The Great Circuit Challenge!

[Aleph(0)]

At the moment, I'm hard pressed to think of any analog device that can't be replaced with some manner of digital gizmo, other than an amplifier...

Cowboybob I say some applications (in addition to just gain/attenuation blocks) don't brook granularity! Given a chance Analog Reality is our best friend! Which isn't to rip on digital just reminding ppl that useful as it is it's still just an an artificial (discrete) construct built over a continuous reality! So in anticipation of anybody citing QM, PL, M-Theory et cetera as argument against continuous reality, I say better revisit basic undergrad science lest you're raised by own petard!

PS cowboybob Plz don't be offended! Cuz snark is definitely NOT aimed at you but, instead, at some ppl (mostly in my generation) who've forsaken appreciation of reality for petty convenience! So your point that gain blocks break bubble of _digital safe space_ wasn't lost on me after all!

 

[cowboybob]

You nailed it, Aleph(0) ...

 

[unclejed613]

I ran into that problem when I was in oil field a lot. The company I worked for used ethernet based interconnections for everything between the data van control centers and everything else and our longest ethernet cords were 200 feet. Problem was that for a few of our higher data bandwidth pieces of equipment they worked on a 200 foot interconnect lineset just fine in a singular 1 - 1 connection but if the site got set up wrong even adding even a 25 foot extension to its 200 foot cordset made that whole machine drop out of the network and refuse to talk to everything else yet still pass a connection ping test from every point in the system every time.

The only cheat was to daisy chain that unit through a midpoint machines systems 1 - 1 - 1 and have it work as a passthrough router of sorts assuming it plus the remote machines combined data rates did not then over load its connection bandwidth limits to the main data van systems, which most did, given the type of ruggedized industrial ethernet cord sets we used had terrible data rate capacity numbers. You could pull a small car with one but you couldn't expect much over 50 - 100K data rates at 200 feet and no amount of software would change that limit.

Analog reality choked the life out of digital function every time and very few understood why. I didn't and I chased that ghost for the better part of day before I finally got a explanation for it. Cord bandwidth Vs machine connection bandwidth requirements at a distance didn't add up.

i have a friend back on the east coast trying to set up a recording studio, and the multiplexer that goes in the studio communicates over a usb cable. he was getting errors with a 30 foot cable. a USB repeater fixed the problem.

 

[Beau Schwabe]

Challenge 5. LVDT (Linear Variable Differential Transformer) to 0% to 100% PWM output

This is supposed to be a fun challenge using as few components as possible.

Using only "off the shelf" IC's and discrete components .... i.e. 555, Op-Amps, resistors, capacitors, diodes, transistors, etc.
Note: If your design doesn't require a 555, then that would be valid as well.

You may use any component configuration you want as long as the intended end function is accomplished.


Intended End Function:

1) 5V operation

2) When metal is equally present (or absent) on both secondary coils (IOW balanced) , the output should be 0% Duty Cycle.

3) If metal is present on one or the other secondary coils, the output should be 100% Duty Cycle.

 

[Beau Schwabe]

No takers? .... Sigh...

 

[unclejed613]

it's been a long time since i've used anything with an LVDT in it, and when it was used, it had two analog outputs, going to a servo amp that controlled the braking of the feed reel in a tape deck. having a single ended output didn't make any sense to me. the purpose of using an LVDT is to keep the sensor centered, and so you need a directional component in the signal to sense what direction the error is in.

 

[Beau Schwabe]

the purpose of using an LVDT is to keep the sensor centered, and so you need a directional component in the signal to sense what direction the error is in.

... Not necessarily, using an LVDT in single ended mode can have other benefits such as temperature compensation negating any kind of drift. It really depends on the application. Another life ago, I worked in robotic research and development for prosthetics. I designed several non-contact sensors. In prosthetics this kind of sensor is robust enough to be placed within a prosthetic foot or hand for pressure sensing.

If you want a directional component to the schematic I posted earlier, then all you need are two transistors and three resistors..... I thought I would make the challenge a little bit easier by not specifying direction.