Hi folks, first post. I joined to get some help with this problem.

I want to drive a solenoid through a transistor configured to latch on & off, using EDIT - 3x - NOT 2x - hall effect sensors to provide the switching inputs.

As the title states; what would be the most effective way to latch the transistor "closed" using the output from another "always on" sensor - then have it turn back on by biasing the transistor the other way again, using yet another sensor?

I have seen many videos using momentary push button switches to perform the junction bias & reset, but I want to use these sensors with an N52 instead.

I presume a BJT - NPN would work best for this, or would an N - MOSFET be better?

Planning to run the circuit between 12-23v, just not sure how best to switch sensor #2 on & off without needing to also power an Arduino.


Many thanks in advance for any help I get with this.

 

Is the solenoid a 2-position latching device or spring-biassed so requiring power to be applied to maintain one position? I'm guessing the N52 you refer to is a magnet and you want to switch the solenoid on or off with the magnet - by its presence or its polarity?

Do you have a specific hall effect sensor in mind? Most simple hall effect devices output 2.5v when no field is present and 0.2v for a S-pole and 5v for a N-pole, so some additional logic needed unless you can guarantee magnet polarity and strength.

Edit: Missed the A3144 reference in the title doh!

 

I presume a BJT - NPN would work best for this, or would an N - MOSFET be better?

A latch circuit using dicrete components would need two BJTs or two MOSFETs, not just one.

 

Is the solenoid a 2-position latching device or spring-biassed so requiring power to be applied to maintain one position? I'm guessing the N52 you refer to is a magnet and you want to switch the solenoid on or off with the magnet - by its presence or its polarity?

Do you have a specific hall effect sensor in mind? Most simple hall effect devices output 2.5v when no field is present and 0.2v for a S-pole and 5v for a N-pole, so some additional logic needed unless you can guarantee magnet polarity and strength.

Edit: Missed the A3144 reference in the title doh!

Hi Irving, thanks for the reply. Might be an N42 actually. Regardless; they are bipolar and have an N face and an S face.
They are called "latching" sensors but they don't actually "latch" on & off (Like a touch lamp will for example)

The * S-N attraction solenoid is to be wound with .8mm wire, not sure on resistance yet; some tinkering to do. (think 3Phase BL motor coils but iron and single winding, unless someone can say why bifilar wind would be better for this)

I will use N52's on the solenoid interface too instead of iron or steel, in that the solenoid can be pulsed in a tighter range of voltage allowing more sensitivity for adjustment, as well as reduced attraction cost per cycle and the spring overcomes the Lenz drag of the iron core.
There is an adjustable weight fixed atop the actuator which will provide inertia and act as a third input to free from Lenz drag.

Imagine a metronome, that's the style of operation we are looking at. (linear back and forth / to & fro)
It will oscillate at some 4-5hz (hence mechanical relays not being a valid solution)

I want it to use the absolute minimum amount of energy to perform this tip in balance.
I want to shunt off the back EMF from rapid coil shutoff through the N-MOS (as has been demonstrated ad - nauseum by now)
charge a capacitor, then time swap the charged cap with battery input, reusing that wasted power.

It's quite an unusual design but I have high hopes for the outcome.

 

A latch circuit using dicrete components would need two BJTs or two MOSFETs, not just one.

Thanks for the info.
Could you please link me to some examples so I can study them ?
That would help me a lot. I have already found a few examples of this situation, I get the basics, it's mostly which resistors do I use and where etc. New to this sort of stuff, and I have no formal training as such so teaching myself what I need to know.

 

Do you actually have the A3144E's ? When I checked for a datasheet it seems the A3144 is no longer in production (although advertised by online sellers via Amazon), and has been replaced by the A1104.
Googling "latch circuit schematic" will get you plenty of examples of latch circuits.

 

One picture is worth a thousand words ............
What are You trying to accomplish ?
.
.
.

 

Do you actually have the A3144E's ? When I checked for a datasheet it seems the A3144 is no longer in production (although advertised by online sellers via Amazon), and has been replaced by the A1104.
Googling "latch circuit schematic" will get you plenty of examples of latch circuits.

Yeah they're definitely A3144E, says it on the chip, got them off eBay for pennies.
Another manufacturer must have too it upon themselves to continue its production.

