50-ohm power resistors

I learned a lot about liquid cooling of liquids when I started brewing again. At the end of boiling the wort you have to get several gallons of it from 210F to under 90F as fast as possible. Do it too slowly and infection can creep in, ruining the beer. If you don't get it cool enough but start fermentation anyhow the yeast dies. I found the swimming pool to be an excellent location to do the heat exchange.

:-)

Two-terminal dummy-load termination resistors exist. I have some here, from Alcatel GSM gear, but I can't find the manufacturer's details. They're 250W types, labeled "RM9-R 250-50".

I loved this idea, until I realised that the extra thermal impedance might be a problem. But a dummy load that cools itself only when needed and without a separate power source, brilliant!

I was totally kidding, but if designed properly it could be practical. The fan uses just a fraction of the power being dissipated so the Peltier devices can interfere with just a fraction of the power being dissipated. Or, if it didn't need to be a purely resistive load, the full load could be the DC motor driving the fan, lol.

Why does it need the conical or hyperbolic taper? What is that accomplishing? Is this so the impedance increases toward the open end of the line or so that it decreases and the line is shorted?

If a large enough piece of wire is used, it wouldn't need to get so hot. A 1200 watt space heater uses some couple of yards of resistance wire. Make it a number of yards and it won't get so hot. I guess that gets unwieldy. Or if dunked in oil it could remain shorter.

Win didn't get the Ph.D. - he bailed out into electronics early. If only I' d had that much sense. You do have to have a respectable mathematical educa tion to even start a Ph.D. in Chemical Physics, but if you have the sort of brain that can do mathematics while you are asleep you are much more likel y to have gone into a mathematics program - I do know a few real mathematic ians and my impression is that their brains are a bit unusual.

Dunk the resistors in a water or oil bath; stir until the soup is done.

There are SOT-227 (ISOTOP) power resistors. Think that'd be good enough?

The inductance and capacitance of a half dozen of those '9100s in series or parallel (as need be) should be pretty reasonable yet.

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

We're involved in a mind-boggling PLL design now (not the TI synthesizer, which is bad enough) with numerical nasties all over the place. I do NOT need another puzzle to think about.

When I was young, I'd delight in chess and puzzles and such. Now I ration out thinking to when it's really necessary.

I don't know. I'm not a transmission line guy. I just know that the article was written by some pretty sharp guys and backed up with measured SWR vs. frequency measurements.

One can argue with how sharp the "sharp guys" are, but it's hard to argue with measurements...

just build a chimney.

If you ground the end of the resistive element, the load resistance changes depending how far down the line you go. That requires a continuous change of geometry to avoid a distributed reflection.

Cheers

Phil Hobbs

And ration pretty stringently at that. When custom-wound transformers end up in the too-hard basket, you have to think that the amount of thinking available is insufficient.

John needs to go out and hire a few more people to do his thinking for him.

Great, thanks.

Chimneys suck for cooling a heat sink. I tried detecting an air current in a "chimney" above a 100 watt light bulb and it was virtually undetectable. The chimney was a three foot section of 4 inch dryer vent.

What did you use to detect the air current? I tried such an experiment long ago using a strip of toilet paper attached to one side of the chimney. Had some very visible movement of the T.P.

Grow up.

Well. _I_ am grown up enough that I _have_ hired some smart people to help me think and design. It's fun to think with other smart people. It's also fun, at my age, to prod them into doing all the work.

One of my guys got interested in the rational-fraction problem and coded up a c routine that finds the best N and D such that N/D = F where F is a real number from 0 to 1 and N and D are 22 bit integers. It's iterative, but runs acceptable fast on an 800 MHz ARM core.

One of my guys has got good with Sonnet, so she could attack the tapered resistor thing if needed. I suspect a lumped Spice model would work too. You can buy huge kilowatt multi-GHz dummy loads, so that must be doing some sort of tricky distributed stuff inside.

Designing electronics isn't about demonstrating how academically smart you are. It's about building and selling stuff that works.

e:

end up in the too-hard basket, you have to think that the amount of thinkin g available is insufficient.

him.