2 hand control circuit for 2 coils

[zplitt]

Hey there!

As you can see, i´m a newbie. not only here, but in terms of electronics too.
i´m working on a project on which i want to switch 2 coils (3,6A together) with a two hand control circuit via TTP223 sensors. when both sensors give an high output, the coils should be activated. i want to run it with a 5V DC power supply, but with a back up battery directly on the coils.
with help from a few nice people and a little bit of AI i figured some things out that could work for my purpose. i have no idea how to use proper circuit drawing software, but i do know photoshop    so i tried to draw a circuit and hope you guys can follow my idea. so please do me a favor and check it out.

am i on the right way? something additional needed? something wrong?

thanks in advance and greetings from Austria!
...and please dont be too harsh on my english knowledge...

[pcprogrammer]

Take a look at the EasyEDA website https://pro.easyeda.com/editor. It is a free schematic and PCB design package that can be run locally or online.

About your circuit, it depends on the TTP223 touch module output if it can work. it has to have a push pull output for this to work. I think the module has a CMOS output pin which is push pull so you are probably good here.

The MOSFET's are probably IRLZ34N because I could not find anything on IRL43N. They are certainly up to the job. The 220 Ohm series resistor and the 10K pull down are probably not needed. The 10K has no use in combination with the CMOS push pull output and for the series resistor it depends on the gate capacitance and the TTP223 output if a resistor is needed. If, it can be smaller like 47 Ohms or so.

It reduces the strain on the output, but delays the transition from off to on and vice versa, which might lead to more heat dissipation in the MOSFET's, if I'm not mistaken that is.

[Uunoctium]

The battery, as it is connected, always supplies power to the coils.
If it is to provide a redundant power supply in addition to the 5 V, the battery terminal must be connected to GND of this

[Zero999]

Why not just use a logic IC, rather than trying to make your own AND gate with two MOSFETs?

Although AND gate ICs are available, I only stock NAND, since they're more flexible.

Here's an example using the 74HC00. Two gates are used, one as a NAND, followed by an inverter, a NAND gate, with both of its inputs tied together. The inputs of the unused gates are tied to either +V of 0V.

C1 is a decoupling capacitor and must be as close to the IC as possible.

[zplitt]

Thanks for telling me about EasyEDA, gonna check that out!
I had to look up "push-pull" and i have absolutely no clue..
CMOS, yeah, about that. Maybe we shouldnt go too deep into details (for now) 
I already knew what a transistor is, but heard about MOSFETs just a few days ago.
You are right, it should be IRLZ34N.
So no resistors at all? I often red about how necessary a pull down restisor is. The MOSFET should be fully open at 4V, so i thought the 220 ohms were fitting.
Heat would not be an issue, as the circuit will only be operated maybe 3 times a day for just a few seconds at max.

The battery will not be connected permanently, it only gets hooked up just in case if there is a failure. But thanks for the GND tip, maybe need this for some future projects!

I thought about an AND gate too, just for my kind of thinking two "switches" in series were easier to understand. The one i found was the SN74LVC1G08, but couldnt tell if this would work for that purpose.

I guess there are MANY ways wich get the job done, i just try try to figure one out that's easy to understand and not likely easy to fail.

[Zero999]

You are right, it should be IRLZ34N.
So no resistors at all? I often red about how necessary a pull down restisor is. The MOSFET should be fully open at 4V, so i thought the 220 ohms were fitting.
Heat would not be an issue, as the circuit will only be operated maybe 3 times a day for just a few seconds at max.

The battery will not be connected permanently, it only gets hooked up just in case if there is a failure. But thanks for the GND tip, maybe need this for some future projects!

I thought about an AND gate too, just for my kind of thinking two "switches" in series were easier to understand. The one i found was the SN74LVC1G08, but couldnt tell if this would work for that purpose.

I guess there are MANY ways wich get the job done, i just try try to figure one out that's easy to understand and not likely easy to fail.

Where did you hear pull-down resistors are necessary? The main reason to use them is in the case of a microcontroller driving a MOSFET, in which case the resistor ensures the MOSFET is off, if the microcontroller's I/O pin is set to an input (high impedance, rather than either high or low), which would cause the voltage on the MOSFET's gate to be an indeterminate level, which could cause it to turn on, when it should be off, or get hot, if it's partly on.

In my circuit, the MOSFET is driven off a logic IC, which has an output, which can only ever be high or low, unlike a microcontroller's I/O pin, so there is no need for a pull-down resistor. If you were going to add pull-down resistors, then they should go on the logic IC's inputs, to prevent them from floating, if the TPP223 were to go high impedance.

