Decode voltage polarity to toggle a flip-flop

[DonEM]

How would I go about setting a flip-flop this way (input A=low) or that way (input B=low) determined by the polarity of the current at some sense point?

But the sense-point swings positive and negative relative to the logic ground.


Can anyone share a sexy way to toggle a flip-flop from this AC sense-signal?

Thanks. The problem sounds simple, but I'm not sure where to start. And, please excuse or read between the lines that I write, as I'm not an electronics tech as much as I'm a newbie woodworker.

I've this glimmer of a thought... Can a capacitor momentarily invert the polarity of a leading edge? Dusty memories.

I'm grateful.


DonEM
rank newbie

 

[crutschow]

A comparator, such as an LM339/393 should work for that.
What is the amplitude of the AC sense point signal that you need to detect?

 

[eTech]

How would I go about setting a flip-flop this way (input A=low) or that way (input B=low) determined by the polarity of the current at some sense point?

But the sense-point swings positive and negative relative to the logic ground.


Can anyone share a sexy way to toggle a flip-flop from this AC sense-signal?

Thanks. The problem sounds simple, but I'm not sure where to start. And, please excuse or read between the lines that I write, as I'm not an electronics tech as much as I'm a newbie woodworker.

I've this glimmer of a thought... Can a capacitor momentarily invert the polarity of a leading edge? Dusty memories.

I'm grateful.


DonEM
rank newbie

Hi

an AC signal has both polarities.
Is what you really mean is to have the flip flop toggle in sync with the AC input frequency? Is this what you want?

Or

Is the input signal expected to be a "half wave" AC signal where the half wave signal polarity is either positive or negative?

 

[AnalogKid]

Way down in the dust, you're close. A capacitor by itself cannot invert a signal, but a capacitor is needed to extract the signal polarity characteristic.

A differentiator circuit can produce a pulse or output logic level based on the direction its input is changing. It works best with rectangular waveforms, but can be adapted to handle other shapes.

What are the characteristics of the signal you are monitoring? Voltage peaks above and below GND, frequency, ***shape***?

ak

 

[rjenkinsgb]

One way is a couple of high CTR opto isolators with the LEDs connected parallel and inverse, then across the AC supply with an appropriate series resistor to limit the peak LED current to a safe value.

Then use the transistor outputs with pullup (or pulldown, as needed) resistors to feed the set & reset inputs of the flipflop.

Edit sorry, forgot it's current not voltage; use a few power diodes in series & duplicated in reverse parallel, all connected in series with the load to give a reasonably constant voltage drop of 2.4 to 3V when current flows, then connect the LEDs and a low value resistor across those.

 

[DonEM]

Hey! Rank newbie here <-- not sure how to do the response thing... 1st attempt...

A comparator, such as an LM339/393 should work for that.
What is the amplitude of the AC sense point signal that you need to detect?

Thanks, Crutschow. The comparator makes good sense, but I was stuck on thinking discreet, in a fog for starting.

Where this approach could go is toward dual-comparators to toggle whatever on either side of a neutral buss.

That would let me zero-balance the half-bridge and tweak the wave-form to balanced positive and negative switching.



Oh wow make me think! Thanks, really, Crutschow. This social mind-on-mind thing brings my brain out of a starting-fog.

The comparators need to output what the half-bridge totem needs for input... differential outputs to differential inputs that drive the half-bridge totem (two power transistors in series tapped at the middle for output that can switch to either power rail. For now my head sees a design-simplicity I'm after. (And I think it will work for later faster versionn like GaN totems and balance GaN drivers.)

Let me attempt to draw an electronic circuit of the dual-comparator scheme.

Then drawing will take grandpa here some time. Another learning curve.

I'll try to explain my intended use in the following newbie response learning curve.

You guys have been really cool.

Kudos,
DonEM

 

[DonEM]

What is the amplitude of the AC sense point signal that you need to detect?

The 1st prototype I will build operate a half-bridge totem through a test-inductor, then through a resistive shunt to the neutral rail between the plus/minus rails.

I'm keeping the resistive shunt low, but its value can be selected with some design flexibility to optimize the comparator design.

Probably a .1 Ohm shunt pulling a 100 amp load or so from plus/minus 40 Volt power rails. For starts. The half-bridge will drive by two pairs of voltage-doubling bridge rectifiers of two pairs of reverse-connected microwave oven transformers. They rectify at about 20 volts each MOT.

The shunt is the sense-point for the dual-comparator scheme to generate balanced switching thresholds for plus and minus sense events.

I'm wondering to myself how undstandable I am. Please advise.

 

[DonEM]

Hi

an AC signal has both polarities.
Is what you really mean is to have the flip flop toggle in sync with the AC input frequency? Is this what you want?

Hi eTech!

And yes, the quote above is exactly what I hope to implement. The AC is the square-wave voltage-drop across the shunt, and the dual comparators discussed with Crutschow will adjust to sense on the square-wave sense-point where a clean edge seems to be... tweak from the hip. Take a picture of the oscilloscope, and ask for help!

I will attempt to make the full-explanation in context as I go along, and will try to answer questions.

The implementation of the intended full mystery circuit is to reverse current on a conductor. A half-bridge drives one end of the conductor with current from either polarity rail. The other end of the conductor has the current-shunt for a sense-point to latch the comparator outputs (like a flip flop, or discreet latching --dunno yet.

So one end sends current this way or that way, and the other end senses and decodes and latches the current direction to a pair of differential outputs.

One conductor current reversing unit (CCRU to save typing) will connect to another identical CCRU, outputs to inputs. So a CCRU drives a second CCRU, and that drives another, etc.

