Heatsink metal--Wikipedia

 

Nice equation in here to figure things like junction temp when on a heatsink:
http://www.jaycar.com.au/images_uploaded/heatsink.pdf

 

Never ever underestimate the value of a good electronics technician, or technician in general.

I did the technician job for 2 years out of school, and have been an EE by title for 1 year.

I needed to comment on this post again with some observations I've formed in some of the various jobs I've had in the long past.

Most true EEs can design well but rarely consider servicability of the product which, as with most things, often becomes a necessity somewhere along their life.

I have worked on so many things that the engineers never even considered how hard it would be to service said product.

Some really bad examples:

1) Consumer TV sets. It was never considerd that along time people clean the CRT screens, usually by spraying a bunch of window cleaner on them to the point that some drips down the tube and onto the underlying circuit board. All the pushbutton switches, as well as some of the remote control circuitry, is exactly where it drips on and over time you get bad switches and eaten up PC Bd traces. On most of these sets the chassis can be pulled out far enough for servicing without having to unplug a dozen connectors, others are so bad you have to pull the picture tube to get the chassis out. Oh well, TVs are almost a throw away item nowadays yet this still looks like it may apply to the many of the flat panel LCD sets too.

2) Mistubishi computer monitors, from the cheapest to the professional CAD monitors and even the 35" - 40" presentation models. Shields abound, and there's really no way to work on most of the main PC boards without buying their $200 set of extender cables, many which were unique to particular models.

Back in the very early 1990s, after we lost some grant funding at the OSU laser & photonics labs I was working at, having an engineer/tech on the team was considered a luxury even though I often saved them 10s of thousands of $$ repairing a failed power supply for a high-powered Argon-Ion laser, designed and built speciatly circuits for them or even hand machined many unique optics mounts and the such since I also had access to a most wonderful machine shop. I honestly don't think there are words in Chinese for "lathe" or "vertical mill" and sometimes these guys would come watch me constuct something with great fascination.

I gave my part, after we lost funding up there I moved back to OKC to help out an old friend of the family who had fallen upon serious heart problems and needed someone to operate his 45 year old consumer electronics sales and service store. I made it a point to hire one or two "still in school" EEs just to give them some hands on and I know the world ended up with some better engineers because of it.

Point is I think all EE programs should have at least one course dealing with "what if it needs to be serviced later" so they might consider that in their designs.

 

Ok I managed to do a successful experiment:

I bolted an LM317 to a 0.8mm steel chassis and set the output voltage to 2.5V. I then increased the loading over a period of about 2 hours, and monitored the output voltage. When the temp reaches its maximum safe value, the thermal protection circuitry kicks in and holds the device dissipation constant regardless of load. This occurred at about 10W dissipation, suggesting something in the region of 10C/W thermal resistance for the chassis!

I also tried with a 1.5mm aluminium chassis, but with my PSU maxed out with 15W device dissipation, there were still no signs of thermal limiting, and the fixing bolt was still under 100 degrees. I would venture to say it would have gone to at least 20W.