Series gate resistors are generally recommended, if the MOSFET is being driven from a low impedance source, which can cause oscillation, due to the parasitic gate inductance. In my circuit, they are also unnecessary, because the 74HC00 has an output impedance of around 50 ohm which is high enough to damp any oscillations.

The SN74LVC1G08 is a good choice. I only suggested the 74HC00 because being through hole, is easier to solder and I only stock NAND and NOR gates, because they're more flexible i.e. it's possible to use them to make oscillators with, which can't be done with non-inverting AND and OR gates.

[pcprogrammer]

I had to look up "push-pull" and i have absolutely no clue..

It is not that hard. A push pull output has a transistor for both levels, so a low is a "hard low" and a high is a "hard high". This is not the case for an open drain or open collector output, where there is only one transistor. In this case only a low is a "hard low" and to make a high a pull up resistor is needed. (Or vice versa, but that is not commonly used)

I guess there are MANY ways wich get the job done, i just try try to figure one out that's easy to understand and not likely easy to fail.

Yep, it is also possible to make the AND function with two diodes and a pull up resistor.

[zplitt]

Where did you hear pull-down resistors are necessary? The main reason to use them is in the case of a microcontroller driving a MOSFET, ...

Series gate resistors are generally recommended, if the MOSFET is being driven from a low impedance source, which can cause oscillation, due to the parasitic gate inductance. In my circuit, they are also unnecessary, because the 74HC00 has an output impedance of around 50 ohm which is high enough to damp any oscillations.

The SN74LVC1G08 is a good choice. I only suggested the 74HC00 because being through hole, is easier to solder and I only stock NAND and NOR gates, because they're more flexible.

It could be i heard and red this in case of using a microcontroller, makes sense.
The SN74LVC1G08 is currently out of stock at my supplier, could be a  CD74AC08E an option? Since i only use this for this project there is no need to be more flexible, but i get the idea why you stock NAND and NOR gates. Would it be necessary to tie the unused gates to either +V of 0V with this too? How about the series gate resistor?


I also like the option with the two diodes, since it seems even more simple. Just cant get my head around how this should work. Arent the diodes just blocking the signal? Also wouldnt it draw the current for the coils from the TTP223? Or should be the source on GND?   I only can find 1N 4148, withoud the D.

[pcprogrammer]

I also like the option with the two diodes, since it seems even more simple. Just cant get my head around how this should work. Arent the diodes just blocking the signal? Also wouldnt it draw the current for the coils from the TTP223? Or should be the source on GND?   I only can find 1N 4148, withoud the D.

The D in the 1N4148 is not of importance. Just a manufacturers addition to indicate what package is used. The diode could be any fast signal schottky diode.

The way the diode AND works is that when one or both TTP223 outputs are low the gate of the fet is also low, due to a small current flowing through the resistor R1 and the diodes that have there cathode pulled low.

If both TTP223 outputs are high the diodes will not conduct and the gate of the fet will be pulled high via R1, because no current will be flowing.

The current throug R1 can be calculated as follows: ((5V - Vdfwd) / R1) = (5 - 0.2) / 1000 = 0.0048A, so 4.8mA. (The 0.2 forward voltage of the 1N4148 is a typical value and might be higher) If this current is found to be to high, R1 can be of higher value, but a to high value will make the fet open more slowly due to the gate capacitance.

To block the battery voltage flowing into the TTP223 circuits you can move D4 to the output side of where the coils are connected, just like in your own wiring diagram.

[Zero999]

I had to look up "push-pull" and i have absolutely no clue..

It is not that hard. A push pull output has a transistor for both levels, so a low is a "hard low" and a high is a "hard high". This is not the case for an open drain or open collector output, where there is only one transistor. In this case only a low is a "hard low" and to make a high a pull up resistor is needed. (Or vice versa, but that is not commonly used)

I guess there are MANY ways wich get the job done, i just try try to figure one out that's easy to understand and not likely easy to fail.

Yep, it is also possible to make the AND function with two diodes and a pull up resistor.

That schematic contains a couple of errors which I've fixed.

The source of Q1 must go to 0V, not the source.

The 1N4148 is an ordinary silicon diode, not a Schottky.
https://www.vishay.com/docs/81857/1n4148.pdf

I also like the option with the two diodes, since it seems even more simple. Just cant get my head around how this should work. Arent the diodes just blocking the signal? Also wouldnt it draw the current for the coils from the TTP223? Or should be the source on GND?   I only can find 1N 4148, withoud the D.