The differential connections are made such that each CCRU conducts in the reverse direction of the CCRU driving it.

This makes one-pair of CCRU-s that implement a magnetic loop antenna by shaping the conductors into half-loops and using the pair to make a whole loop.

The whole system is a 6-qty CCRU ring-amplifier to energize copper loops on a donut in strange ways in high-power over-driven pattern-seeded probing of the tensile matter lattice (a rogue science could term it nucleosonic resonance hunt, but I'm not a scientist... rogue maybe). We be experimentalists.

The experiments are nothing about overunity <-- nor is the Earth flat. But I'm not sure about aliens! This project is just a personal compulsion. The study goes into time-domain of the edge-jitter of the magnetic ring-amp. All of this is to be a digital time-domain version of a spectrum analyzer, working of a log of time-stamps of the oscillator cycle-times. It gets really nerdy and out of scope of this forum I fear. But I'll talk off topic off forum if such is prudent.


I really appreciate this help from this knowledgeable forum!

Standby patiently for illustrations. Grandpa's on a slow roll here. Learning curves everywhere.


DonEM

 

[DonEM]

Way down in the dust, you're close. A capacitor by itself cannot invert a signal, but a capacitor is needed to extract the signal polarity characteristic.

A differentiator circuit can produce a pulse or output logic level based on the direction its input is changing. It works best with rectangular waveforms, but can be adapted to handle other shapes.

What are the characteristics of the signal you are monitoring? Voltage peaks above and below GND, frequency, ***shape***?

ak

This is good information, AnalogKid! Thanks.

I'm eager to see how you flow this together. This may optimize or stabilize things especially for higher speed versions of the ring amp in RF ranges. The effect I'm after works very high frequency on small (affordable) copper arrays. To make this cheap (small) these optimizations may be downright necessary, as they may already exist in parasitic resonance of component noise, etc.


I feel humbled. This could lead to some spiffy transmission line optimization of the signal paths too, perhaps. This is all over my head, beyond present skills, but sounds most applicable to improvement of versions past prototype/proof.


Cheers!
DonEM

 

[DonEM]

One way is a couple of high CTR opto isolators with the LEDs connected parallel and inverse, then across the AC supply with an appropriate series resistor to limit the peak LED current to a safe value.

Then use the transistor outputs with pullup (or pulldown, as needed) resistors to feed the set & reset inputs of the flipflop.

Edit sorry, forgot it's current not voltage; use a few power diodes in series & duplicated in reverse parallel, all connected in series with the load to give a reasonably constant voltage drop of 2.4 to 3V when current flows, then connect the LEDs and a low value resistor across those.

Rjenkinsgb that's genius! Signal translation on either side of a light beam.

Whoa! I do think me maybe getting rummy on the high of so many design-features falling together. Can a fellow over-dose on dopamine learning-reward? Make my head spin more! Please.

I'm excited because I've wanted to explore remote sensing of timing-parameters on the copper coils on a donut.

Also, the opto-isolation-design may be more tolerable of transient voltages that could spit and spat on the oscillating system at high power levels.
If a DIY opto-isolator could implement a beam splitter, (at sufficient frequency response --up to a few dozen megaHz, ish, don't quote me), then the same components isolating the current-sense-voltage could split-the-beam to route timing signals away from the kilowatt oscillator array on fiber optics to a panel of blinky lights some safe distance away.

To the thread: This project is not related to anything real-world, but is something like a 3-phase current chopper that could --in principle-- energize copper inductive-heating coils by several kiloWatts (pending power-transistors selected, etc.) So, keep in mind to keep high-delta magnetic fields clear of ferrous material... it would mess up the harmonics and fade the pink pixie dust to the wrong shade on the spectral abstration of the vapor-wave control console (not yet designed). Keep your head clear of the magnetics for that matter. TMS is way out of scope here!

DonEM

 

[rjenkinsgb]

To the thread: This project is not related to anything real-world, but is something like a 3-phase current chopper that could --in principle-- energize copper inductive-heating coils by several kiloWatts (pending power-transistors selected, etc.)

OK; for high currents you don't really want to be interrupting the circuit with diodes or a shunt resistor.

For that application a current transformer feeding a comparator or opto circuit would be more suitable.
Or a hall effect current transducer.
More info:

Current transformer - Wikipedia

en.wikipedia.org

https://uk.farnell.com/murata-power-solutions/56200c/current-transformer/dp/2126292

An example of the hall effect electronic type; the output on these is 2.5V (same as "ref") with no current and varies positive or negative of that, in proportion to the current through the device main terminals.
It would connect directly to a comparator, on a 5V supply, and also give a proportional measurement of the circuit current.

You can get similar units for different current ranges, anything from a couple of amps to hundreds or thousands of amps.

CKSR 25-NP | LEM CKSR Series Current Transformer, 25A Input, 25:1, 5.25 V | RS

uk.rs-online.com

 

[DonEM]

Dear Electro-tech-online community,

You guys are great! I have for a first in my life asked a question in a forum just one notch above newbie and gotten no flak, no insults, no advise without reading my post, and all around enjoyable learning experience laced with warm fuzzies.

Unlike any other forum in my several decades of self-study and community advice, this community launched a brain-storm on me of information from four points of the cerebral compass.

Wow people! Thank you from the bottom of my heart cockles! My birth-breached brain-child may survive this design-canal yet! My hacker is taken with this strange feeling of hope I've never gotten from any other forum ever. Ever. Since the 1990s I've been looking for you! Where have I been?


Warm regards and robust designs,
DonEM
<s>
Amateur Engineer and Newbie Woodworker
Director of Bucket List Operations (BLO)
</s>