Tired of manual switches? Simple Touch switch pass thru for under $15.00
- Thread starter Scubabatdan
- Start date
- th_trl_thread_readers 0
- Status
Just looking over this... I notice the IRF8736PbF is a dual mosfet, but only one source, gate, and drain is used. Is there any reason why you couldn't substitute an IRLZ34N or IRL3103 in its place?
It is not really a dual mosfet . The drain and source pins are connected. To get the full amps and maximum heat dispersion all the pins should be used. To answer your question the AT42QT1010 touch sensor chip should be able to drive the gate of any mosfet you want to use.
Got some AT42QT1010 in my cart at digikey. Going to try one of these with a P channel mosfet . With my project I need to switch the Positive side.
Ahh, after going back over that part number I see what you mean about it not actually being a dual mosfet. For our purposes though I don't think we need all the extra current to drive an n-channel mosfet.
Since you brought it up, could you suggest a good p-channel? I have been working on a few projects with the ptn04050C booster, which doesn't play nice with n-channels. I found one a couple weeks ago that had perfect specs, but I didn't pay attention to the measurements. When I got it in the mail the FET was about the width of 3 human hairs, and had 8 tiny square dots on the bottom...I'm not THAT good with a solder gun LOL.
I am going to use some of these. FDS6679AZ Fairchild Semiconductor MOSFET PowerThey are protected agaist static shock up to 6kV. CraigHB point them out in another thread. Going to be another part needed if you plan on using this touch sensor with a P Channel besides what Dan has for the NChannel. Something like a 2N3904 is needed i believe for Pchannel. But i need more coffee before i figure this all out.
SHUCKS!
I just ordered the parts to do the touch sensor, and THEN read further into the thread.... capacitive touch...NICE.
But i'll try that after i get the resistive touch sensor right.
Thanks Dan (and everyone else)!
I just ordered the parts to do the touch sensor, and THEN read further into the thread.... capacitive touch...NICE.
But i'll try that after i get the resistive touch sensor right.
Thanks Dan (and everyone else)!
Any luck with this? Would love a few more specifics on using this with a P-channel.
For a P-channel, you use a pull-up resistor (connects gate to source voltage) with the FET on the high side. You ground the gate with the touch sensor. With an N-channel, you use a pull-down resistor (connects gate to ground) with the FET on the low side. You provide power to the gate with the touch sensor.
High side (P-channel) is usually better because you don't have to concern yourself with any ground offset caused by a low side voltage drop across a FET. Ground offsets can cause problems for chips like regulators. For a simple mod with no additional electronics, it doesn't matter, N or P-channel is going to work equally well. Even with additional electronics, you can still use an N-channel, but you have to be careful not to connect any grounds after the FET. That can be tricky with traditional circuit design.
High side (P-channel) is usually better because you don't have to concern yourself with any ground offset caused by a low side voltage drop across a FET. Ground offsets can cause problems for chips like regulators. For a simple mod with no additional electronics, it doesn't matter, N or P-channel is going to work equally well. Even with additional electronics, you can still use an N-channel, but you have to be careful not to connect any grounds after the FET. That can be tricky with traditional circuit design.
I put the ATMEL AT42QT1010 captive touch circuit on the back burner for now. There is some metal around where i want to put it so I went with just a regular touch switch with the Pchannel. If you take the first circuit posted in this thread and just switch the polarity of the wires it will work with a P channel mosfet. The FDS6679AZ worked well for me using a 2Mohm Pull up resistor. But i didnt run a 47k on the gate input. I ordered up some 15kV TVS diodes to protect the gate of the mosfet even further from ESD events. The FDS6679AZ has 6kV built in. Not sure if the TVS diode is going to effect the gate so i might need to change the value of the resistor. I ran a uA test with the 2Mohm and the mosfet was fully unlatching so i was getting very little leakage threw the fet.
Might be awhile, but i will work on the ATMEL AT42QT1010 again at some point. Here is a basic touch with the FDS6679AZ Pchannel
Last edited:
There is a little less concern with something handheld for ESD protection. You have to make a circuit for the static so the mod would have to be in contact with some sort of ground to allow a static charge to pass into the gate. The 6kV offered by the MOSFET's internal clamping didoes seems adequate to me, but there's certainly nothing wrong with beefing it up. You never know, you might just inadvertantly pick if up off the floor with a big charge in your body and your finger in contact with the switch. Though, a 15kV static discharge is pretty big. Those are the ones that make a noticeable sound and even cause a little pain when you touch a doorknob or something.
Would that same P-channel work with only a 3.7v input, for instance to power the ptn04050C booster?
I put the ATMEL AT42QT1010 captive touch circuit on the back burner for now. There is some metal around where i want to put it so I went with just a regular touch switch with the Pchannel. If you take the first circuit posted in this thread and just switch the polarity of the wires it will work with a P channel mosfet. The FDS6679AZ worked well for me using a 2Mohm Pull up resistor. But i didnt run a 47k on the gate input. I ordered up some 15kV TVS diodes to protect the gate of the mosfet even further from ESD events. The FDS6679AZ has 6kV built in. Not sure if the TVS diode is going to effect the gate so i might need to change the value of the resistor. I ran a uA test with the 2Mohm and the mosfet was fully unlatching so i was getting very little leakage threw the fet.
