Guilty but, I'm working on it and I appreciate all you knowledgeable folks helping out.
MOSFET question for tactile switch
- Thread starter Dougiestyle
- Start date
- th_trl_thread_readers 0
-
- Tags
- madvapes
- Status
Guilty but, I'm working on it and I appreciate all you knowledgeable folks helping out.
Ok, my understanding is that you want to drive the gate at full voltage for efficiency. The gate ratting will determine how much amperage it draws and is what you want to use to determine the mosfet you want to use to limit the current across the switch.
Did I get that right?
This is how a P channel would be used if my assumption is correct.
Edit: Fixed image
Edit: Added pull up resistor
Did I get that right?
This is how a P channel would be used if my assumption is correct.
Edit: Fixed image
Edit: Added pull up resistor
Last edited:
Ok, my understanding is that you want to drive the gate at full voltage for efficiency. The gate ratting will determine how much amperage it draws and is what you want to use to determine the mosfet you want to use to limit the current across the switch.
Did I get that right?
This is how a P channel would be used if my assumption is correct and flip the battery for an N channel.
Perhaps I am looking at it wrong, but I think you need the switch between the gate and ground for p-channel. With a p channel you want the gate voltage to be less than source voltage to open the gate. For n channel the source would be hooked to ground and gate would have a switch connected to positive. N channels need a voltage higher than source to open the gate.
I do believe you are correct. I was drawing in a hurry and not paying much attention to the specifics. Re-drawing now and will edit post.
I'm sure you would make something better than anything she can get off the shelf. OTOH.... There is a neat one that has the "off the shelf" finished look and uses 14650's:
Titanium E-Power Replacable Battery kit!
in case you haven't seen it. No boost tho.......but it is true 3.7. Haven't seen any reviews... like how the button is, etc. Just an FYI.
Last edited:
Sounds like you getting it for the most part. For FET switches, higher gate voltage is better. But, you also have to look at the specs on your FET. Some are what they call "compatible with CMOS logic" and can be more sensitive with lower maximum allowable gate voltages. Vgs as listed in the data sheet is essentially your gate voltage. For an N-channel, it's the voltage you put on the gate with respect to ground. For a P-channel, it's the voltage on the gate with respect to the source or power supply. Rdson is the resistance the current sees in going through the FET. Higher Vgs = lower Rdson and that's good. Less resistance = less heat and less power wasted.
Your P-channel circuit looks good except you're missing a pull-up resistor on the gate (4.7kΩ is fine) . The resistor needs to go were the tactile switch is and the tactile switch needs to go between the gate and ground. With the switch relaxed, the pull-up resistor makes the gate voltage equal to the source voltage or Vgs=0 and the FET is off. When the switch is pressed, gate voltage is 0 and source voltage is battery voltage. Vgs = battery voltage and the FET is on.
Thanks for that. It may be a stop gap for me until I can come up with a small LiPo powered boost mod for my wife. It's going to be tough because I need to go super small on the board and components, but it's a challenge and I'm all for it. The one I use now has a big board with liberal space so the layout wasn't terribly hard.
Last edited:
If she really likes the ego's I just bought a couple of NoEgo's from Jazz Carto Pipes. They are a bit bigger than an ego with the same look. They are basically an e-power type mod. They have a voltage cut off so the batteries don't drop below 3.6V. This reduces the usable vape time but you get the 4.2V through 3.6V. I have been real happy with them and they look real nice. He is currently working on either a VV version or regulated to voltage of choice switch bodies that will be able to be purchased separately.
I realized I missed the pull up or down resistor after reviewing a couple of circuits. I will edit my picture.
Wow, thanks for your input! I suppose getting a college degree in engineering wouldn't be such a bad idea before I start making some hacked up mods. The FET I'm using is a CSD16325Q5 and it's in a simple circuit with two 3.7V 14500 batteries in parallel. On the breadboard, I was reading 4.11 at the tactile switch and was under the impression that in order to preserve the tiny switch, I needed to reduce the current going through it. I applied several resistors until I got to the 47K which brought it down to 1.85V at the switch, which I thought was awesome! Guess not, eh?
