insult

On Wednesday, January 17, 2018 at 1:05:37 PM UTC+11, snipped-for-privacy@notreal.com wrote :

at

e

oot.

I wonder what such a check-up would show for krw? Probably nothing - the pa rasitic load of his silly ideas presumably mimics the normal cognitive load , even if his brain is so busy articulating right-wing delusions that it ha sn't got time to do anything useful.

--
Bill Sloman, Sydney

ASR33's didn't have lowercase. Neither did 029s.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

They aren't...

niether are Via (viatech.com) according to this SX:

The raspberry Pi's Broadcom branded ARM processor is reportedly safe.

--
This email has not been checked by half-arsed antivirus software

nah, it seems to translate straight to machine language, there's other compilers like that. Turbo C for example.

--
This email has not been checked by half-arsed antivirus software

r ZIP

IC)

it can

on to the

s.

ng time, and with good reasons.

which), GIMP, I don't, *, Firefox. It all runs on Linux, and not an .exe to be seen.

they're executables that are downloaded from the producer of the distro, no t just wherever on the web. There's a rather large difference.

hardware design, writeups, experimental projects, and a lot of other stuff too, most of it not electronic. Your forte is speed, mine is cheapness. I s pent my early years at the high price high reliability end of electronics, and disliked the gross inefficiency. Gradually I pieced together how to cut costs everywhere every way.

e to compile their own extensive application software respositories is that downloading exes from 3rd party sites has been a major source of malware f or windows users. And no, that's not just from fringe sites & dodgy authors . The whole process is inherently vulnerable.

lse posts, even with the best of intentions.

you persist in completely failing to understand the security situation. It' s not you that's distrusted.

NT

Indeed. Time was when you never ran anything you downloaded from the internet without inspecting the sourcecode first and then compiling it. Those days are long gone but you have to admit that a single word title like "insult" and a zip file make it look exactly like a hacked account.

If you had titled it thermal resistance calculations and linked to a few interesting images then you would have had a different response.

--
Regards, 
Martin Brown

Luxury. MITS Altair 8800's only had toggle switches.

Hey, that program's pretty cute. I already used it to figure a sub-pF compensation cap I'm thinking of making from FR-4.

(I usually use a spreadsheet I wrote ages ago for PCB and wire calculations, but your calculator sure is a lot lighter and nimbler.)

Cheers, James Arthur

If a bunch of engineers meet at a bar, or a design review, they might discuss some things that aren't textbook theory. Some things are interesting from a system dynamics standpoint, or peripheral to engineering, like economics and the psychology of design. And lots of engineers cook and like (or brew) beer.

What degrades and damages a forum are ritual personal insults, raging profanity, partisan politics, and low-rent rants from people who don't design electronics.

A war in Korea would sure affect the semiconductor industry.

Of course Tennessee is occupied. It would be awfully quiet if it weren't.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Just remember the roughly +900 PPM/K tempco! I had to special-order a reel of N4700 caps to compensate a 600 MHz coaxial ceramic resonator oscillator. I almost got it right... missed a small temperature gradient from another circuit.

I wrote it so I could seesaw between heat sinking and capacitance while I'm Spicing my pulser circuit. But it's arithmetic that I do often enough to make it a little program.

Really, the transition from paper tape to punch cards was wonderful. Editing became shuffling cards. You could color code subroutines by marking the tops of the cards, and copy them to use in other programs.

It was cool to look at a card deck and see the program structure. You can't do that any more.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Good point. It's a feed-forward for a 'scope-compensator type situation--I have a HV divider feeding a debugging testpoint.

HV >----+-----. | | .-. | | | --- Ccomp | | --- '-' | | | +-----+------+----> t.p. | | | .-. | | | | --- --- | | --- --- '-' |FR-4 | trimmer | | | === === ===

I'll have a trimcap on the cold side, but it wouldn't hurt to make most of the cold-side capacitance out of FR-4 too. That way the ratios should track, which is what matters most here anyhow.

If you could we'd all be horrified at what we'd see!

Cheers, James Arthur

Remember when we went to Halted & I got you some of those ceramic resonators? I lost some of mine, probably still in your car somewhere.

This looks fun:

signal >-+---------------. | | .-. === (top plate) | | R1 '-' +-------o=============== (middle plate) .-. | | R2 '-' '--o============+=========== (ground plane) | -+- GND

in >--. | .--------|--------. | | | | .-----|-----. | | |/////|/////|

Tektronix did an appnote once on pcb capacitance "hook", which was I think strange dielectric absorption that varied between board batches. It made uglies in scope step response. I don't know if your app is that pickey.

FR4 is fairly nasty stuff.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

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

It's nasty but if everything's FR-4 all of that balances out, right?

My main trouble here has been that I've had to cobble multiple parts together to meet my voltage requirements, producing intermediate nodes (of various unwanted time-constants) with stray FR-4 'capacitors' to ground.

Bootstrapping or guarding those with structures in FR-4 seems like a cute trick.

But even better is getting rid of those intermediate nodes, which, by more digging at Digikey, I was able to do with higher-voltage components. This time.

As we go higher and higher, though, at some point I may resort to at least bootstrapping the strays with guard plates buried in the FR-4.

Cheers, James Arthur

I'm about to haul my lazy carcass to work and breadboard my kilovolt pulse generator. If it works we'll show it to a potential customer next week.

The ultimate load is roughly 10 pF, so I can snoop the output and simulate the load with a 1000:1 capacitive divider whose first cap is

10 pF, 1KV, which I happen to have in stock. I shouldn't have any PCB problems with that. I'll have thermal problems as I crank the pulse rate up. The customer can provide water cooling!

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

Can you be sure the various layers (core, prepreg, etc.) are similar across boards and manufacturing runs?

I think we can be sure that they won't, but that's not necessarily important. If they vary 20%, I wouldn't toss away nulling 80% of the stray on that account.

For certain ratios (such as equal series resistors) you can design a structure that does the job on one layer, eliminating the FR-4 variations' effects.

.-.---.-. .-.---.-. | |R1 | | | |R2 | |

No! You're lucky if they get the stackup sorta close.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

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

Sure (rather slick, actually) but I thought you were talking about a more complicated structure, using more layers. Here, you really only have the variation of one layer and even it's pretty much nulled out.

SNIP

You idiots never snip any more. What "We have bandwidth now so who cares" mentality. Clutter is clutter, dumbshits.

From the graph you posted, it looks like you are making half a kilowatt at 1000 volts.

That is pretty powerful as HV power supplies go (not referring to AC power distribution schemas, of course). Does it fit in a 1U rack slot or 2U or?

You should use a 10pF 2 or 3 or greater voltage cap. Your cap choices, especially when working with HV, should always be twice your expected EMF levels.

Otherwise...

"Holy holey caps, Batman."