Might be awhile, but i will work on the ATMEL AT42QT1010 again at some point. Here is a basic touch with the FDS6679AZ Pchannel
Since am going to have on-board charging it will be connected to a ground at those times. Even so it is pretty unlikely the ESD discharge will hit the touch pin ( gate ) . More then likely going to go to the grounded case unless you pick it up just right. I guess it would be a one in a million shot. The TVS diodes are 0402's Never messed with anything that small yet i can do 0603's no problem. Got my order a few hours ago those jokers are small. But for 42 cents the added protection is worth it to me considering the PV wont work if by some chance the gate did get fried. Really wanted to add 30kV protection but there is no way in hell i can solder 0201's at least not without a microscope and even then i don't think my hands would be steady enough. But that's just my style i over kill the stuff i build that i know i am going to be using a lot.
I think it would work might be a bit touchy on the latching at that voltage. Might want to go with a 3 or 4 Mohm resistor. I can check it on the bread board and see how it acts at 3.7 volts with 2Mohm. I will report back in a bit with results
Actually i would go with a 10Mohm resistor @ 3.7 volts. 2Mohm wouldnt turn it on most of the time with a touch @3.7volts.
EDIT: Yeah for sure a 10Mohm it will still work decent down @2.75volts which would be the cutoff for a protected cell
EDIT: Yeah for sure a 10Mohm it will still work decent down @2.75volts which would be the cutoff for a protected cell
Last edited:
I hear ya, the 0603 package is about my limit at this point. I try to avoid chips with .5mm pin pitches, but I can to them if I have to. .65mm is my preferred minimal pin pitch. Reflow soldering makes that somewhat moot, but like I've said before, I'm stuck in a rut on the hand soldering.
For a single battey mod, the FDS6679AZ might get pushed a little close to the active region (Vgs too low). That's not good, makes a lot of heat and can burn out the FET. You want a 2.5V specified MOSFET for that. I'm using the FDS6576, but it's not for a touch switch and that FET is not protected. You can either add a regular 10V zener diode to protect the gate or find a simlar part with the clamping diodes built-in.
For a single battey mod, the FDS6679AZ might get pushed a little close to the active region (Vgs too low). That's not good, makes a lot of heat and can burn out the FET. You want a 2.5V specified MOSFET for that. I'm using the FDS6576, but it's not for a touch switch and that FET is not protected. You can either add a regular 10V zener diode to protect the gate or find a simlar part with the clamping diodes built-in.
Yup thats a good point you really dont want to use a FDS6679AZ below a 4.5 gate voltage. I would say slap a TVS diode between the gate and source just like a resistor for the FDS6576 but for our voltages they are small as hell. I guess a zener might be a ( bigger) better package.
A TVS diode is the best thing to use. They're specifically designed to protect against ESD and voltage spikes. A plain ol' zener will provide some protection against ESD, they're typically good for 16kV, at least once. Zeners are cheaper and more readily available. TVS diodes are usually larger and more expensive and you won't find them at your local Radio Shack. They're designed for multiple hits with much larger silicon junctions. A regular zener will break down after a few big strikes.
I dont know about larger here is a 30kV thats in a 0201 package.Digi-Key - F3572CT-ND (Manufacturer - SP1005-01WTG) Maybe I missed something but i didn't find any larger ones i though would work at my voltages. I for sure would rather solder something bigger then a 0402.
It depends on what kind of protection you're looking for. I use the 400W ones to protect against both power spikes and ESD on power inputs, I use an SMAJ6.0A for power input protection on my mod's charging circuit.
For things like tactile switches, the chips they connect to already have an amount of ESD protection plus the debouncing circuit for the tactiles offers an additional level of protection. So, I'm usually only interested in both protection against ESD and power spikes.
I've not yet used a TVS diode solely for protection against ESD. For the rare occaisions I've needed it, I've used plain ol' zeners. The small SMD ones may in fact be the best solution. Thanks for pointing that out.
Oh, BTW, the voltage rating on your TVS only needs to cover the limit of your part. So, if you have a 25V max Vgs, a TVS with a 20V working voltage should cover it. So, that may open up your options on finding a SMD TVS in a 0603 or larger package. Also, you don't need a dual sided one, just single sided. Since you would never apply a negative voltage, you don't need a clamping diode for the negative direction. The two sided ones are usually only required for AC inputs.
For things like tactile switches, the chips they connect to already have an amount of ESD protection plus the debouncing circuit for the tactiles offers an additional level of protection. So, I'm usually only interested in both protection against ESD and power spikes.
I've not yet used a TVS diode solely for protection against ESD. For the rare occaisions I've needed it, I've used plain ol' zeners. The small SMD ones may in fact be the best solution. Thanks for pointing that out.
Oh, BTW, the voltage rating on your TVS only needs to cover the limit of your part. So, if you have a 25V max Vgs, a TVS with a 20V working voltage should cover it. So, that may open up your options on finding a SMD TVS in a 0603 or larger package. Also, you don't need a dual sided one, just single sided. Since you would never apply a negative voltage, you don't need a clamping diode for the negative direction. The two sided ones are usually only required for AC inputs.
Last edited:
- Status
Similar threads