Last edited:
Need a second opinion, would either of these mosfets work, with an input voltage being ~3.7v. I'd be using it with a boost circuit.
http://focus.ti.com/lit/ds/symlink/csd25302q2.pdf
http://focus.ti.com/lit/ds/symlink/csd25401q3.pdf
http://focus.ti.com/lit/ds/symlink/csd25302q2.pdf
http://focus.ti.com/lit/ds/symlink/csd25401q3.pdf
Craig, So we should be looking for the datasheet spec to show LOW RDS(on) resistance. I'm using a P-Channel that states: TYPICAL RDS(on) = 0.18 Ω where as I've seen other mosfet with the RDS(on) rated in the Meg ohms.
Or: perhaps it's mOhms meaning milli ohms.
I did see one mosfet specifications as 75 mOhms but inside the datasheet it said 0.075 ohms
Last edited:
I do think they would work but not at their peak efficiency. If I am reading the data sheet correctly the ideal gate voltage on them is 4.5V or above. And I make no claim that I am reading the data sheet correctly. =] I usually have to read the 5 or 6 times to get it right.
Also I do believe it is milli ohms as it is a small M.
Also I do believe it is milli ohms as it is a small M.
That's what I thought. Tring to find a p channel FET from TI, so I can just grab a sample for this one time venture.
Always consult your data sheet. My previous comments are based on values you see with the more common parts.
@Solder-Meister, that FET is "CMOS logic compatible". It has a low absolute maximum Vgs. You're getting close to its maximum with two 3.7V Li-Ion batts. You can use a divider to reduce Vgs, but I would just use a FET with a higher rating.
When they give absolute maximum Vgs it is the absolute maximum and the part will fry if you go over that. The gate junction on any FET is delicate. They can can be damaged easily by ESD, doesn't even have to be enough to make a spark. It makes a good case for using the protected ones. It's usually not an issue once they're in a completed circuit, but you can easily damage them with static by handling them. Always use static protection when handling unprotected FETs. The protected ones have clamping diodes built in that make them much more robust to ESD and you can handle those more carelessly.
For the unprotected FETs in a circuit that could be vulnerable, it's not a bad idea to throw an appropriately rated zener diode between the gate and source. Zeners typically have a 16kV human body model tolerance to ESD. That's a pretty good static shock, like the ones that make a visible spark and a noise.
I typically use logic level FETs only for small signal applications. For power FETS, I try to use ones with an absolute maximum Vgs 12V and above for robustness to power spikes. Power FETs with a 12V or 20V max Vgs are probably most common and you can find lots or them with low Rdson and low gate source threshold voltage (gate voltage where they first start to turn on).
Switch current through your tactile switch is solely a matter of your pull-down or pull-up resistor value and source voltage. You can go as low as 4.7k for a tactile switch, but 10k, 33k, or 47k are also fine. The advantage of a lower resistance is less susceptibility to noise. The disadvantage is it results in more current flow through the switch and more power consumption.
@Solder-Meister, that FET is "CMOS logic compatible". It has a low absolute maximum Vgs. You're getting close to its maximum with two 3.7V Li-Ion batts. You can use a divider to reduce Vgs, but I would just use a FET with a higher rating.
When they give absolute maximum Vgs it is the absolute maximum and the part will fry if you go over that. The gate junction on any FET is delicate. They can can be damaged easily by ESD, doesn't even have to be enough to make a spark. It makes a good case for using the protected ones. It's usually not an issue once they're in a completed circuit, but you can easily damage them with static by handling them. Always use static protection when handling unprotected FETs. The protected ones have clamping diodes built in that make them much more robust to ESD and you can handle those more carelessly.
For the unprotected FETs in a circuit that could be vulnerable, it's not a bad idea to throw an appropriately rated zener diode between the gate and source. Zeners typically have a 16kV human body model tolerance to ESD. That's a pretty good static shock, like the ones that make a visible spark and a noise.
