insult

I suppose I should get serious about Python, one of these days.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

I like PowerBasic. One of my guys did an embedded c thing that involved signal averaging. I thought it was slow so I did the function in PB. Mine ran four times as fast, and I used the dumbest possible subscripted FOR loop, where he used fancy pointers. After a day or so of playing with code and compiler switches, he got close to my speed.

The zip that I posted included a 23 kbyte .EXE file. PBCC has a #BLOAT option if you want megabyte exe files.

Bob Zale, the (somewhat crabby) inventor of PowerBasic, died a few years back, but it looks like the product lives on.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

I can't answer as the reliability of heat pipes other than to say that Apple uses them in their Macbook Pros and I have yet to hear of failures. I was hoping they may help your situation as I understand that heat pipes are extremely efficient at moving heat.

A bit expensive though for diamonds although there have been experiments using diamonds and heat pipes:

The potential problem with heat pipes and corrosion would be dependent on the purity of the water if used as the coolant.

John

Hint: it's not just you and the recipient in the chain. Hint 2: almost no-one has been willing to touch exes for a very long time, and with good reasons.

NT

The fins don't need to be diamond. You can get respectable layers of vapour-deposited diamond by chemical vapour deposition

The process is slow and expensive.

Everybody else did, years ago. My boss at Fisons Applied Sensor Technology (where I worked from 1991 to 1993) definitely had it in mind back then.

Heat pipes are nicer.

--
Bill Sloman, Sydney

the

ermal

I did

onstant

ld

ZIP

)

t can

n to the

eaders

is

n

-solution

Typically it's pure water (de-ionised) and copper pipes and heat sinks.

Plumbers have been coping with corrosion in copper pipe-work for quite a wh ile now. You do have to clean out any flux before you put your pure water i nto the (evacuated) heat-pipe and seal off the system, but if you treat the heat-piping as if it were a vacuum system you can't go far wrong.

One trap for the less careful is to let some air (or other non-condensible gas) get into the sealed system. Water vapour diffuses through non-condensi ble gas much more slowly than it flows towards a cold point to condense - i t used to whistle through my vacuum line back when I was a graduate student , and I could hear the whistle above the boiling noises in the liquid nitro gen that was condensing it.

Hydrogen gets everywhere, but there's not a lot in air to start with - thou gh the mammalian digestive system is always generating some.

--
Bill Sloman, Sydney

PowerBasic is probably transpiled to C/C++ "under the hood", so if you were using the included list data structure with PB and it's transpiled to some (highly optimized) standard template library vector-like data structure for the PC, and he's trying to roll his own for an embedded platform without STL support that might explain some of the difference

Also, cache locality. If your dataset is small enough that it can fit in its entirety in a PC's say 4 or 8 megabyte L1 cache so that it never has to go out to RAM to fetch more while processing a block that will certainly speed things up considerably

Nobody in their right mind downloads a foreign .exe file.

John Larkin might not have put malicious .exe file into his original upload, but malicious people have been known to replace innocent files with more dangerous stuff.

The source code might have been a more considerate option.

--
Bill Sloman, Sydney

The PowerBasic compiler and the IDE are both written in PowerBasic. The compiler makes x86 code directly. It also allows inline assembly code.

I've written useful FOR loops that execute at 100 MHz.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Well, maybe you don't trust Dropbox.

How do you run anything on your PC?

My browser is

"C:\Program Files (x86)\Mozilla Firefox\firefox.exe"

LT Spice is an EXE.

Heck, almost everything runnable is an EXE.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

From trusted sources. Nobody trust you to find your ass with both hands.

For one thing you don't spend as much on keeping your systems secure.

--
Bill Sloman, Sydney

Some errors: the dielectric constant of Al and Cu is quite large, certainly not zero.

This could be scratched up in JS and placed on a webpage, where it runs (reasonably safely) within a browser's sandbox. That would've avoided complaints. I can do that for you, if you like.

Tim

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

Surely you know there is a whole list of intermediaries in the process.

Nothing on my computer is except LTspice.

NT

How do you look at PDFs, edit text, browse the web, create documents, manipulate images, compile programs, lay out PCBs, send and receive mail?

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Since they are dead shorts, it didn't make sense to calculate capacitance.

You can do that if the whiney scairdy-cats here want it. My version works fine for me, and is totally safe.

But the people complaining have no use for complicated stuff like heat flow and capacitance anyhow.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

That settles for "complicated" these days? I just let Google Calculator handle it:

*kelvin%29%29+*+2+inch+*+3cm%29+in+kelvin%2Fwatt

I was at least hoping for a thermal diffusion applet or something.

On that subject... it looks like there is such a thing for free:

Can't find a web-based one (surprised Falstad didn't do one), might be neat to whip up a simple 2D one some day.

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
Website: https://www.seventransistorlabs.com/

A TO247 insulated from a copper block is pretty simple. The only serious 3D thermal analysis would be involved in the heat spreading inside the copper, out to the ultimate heat sink.

Being 2D limits it a lot. It could be used for planar heat spreading, like PCB copper pours. But I can use Sonnet Lite, or dremel some copperclad.

The math solution is trivial. The work is in the GUI.

A lot of thermal problems are best solved by breadboarding. Often faster, always more accurate than simulation.

I have an experiment sketched on my whiteboard, waiting to be run. Heroic cooling of SOT23 transistors, with actual junction temperature measurement. I wouldn't know how to simulate that.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

That sounds interesting. What is needed to measure the junction temperature of a SOT23? Do you use a FLIR?