A non-traditional use of LM431

No pun intended but didn't you just reinvent the concept of the 4-20mA loop?

It is not about the concept, it is about the schematics. This circuit was initially designed for factory programming of the sealed wireless sensor. After the programming is done, the battery is connected to the power terminals. There is no additional drop on the power rails.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Somewhat reminiscent of a design I did many years ago. I needed to power and communicate with wellhead controllers over 3 miles away over a single twisted pair. I used a constant current 50ma source in series with a 600 ohm telco transformer on the local side and a 5.1 volt zener in series with a 600 ohm tranformer on the remote side.

I took 25ma off of the 5.1 volt zener to run the remote and used 300 baud single-chip modems on the transformers to send and receive data.

You almost (re?) invented HART protocol, but 300 rather than 1200 baud.

Best regards, Spehro Pefhany

Or though a good idea, it isn't new. sorry.

Maybe the way you're implementing it with the arrangement of components maybe but the net results is still the same.

for example:

We modified some equipment that operates a DC magnetic tension clutch supplied via a set of slip rings (collector rings) to energize the coil on the spinning mast. We needed a way to detect if the clutch was locked or slipping with the driving mast. So, we mounted a prox sensor on the mast just behind the clutch disc where it has timing teeth which was used for the prox sensor. I used the DC that's supplied the clutch and past it through our current amplifier that was mounted on the mast. this voltage also served as a voltage supply to operate the prox via a regulator. Any ways, the circuit simply varies the current slightly that represents the proximity sensor activity. At the supply end on the other side of the slip rings, one leg is past through a 1:1 xformer which isolates the raw DC (0..90V DC) and gives us the pulsed signal we used to determine if the clutch is in slip or locked rotor.

P.S. You need a clean DC power supply to operate the clutch other wise, you'll get noise in the detector circuit.. The other method would be sending burst of 50khz for each prox pulse which I did on one. It's all about cost, clean power supply verses using a dirty one with high freq pass detection.

Is that a real circuit, or just a spice-model circuit? The LM431 has a rather high voltage gain. It also has a DC voltage tolerance, up to 50mV, or 2%. I noticed the LM431 amplifier in your circuit is biased without any dc feedback, from two 33k resistors (5% resistors) dividing the 5 volts (5% is a common tolerance for 5V, where does the 5V come from anyway?).

I wonder if in real life some fraction of LM431s might not be saturated ON or OFF from the random accumulation of voltage tolerances in the circuit, especially as a function of the applied 5 volts?

Please do not give references to nasty URLs like above; it is so browser hostile that i had to use power-off reset to regain use of my PC.

It's a Larkin-style circuit... you just have to "tap tap tap" it just right ;-)

...Jim Thompson

It works either way. You can try soldering iron or spice.

It has been accounted for.

Is it a problem?

The current flows through R1, the whole receiver circuit operates as the voltage source. The voltage across R1 is proportional to the input current.

, from two 33k resistors (5% resistors)

???? Look at the schematics.

Applied is the current, not voltage.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Oops, may be time for a nwew PC ... or at least a new browser. Works fine here. It's a regular htm file on his company's web server or provider.

Works fine here also. But I normally run Firefox with severely throttled script capability.

...Jim Thompson

Same here. Enjoying the new PC now. Someone suggested I should plunk down some more dough and get the Intel dual-core. So I did and also upped to 2Gb of RAM. Had to do some Spice today. Man, this thing rocks and the fans didn't even speed up.

I'm enjoying my quad core. ;-)

Best regards, Spehro Pefhany

PC.

provider.

I write this from 150MHz/64M Pentium-I Win98/IE6 machine. No problems with url. I think Robert Baer needs to reinstall the driver skilfulhands.sys.

Vladimir Vassilevsky DSP and Mixed Signal Consultant

Is that made clear?

Spehro, you just had to say that, did ya?

I used such a machine until a couple weeks ago. Worked fine for Usenet and Gerber check but then it let off a few suspicious hisses and needed several PS start attempts on some days.

Maybe it's a browser config issue. Your link was one of the faster loading ones, no problems here.

I keep this ~12 year old workhorse. It is handy to have the ISA bus and the direct access to the true LPT/COM ports on occasion. Everything works fine, but all video applications are immediately out of class, opening PDF files is a torture, web pages are loaded much faster then rendered.

If you don't mind the dirt and dust, there are the biasing resistors at the power transistors in the PS. Reduce those resistors to ~100k value.

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

I know, I could kludge it. But there were also some pop and crackle noises on occasion, not so good...

I bet the reason "it works" is because you are using IE. It cannot be "a regular HTML file" and do that; perhaps there is a lot of Java or related fiddling going on.