I typically use logic level FETs only for small signal applications. For power FETS, I try to use ones with an absolute maximum Vgs 12V and above for robustness to power spikes. Power FETs with a 12V or 20V max Vgs are probably most common and you can find lots or them with low Rdson and low gate source threshold voltage (gate voltage where they first start to turn on).
Switch current through your tactile switch is solely a matter of your pull-down or pull-up resistor value and source voltage. You can go as low as 4.7k for a tactile switch, but 10k, 33k, or 47k are also fine. The advantage of a lower resistance is less susceptibility to noise. The disadvantage is it results in more current flow through the switch and more power consumption.
Do the Fet's that show a diode between source and drain have an effect on our application. Most of the Pchannel ones I have found on Digi Key that are cheap show this diode.
That's called the body diode and is actually not a discrete component of the chip. It's naturally formed by the junction of the two types of semiconductors (the P and N substrates) and the center connection you see identified in the part schematic. That connection is there to keep the substrate from floating. A floating substrate can cause the transistor to latch under certain conditions.
In some cases, you are interested in the behavior of the body diode. For example, synchronous switching regulators typically use a P-Channel FET in lieu of a Schottky diode as a rectifier (notably increased efficiency). In that case, you rely heavily on the body diode to conduct forward surge current and generally have to pick a FET with favorable body diode characteristics. Other applications like heavy motor control also rely on a power FET's body diode. Atypically, some specialized FETs do not have that internal connection, instead running it to a pin which allows them to be utilized with or without a body diode.
Protected FETs will show an additional TVS (transient voltage suppressor) type diode between the gate and source. That one is a discrete component on the chip.
In some cases, you are interested in the behavior of the body diode. For example, synchronous switching regulators typically use a P-Channel FET in lieu of a Schottky diode as a rectifier (notably increased efficiency). In that case, you rely heavily on the body diode to conduct forward surge current and generally have to pick a FET with favorable body diode characteristics. Other applications like heavy motor control also rely on a power FET's body diode. Atypically, some specialized FETs do not have that internal connection, instead running it to a pin which allows them to be utilized with or without a body diode.
Protected FETs will show an additional TVS (transient voltage suppressor) type diode between the gate and source. That one is a discrete component on the chip.
Ummmm..... Ok..... So I understood most of what you said. Surprisingly more than I probably should have. Please have pity on us poor technicians. We can tell when it's bad and replace it but we don't always know the specifics of what it's doing.
So does that mean it will have an effect on our use as a switch? =] I guess I should have asked - should we use them or stay away from them?...
Sorry, I hate to insult people by talking at a level too low for them. There's always Google if you don't understand a term. I still have to do that when reading technical articles sometimes. One got me just recently, what was it, oh, parasitic impedance. I knew what that meant at one point, but I forgot when I was looking at some technical paper a while back. Hehe, sometimes I think I've forgotten more than I've learned.
Anyway, the ones with the body diode are the ones you normally use. If it doesn't have one, it was either neglected in the symbol diagram or it's one of the 4 pin type of FETs you wouldn't normally use. Basically, any 3 pin FET has one whether it was printed in the symbol or not.
Anyway, the ones with the body diode are the ones you normally use. If it doesn't have one, it was either neglected in the symbol diagram or it's one of the 4 pin type of FETs you wouldn't normally use. Basically, any 3 pin FET has one whether it was printed in the symbol or not.
That clears it up. I was not aware that most of them actually have it built in. That's why when I saw it in the data sheet it kinda worried me a little. The schematics I have dealt with don't show them with the diode and I have had limited exposure to them from a design standpoint. I have replaced many, just haven't delved too deep in how they work beyond the basics.
It is a fine line between talking down and talking over =] and it all depends on who your talking to. =]
Now that I know it's a common design characteristic I will research a little deeper into the implementation.
- Status
Similar threads