The D in the 1N4148 is not of importance. Just a manufacturers addition to indicate what package is used. The diode could be any fast signal schottky diode.

The way the diode AND works is that when one or both TTP223 outputs are low the gate of the fet is also low, due to a small current flowing through the resistor R1 and the diodes that have there cathode pulled low.

If both TTP223 outputs are high the diodes will not conduct and the gate of the fet will be pulled high via R1, because no current will be flowing.

The current throug R1 can be calculated as follows: ((5V - Vdfwd) / R1) = (5 - 0.2) / 1000 = 0.0048A, so 4.8mA. (The 0.2 forward voltage of the 1N4148 is a typical value and might be higher) If this current is found to be to high, R1 can be of higher value, but a to high value will make the fet open more slowly due to the gate capacitance.

To block the battery voltage flowing into the TTP223 circuits you can move D4 to the output side of where the coils are connected, just like in your own wiring diagram.

Being a standard silicon diode, rather than a Schottky, the 1N4148 has a voltage drop of around 0.7V, together with the internal voltage drop of the TTP223, the low voltage will be about 0.8V, which is fine in this case, but could be an issue, if the MOSFET is a logic level type, with a  much lower threshold voltage.

Where did you hear pull-down resistors are necessary? The main reason to use them is in the case of a microcontroller driving a MOSFET, ...

Series gate resistors are generally recommended, if the MOSFET is being driven from a low impedance source, which can cause oscillation, due to the parasitic gate inductance. In my circuit, they are also unnecessary, because the 74HC00 has an output impedance of around 50 ohm which is high enough to damp any oscillations.

The SN74LVC1G08 is a good choice. I only suggested the 74HC00 because being through hole, is easier to solder and I only stock NAND and NOR gates, because they're more flexible.

It could be i heard and red this in case of using a microcontroller, makes sense.
The SN74LVC1G08 is currently out of stock at my supplier, could be a  CD74AC08E an option? Since i only use this for this project there is no need to be more flexible, but i get the idea why you stock NAND and NOR gates. Would it be necessary to tie the unused gates to either +V of 0V with this too? How about the series gate resistor?


I also like the option with the two diodes, since it seems even more simple. Just cant get my head around how this should work. Arent the diodes just blocking the signal? Also wouldnt it draw the current for the coils from the TTP223? Or should be the source on GND?   I only can find 1N 4148, withoud the D.

The 74xx08 is fine. It gives you three spare gates. I would still go for the 74xx00, because the circuit isn't any more complex and you might find you need to order more than one, in which case, the surplus will be more useful for future projects.

The AND diode is also a good idea and simpler, so go with that. Hopefully, it now makes more sense, now I've posted the corrected schematic.

[pcprogrammer]

Replace the 1N4148 with a BAT85 if the 1N4148 does not do the job. https://www.vishay.com/docs/85513/bat85s.pdf That for sure is a Schottky diode. Vfwd depends on the current through the diode, and if still to high with the 1kOhm resistor, try it with 2k2 or 3k3.

@Zero999, my bad, you are right about the diode, but I don't see any difference between your MOSFET drawing or mine. Body diode pointing up, and arrow pointing toward the gate in both cases. https://www.vishay.com/docs/91327/irlz34.pdf

[Zero999]

@Zero999, my bad, you are right about the diode, but I don't see any difference between your MOSFET drawing or mine. Body diode pointing up, and arrow pointing toward the gate in both cases. https://www.vishay.com/docs/91327/irlz34.pdf

You connected the gate to the source, so it can't possibly work. Look at the two schematics side-by-side.

Would it be necessary to tie the unused gates to either +V of 0V with this too? How about the series gate resistor?

I missed this. Yes, connect all unused inputs to either power supply rail. This is the case for all CMOS gates, which don't have built-in pull-up resistors.

Series gate resistors are recommended for MOSFETs driven from low impedance output gates such as the 74AC08, but unnecessary for the 74HC00 or CD4011, which have higher impedance outputs. Oh and I recommend 74HCxx or CD4xxx for this kind of application, especially for beginners. The 74AC and 74LVC logic families are much faster and can create problems with ringing and oscillation, unless careful attention is paid to the layout.

[pcprogrammer]

You connected the gate to the source, so it can't possibly work. Look at the two schematics side-by-side.

Ah yes, that is what you get when you willy nilly draw a schematic. 

Also falls under the category one sees what one wants to see. I honestly did not see it until you posted them side by side.