linear voltage regulator suggestion

[alm]

TO-3 has a lower thermal resistance to the heat sink, but is larger and usually more expensive. It's also harder to directly mount on a PCB. Hence the popularity of TO-220 (and similar packages like TO-3P). TO-3 would require a (slightly) smaller heat sink due to the lower thermal resistance. Performance and efficiency would be identical if the junction temperature is the same.

[shebu18]

Why not use the LT1083 instead. It also has a 1.25V minimum voltage but is rated to 7.5A

[Cali-Sam]

where can i buy the LT1083 from? and how much is it?

[digsys]

LM338 is only a dollar each on ebay, packaged as TO-3. could this be fake?
as far as heat dissipation, efficiency, and performance ,which package is better To-3 or To-220?

YIPES, that sounds like bad silicon, IF it's NEW. There are some trusted places where I buy surplus lots,
usually quite cheap. eg   www.x-on.com.au     Search your part # and check ALL the hits.
If you're not doing production, TO3 is better thermally plus larger transfer area (as suggested).
In production, it's a pain in the ass, which is why I'm designing them all out of a product range, too much labour.
The LT1083 also comes in a TO264 package, which is my MOST favourite package for assembly !!!
They are much dearer than LM338s, but an awesome regulator. Check X-ON

[Cali-Sam]

this might sound strange!!!
i was thinking about dumping my voltage regulator in a small contater filled with car engine oil instead of ataching a heat sink... is engine oil an insulator?
forgive my ignorance, im just trying to come up with better solutions.

[digsys]

this might sound strange!!! i was thinking about dumping my voltage regulator in a small container filled with car engine oil
instead of attaching a heat sink... is engine oil an insulator?  forgive my ignorance, im just trying to come up with better solutions.

They do that for seriously large transformers and similar units that have large surface area. You can't use any oil,
you'd have to check specs (google will tell you). But for small devices, you have no hope. Practicalities of containment and sealing,
just a lot of effort for little value. Besides, the MOST important spec is how FAST you can get the heat away from the regulator.
If you want to try it, go for it, practical is the best way to learn. The theory itself is fine.

Added: Some of the new EVs are cooling their Lithium battery packs with liquid !! ie Tesla etc Actually pumping it OVER the 44-66
lithium cells @ 180-280VDC 100-300A. We're still considering it ... lots to worry about :-)

[SeanB]

Using transformer oil to remove heat will work, but you will need to still attach a heat spreader to the device, and arrange the oil to flow past it. Otherwise it will just cook. You can try using a flat aluminium plate to increase surface area, and keep it vertical to assist convection. Then you will have to keep the oil cool.