measurement of the metal
substrate.

All this is very entertaining. But you're shooting in the dark.
It's like watching a bunch of drunk guys sitting in a bar talking about
women.
Nobody has a clue, but they all have a technique. ;-)

Why not measure the temperature of the &*&&^%G tab and DESIGN a solution?
You remember...temperature delta, thermal conductivity, path length, air
flow...? Understand the problem before designing the solution!!!

mike

--
Bunch of stuff For Sale and Wanted at the link below.
Toshiba & Compaq LiIon Batteries, Test Equipment
Honda CB-125S
TEK Sampling Sweep Plugin and RM564
Tek 2465 $800, ham radio, 30pS pulser
Tektronix Concept Books, spot welding head...
http://www.geocities.com/SiliconValley/Monitor/4710/
 
D

DaveC

Jan 1, 1970
0
Looking at the pix on your site,it seems to me you have adequate contact
already.Maybe if you can find a surplus HS a tad longer,sure,go ahead,but I
doubt you'll gain much in dissipation over what you have now.
IMO,a different style HS with fins closed to the center of the IC substrate
would be best,as the HS cross-section(thickness) limits what heat gets
dissipated.

The contact between the tab and my heat sink is provided by a #4-40 nut. The
nut is come kind of steel, not real flat on either side. It is about maybe 50
percent of the area of the tab. That's not real good conductivity between the
two. I'd like to make something of aluminum that is closer to the size of the
tab. And I'd like to also provide a similar heat "pipe" for notch at the
other end.
 
D

DaveC

Jan 1, 1970
0
The contact between the tab and my heat sink is provided by a #4-40 nut.
The nut is come kind of steel, not real flat on either side. It is about maybe
50 percent of the area of the tab.

And my heat sink doesn't even cover the nut completely. Looking at the photo,
the heat sink ends before it extends completely under the nut:

http://home.covad.net/~peninsula/After.jpg

The heat sink covers the majority of the IC, but that's mostly plastic, and
the heat is radiated mostly by the two exposed metal parts.

it needs reworking...
 
J

Jim Yanik

Jan 1, 1970
0
And my heat sink doesn't even cover the nut completely. Looking at the
photo, the heat sink ends before it extends completely under the nut:

http://home.covad.net/~peninsula/After.jpg

The heat sink covers the majority of the IC, but that's mostly
plastic, and the heat is radiated mostly by the two exposed metal
parts.

it needs reworking...

ISTR that the metal substrate is exposed on the top side of the IC for most
of the length of the IC.
You should be getting pretty decent thermal contact.
 
M

mike

Jan 1, 1970
0
Jim said:
ISTR that the metal substrate is exposed on the top side of the IC for most
of the length of the IC.
You should be getting pretty decent thermal contact.

Jim, I'm really confused by this.
Are we looking at the same picture?
http://home.covad.net/~peninsula/Before.jpg
If the terms "metal substrate" and "tab" are different terms for
the same heat sink structure of the IC, your statement conflicts with
the picture.
What am I missing?
mike

--
Bunch of stuff For Sale and Wanted at the link below.
Toshiba & Compaq LiIon Batteries, Test Equipment
Honda CB-125S
TEK Sampling Sweep Plugin and RM564
Tek 2465 $800, ham radio, 30pS pulser
Tektronix Concept Books, spot welding head...
http://www.geocities.com/SiliconValley/Monitor/4710/
 
J

Jim Yanik

Jan 1, 1970
0
mike said:
Jim, I'm really confused by this.
Are we looking at the same picture?
http://home.covad.net/~peninsula/Before.jpg
If the terms "metal substrate" and "tab" are different terms for
the same heat sink structure of the IC, your statement conflicts with
the picture.
What am I missing?
mike

OK,I was thinking(perhaps it's the earlier version I recall) that the metal
substrate was exposed for the whole length of the IC,not just the tab.
My fault.
So,what I would do is cut a small square piece of sheet copper(or
aluminum),drill a hole for the stud,place it between the tab and the HS,and
a smaller one for the narrow slot,use thermal grease,too_Or trim a copper
washer to size.
Perhaps an old penny? not the new zinc ones.

Heck,maybe THATS why the IC is prone to failure;they changed the packaging
and thermal dissipation with a full plastic DIP package,instead of the
bared substrate IC as ISTR.
(Of course,being an old geezer,my memory sometimes fails me) :cool:
 
Top