Square root circuit

You may have been there and done that already, but if you drive the resistor with a PWM'd voltage then aside from the change of resistance with temperature, the power consumed will be proportional to duty cycle. Generating a variable duty cycle from a voltage is fairly doable.

Alternately, if you can have an inner loop that's closed by even a crude estimate of the resistor temperature you may get linear enough. You can linearize a thermistor over a moderate range with the right series and parallel padding, you could use a two-terminal temperature-to-current converter, or you could use a transistor's VBE as a temperature-to- voltage converter.

Have you changed your smokes or something?

lol

Sadly, you can't. The sum of the squares of the filtered outputs doesn't eq ual the sum of the squares of the unfiltered output.

PWM can be a really cheap way of getting a square root. Around 1975 I put t ogether and tested out my boss's idea of a really cheap PWM multi-plier and divider system, using LM301's - uA741's had some nasty habits - and an int eresting switch consisting of an normal NPN saturating switch (which only p ulled down to about 30mV) followed by an inverted NPN saturating switch (wh ich pulled down to about 1mV). A MOS-Fet would have been better, but not as cheap.

For 10V signals, we were getting 0.1% accuracy without any trimming. It was slow, but fine for process control.

A uC can be used for transcendental functions, but using it for square root is a bit irrational.

On Thu, 02 Jul 2015 07:19:44 -0400, Spehro Pefhany Gave us:

Rations? We don' need no stinking rations!

You guys know you use the moniker wrong. It isn't uC as in lazy man's

of a "micro-chip" It was "uc" as in "Unitary Circuit".

On Thu, 02 Jul 2015 07:19:44 -0400, Spehro Pefhany Gave us:

Rations? We don' need no stinking rations!

You guys know you use the moniker wrong. It isn't uC as in lazy man's

of a "micro-chip" It was "uc" as in "Unitary Circuit".

You don't realize what a dick you are. Get a life!!

Thanks Phil, Simon A. posted a link to the circuit, so I'm good to go in that regard.

George H

Really? Why? Bill S. said the same thing so I must be missing something.

Oh.. I'd have to use the PWM to make the sqrt. Control voltage sets PWM width and pulse height. The four transistor thing looks easier.

George H.

Well a small company and we play to our strengths. I guess part my job is to get little signals up to the volt level and then use some other device to digitize/ record it. Another "joke" I tell is that we use the interface standard that will never go out of style... The Volt. :^)

George H.

The two pots in the drawing are not used that way. It would be better if each pot ran from +12 to ground, and then sum the wipers with two resistors.

I think the dashed thermal coupling bit should have been between the heater and the cryo-diode, and whatever klutz that redrew the schematic put it in the wrong place. How would you thermally couple two pots anyhow?

I don't understand the LM399. The LT1004 sets the cryo-diode current, so the +12 doesn't matter too much. The IC3A bootstrap has a legal right to never start up, given the common-mode violation.

Piece-o-junk. The square root part is overkill but cute.

On Thu, 2 Jul 2015 20:09:22 +0800, "Rheilly Phoull" Gave us:

Says the retard who doesn't even know how to quote properly in Usenet. You're an idiot, child. I was making light of basic stupidity, but you take basic to a whole new low.

As a small company I try to work a little differently. I try to augment my weaknesses so that they are not liabilities. That said, my main focus is to find product areas that allow exploitation without attracting the attention of the big fish. Often agility is as important as any technical prowess. After all, you can't do the engineering at all if you don't do a good job of marketing.

Two women, R and my daughter L. And I just appointed a new engineering manager that, technically, I work for. Given how much I dislike management, I don't want to do it, either.

But we're not bad, in that we almost always do things in a reasonable time frame, that make sense. I have customers that have TEN levels of management, the middle 6 of which are worse than useless and take years to make decisions, if ever. The few people who do get things done there mostly ignore management.

I found that, most of the time, if you ignore management they don't call you on it, as long as your electronics works.

You could likely PWM into your heater at a lowish frequency and soften the edges, to reduce coupling into your signal circuit. But there would be some risk.

I did PWM into the sample heaters in the Varian/Agilent NMR systems and got away with it. I ran around 130 Hz, with slew-limited edges.

Have, have you ever posted anything friendly, or helpful, or funny?

All that rage is going to kill you.

Lol. Managers that don't manage is what you are talking about.

Criticizing managers that don't manage? Sounds familiar somehow.

Lol. Managing engineering is only partly about getting the designs to work. That's much of the point of managers. They deal with the whole picture. I once completed a design 1 month late because my vendor screwed me on a part delivery by selling the reel to someone else when it came in. I got paid for the product, but it was never used and so no follow on sales. Had I been working for an experienced manager they likely would have focused on *all* aspects of reducing risk and made sure there was a plan B for the part. My design was great, but it was not worth much to the company by getting manufactured late.