What's the Toughest Branch in Electronics?

A cheap, like 85-cent, uP like a HC05, with an onboard mux'd 8-bit ADC, digitizing the voltage and current waveforms. The tricks are in the signal processing.

John

My next project is to make a very long lifetime 6 button control pad for a high humidity environment + salt. Yah..Yah... it's a dum project but I quoted high and got the job. D from BC

Familiar is easy .. New is hard...

I'll admit I'm least knowledgeable in RF... stuff like antenna patterns, smith charts, Maxwell stuff, s parameters... Scary...

I'm currently trying smps design.. A smps is like a pile driver that makes everything shake and I'm using duct tape to control little things from vibrating :P

D from BC

I was reading along anticipating something snappy about digital...and ohhhhh....

There might be a digital design guy in the prosthetic department that does CAD for fingers, thumbs and toes. It must be tough getting skin tone right and figuring how to keep fingers from falling off. :P

D from BC

Le Sat, 08 Sep 2007 16:42:12 +0000, D from BC a écrit:

You can't have the trophy wife and get or remain rich.

Pot the electronics, and use something like capacitive finger-sensing or potted switches (or possibly hall-effect sensors) activated by buttons that move magnets. Alternatively, pot in clear material and use optical sensors which are interrupted by a "blade" on the buttons...

Take some extra pains with how the cable or wires coming out of the potting are routed, making an effort to provide strain relief and a long leakage path, as that will be the major exposure to damage with a properly potted system. Consider two potting compounds - a harder epoxy-type for the electronics and a layer of silicone or urethane softer material as an extra layer to provide a longer leak path for the wires leaving the unit. Pot with vacuum. Put a few tight-fitting o-rings over the wires/cables before potting.

D from BC hath wroth:

Sure. Digital design (using bits, not digits) is a dying pre-occupation. Most digital designs can best be produced by a digital computer. Soon, the computers will all be designing themselves. You can try to do that with RF, but package size limitations prevent even coming close to automating RF design. Much of the real difficulty in digital remains with the GROW (Getting Rid Of Wires) problem, which is the curse of all things digital.

Is that sufficiently snappy (and insulting)?

Ummm... Nobody in the CAD department does much design these days. What happens is that the chip manufactory supplies a "reference design" to their customers. It's usually easier to just copy the design, than to do anything new. So, CAD has become (Clone And Document). I note that Google finds 1440 hits for "PBC reverse engineering" in quotes.

I can always grab someone passing by and borrow some more digits.

Incidentally, I are (or was) an RF engineer. I vaguely recall taking a class or attending a seminar in how to properly look down my nose with disdain at digital designers.

Jeff Liebermann wrote: [... snip...]

What's the story on that microwave leakage detector you have in pics on your website?

The salvage jobs are the best ones. It's impossible for ignorant people to realistically assess the impact of that ignorance upon the project, and this partly defines ignorance. Ignorant engineers are the rule and not the exception, so much so that it partly defines engineer.

Potting solves nothing in this case.

Perhaps but why bother with the extra electronics?

Bad idea. Magnets get lost.

This doesn't work very well in bright sunlight. If he has that problem too, it won't fly.

If you must pot, then yes for sure. Also, you want to strip and tin a section of the wire within the potting. Wires work like little hoses to the outside world if you dont.

No, that is a bad idea. It causes stress under temperature cycling. You want the material on the PCB to have some give to it. Outside that, you want the strong stiff stuff to keep the external forces out.

That doesn't work. Water goes down the > >

I like the hall effect sensor idea.. I think it might make a slimmer control pad compared to the opto method..

Thanks..gives me something to think about before I have to think about it :) D from BC

That's better.... :) D from BC

I was working with one guy many years ago.. I looking at his constructed pcb and thought ..wtf?... He used something like two gigantic 10000uf capacitors for a 20W audio amplifier module used for audio. I wanted to ask the boss if I could see his resume..

D from BC

Fred Bloggs hath wroth:

It's an MD-2000 microwave oven leakage detector.

I spend most of my time infesting alt.internet.wireless and various RF related mailing lists. Most of the stuff on my web pile are photos posted to answer specific users questions. Also see:

The plan was to improve the sensitivity of this RF leakage detector sufficiently to be used to measure antenna gains and detect non-802.11 interference sources. I was going to attach an external antenna and possibly an RF amplifier. However, the device turned out to be a diode detector (D1 on the left of the photos) which was seriously insensitive. So, I built a sniffer and field strength meter from scratch but left the photos of the MW oven leak detector online.

Which color ?:-)

...Jim Thompson

Phil Hobbs snipped-for-privacy@SpamMeSenseless.pergamos.net posted to sci.electronics.design:

He did not say wavelength, he said bandwidth. There is a difference.

OIC- there are several technologies floating around to do this and the sensitivity does not have to be that good because the safety standard is way up to 4mW/cm^2 for domestic appliance leakage. The niftiest architecture around was a Narda fixed frequency job with three dimensional coverage by a small set of tuned dipoles. The dipoles were also thermocouples, the DC voltage resulting from the minuscule amount of RF heating was marginally conditioned and input to a jellybean V/F converter, and from there into a jellybean uC external interrupt which simply timed intervals to derive an RF field strength. So you ended up with a $300 antenna driving $1 worth of electronics to decide whether to alarm or not. It was measuring an ANSI threshold for human body absorption which is actually quite a huge field, so again sensitivity is rather poor.

Details, details. ;) I could just as easily have said that the readout performance is _improved_ by UHF-modulating the laser and making it 1 THz wide, because it is. Getting rid of mode hopping is a real genuine benefit, but FM noise is actually a big help in getting stable AM readouts with a diode laser.

I get so weary of people 'questioning' things but not being willing to stay and listen to the answer. My children are much better at that than many grown-ups I meet. In theology it's called 'the hermeneutic of suspicion', but it shows up in all fields these days.

Clarity of thought matters a great deal to me, which is why I'm grateful to people who show me my errors (as in this case). It's a great pity that our culture has retreated so far from intellectual rigour, to the point where systems that are demonstrably almost error-free, and that can be made completely error-free by simple block coding, are 'questioned', but not actually _examined_, by their critics.

It would be a wonderful sport if it weren't so destructive.

Cheers,

Phil Hobbs

That kind of thing is even on the screen...