Why is the LM317 TO-3 steel case so expensive?

I needed some LM317's in a TO3 case for a project. I went to My favorite vendor and was shocked
to see the LM317 in TO3 steel case for $37.50us ea! I also found it in an aluminum case for $3.25ea.
Anybody know why the steel case is so expensive. What application would require a steel case that
aluminum wouldn't work?  I also compared prices with several other vendors and it was the same.

Just curious. Anybody know?

RonL

To-3 are becoming more of an obsolete package.

IIRC Motorola, last US mfr of steel TO-3 transistors, sold their Mexico factory to mexican company over a decade ago.

I suspect most TO-3 business is not for new designs but repair parts, crazy expensive military, etc.

WRT 317 specifically TO-3 package is to handle high heat dissipation for a pass regulator,, modern approach is often to use a switching regulator without so much heat generation.

JR

PS; Sometimes pricing distortions are just because part is low sales volume.

Steel case is usefull when you need to get your regulator hammered without much damage, but I don't see the point to use steel as a pro when what you need is to take heat out of it, aluminium is a better conductor. So maybe what's happening is they are different parts, I mean, maybe aluminium is a fake part or something like that.

JS

Steel is cheaper so the old school mass production part... but these (all to3) are falling from favor.  IIRC the lateral power mosfets from Hitachi were only made using aluminum cases.

I think the steel TO-3 may be an alloy for better thermal or electrical behavior than cheap steel.

JR

Is it all-steel?

IIRC, TO-3 used to be tin-plated Copper base with a steel lid. Price of copper is UP. Must be 20 cents of copper in there. Al is not up as much.

But as-said, low-demand parts sometimes tend to premium price.

I don't think its a TO3 quantity issue. Mouser and Digikey both have the aluminum version for $3.80 ea. The steel version is $37.00.
Both vendors have a lot of in stock for both types. The published spec's are the same Temp, current, ect. Maybe the steel version
is Mil rated or something. I just buggs me that steel version is 10 X the price.

RonL

Aluminum has higher thermal conductivity than steel, so I'd bet it's a Mil spec for ruggedness.  What's the part number of the steel version?  Maybe the datasheet will give some clues at to standards the two versions meet.

mattamatta said:

Aluminum has higher thermal conductivity than steel, so I'd bet it's a Mil spec for ruggedness.  What's the part number of the steel version?  Maybe the datasheet will give some clues at to standards the two versions meet.

Click to expand...

Mouser Link with data sheet:

http://www.mouser.com/ProductDetail/Texas-Instruments/LM317KSTEEL-NOPB/?qs=sGAEpiMZZMticT%2fTNjssTgOMQqURCv27

rlaury said:

The published spec's are the same Temp, current, ect.

Click to expand...

No they dont have the same thermal specs as you can see from this extract of the data sheet you provided:

It has been a while since I messed with serious power but even 2'C/W seems high for a TO-3, and 21'C/W is for TO-220 plastic package. My recollection for high power TO-3 is closer to 1'C/W and potentially a little lower.

Low volume pricing anomalies happen...  perhaps they are not interested in selling many of those.  Dynamic pricing factoring in supply,demand, and reorder points may cause such distortions.

Indeed the theta junction to case depends on the header material, high power devices can use copper as PRR mentioned, but in my experience that is not common for mass market normal power products.

2'C/W does not sound like copper to me.

I spent most of my career effort in this area focussed more on resistance between case and sink and sink to ambient. I have one patent related to heat sink design (6,515,859 actually co-inventor).

With to-3 devices, the grease, insulator (if used), and even screw tension matters for theta to sink, then we have sink to air resistance.

JR

user 37518 said:

rlaury said:

The published spec's are the same Temp, current, ect.

Click to expand...

No they dont have the same thermal specs as you can see from this extract of the data sheet you provided:

Click to expand...

Look again.

The TO package is shown on page 32, a ~9mm metal transistor package, generally referred to as TO-39. TO-3, page 33, is rather larger.

Nowhere does the data sheet mention aluminum (or aluminium) in the context of the TO-3 package.

JD 'oh yes they do' B.

jdbakker said:

user 37518 said:

rlaury said:

The published spec's are the same Temp, current, ect.

Click to expand...

No they dont have the same thermal specs as you can see from this extract of the data sheet you provided:

Click to expand...

Look again.

The TO package is shown on page 32, a ~9mm metal transistor package, generally referred to as TO-39. TO-3, page 33, is rather larger.

Nowhere does the data sheet mention aluminum (or aluminium) in the context of the TO-3 package.

JD 'oh yes they do' B.

Click to expand...

I never said anything about aluminum, just saying they have different thermal specs

Thanks all. Looks like the mystery is solved.
If you have a lot of dissipation, you need the steel package.

RonL

rlaury said:

Thanks all. Looks like the mystery is solved.
If you have a lot of dissipation, you need the steel package.

RonL

Click to expand...

More specifically the to-3 package... steel is not a great thermal conductor, but two screws and larger surface area improves heat transfer.

JR