The ones I have seen all seem to use a momentary push button or two, Can't find a good example to copy for using these threshold sensors.

 

One picture is worth a thousand words ............
What are You trying to accomplish ?
.
.
.

Have a weighted actuator arm swing to and fro, hitting a spring either side, and on one side a coil pulses attraction at the last second, then is immediately cut off by sensor #3 making the oscillations cheap in terms of power.
There is a second stage to the machine consisting of a 2 weights off balanced, which will spin and turn a gear reduction.

Here is the original idea from 1939. The William F Skinner device. Yet to be actually disproved; lots of loose ends. Nothing comprehensive to copy, however.

 

As the title states; what would be the most effective way to latch the transistor "closed" using the output from another "always on" sensor - then have it turn back on by biasing the transistor the other way again, using yet another sensor?

Is this what you are trying to accomplish?
H1 activates the solenoid and Q1 performs the latch.
H2 is powered through Q2 and resets latch.

 

Is this what you are trying to accomplish?
H1 activates the solenoid and Q1 performs the latch.
H2 is powered through Q2 and resets latch.
View attachment 287887

Could be the ticket! Thanks; that's super helpful.
It gives me something to go off at least.

Any special consideration for the resistor values? Or just calculate using V=IR as usual?

Also, am I right in saying that were the sensors in the diagram momentary push buttons, it'd operate the other way around, seeing as these sensors provide a "negative" signal, or am I mistaken? (EDIT, they seem to put out about 3v + on D.O. , not negative output.)

(It makes little difference to the operation of the thing, just for curiosities sake.)

A push button would provide a positive signal, iinm?
Or does it just depend on the polarity of the connections?

 

Also, am I right in saying that were the sensors in the diagram momentary push buttons, it'd operate the other way around, seeing as these sensors provide a "negative" signal, or am I mistaken?

Not sure your meaning but a push button could provide either a negative or positive signal depending on how it's configured.
The hall sensors have an open drain FET transistor in their outputs which would be the same as a momentary switch with one side connected to ground providing a negative output.

 

Not sure your meaning but a push button could provide either a negative or positive signal depending on how it's configured.
The hall sensors have an open drain FET transistor in their outputs which would be the same as a momentary switch with one side connected to ground providing a negative output.

Thanks for the help. That clears up a lot of questions I had.

 

Any special consideration for the resistor values? Or just calculate using V=IR as usual?

R1 = 10K
R2 = 100K
Q1 = 2N3904

 

The Skinner device has a claimed 1200% energy increase and is a scam. Any 'over-unity' device is a scam. By all means build a latching circuit for your education, but don't expect free energy.

 

The Skinner device has a claimed 1200% energy increase and is a scam. Any 'over-unity' device is a scam. By all means build a latching circuit for your education, but don't expect free energy.

I know .. I have done a fair bit more research on it than most and I disagree.. Were it a "scam" where's the profit to be had??

Would be nice to see some evidence of what you claim.. None of us know, hence the replication.
I'm not expecting anything just playing with the possibility. Laws are made to be broken,
and were it true, there's only one way to find out. No corporation will make this. Where's the ROI?

The whole universe is in perpetual motion.. I think people forget this.
We know physics is an incomplete model.
Why the certainty?

 

R1 = 10K
R2 = 100K
Q1 = 2N3904

Awesome, thanks.
Does it have to be a P MOSFET? Or could I use an N-MOSFET instead but wired the other way round?

 

Does it have to be a P MOSFET? Or could I use an N-MOSFET instead but wired the other way round?

P- Mosfet required and recommended or PNP transistor. How much current does the solenoid draw?

 

P- Mosfet required and recommended or PNP transistor. How much current does the solenoid draw?

Using a step down buck with I/V control, circuit will be running at 18v / 1A for now.
planning to step it up to some 80v and probably more current though as this will work better (shorter on time)
Prototype need not use such high voltage.

One more question- were I to use PNP BJT and N-MOSFET how would the connections differ?

 

The whole universe is in perpetual motion.. I think people forget this.

Or people know that this is not true. The universe is not in perpetual motion. Per this article, the energy in the universe has gone down by 50% over the past billion years or so, give or take a few million. Read the article.