Capacitive touch with MTCH101

I’m developing a product with two touch buttons. The pads used for sensing is on a circuit board which is roughly 4 mm from the touch surface (glass between). The pads are directly connect with a chip called MTCH101. Datasheet. This gives a signal out when capacitance is building up on the pad. I bought an evaluation kit with this chip (as it’s really small and in an smd package), and it came with a sensor board included. Datasheet. The chip triggers when my hands are many centimeters away, which seems logical as the pad on the board (circle) is very big. My issue is that I have no way of knowing how small to make such a pad on my own circuit board to get it to react when my finger is touching the glass (exactly 4 mm away). The capacitance threshold value for the chip input isn’t specified in the datasheet, so I can’t calculate and decide. Any idea about how to go about calculating the pad size in order to get this chip to trigger when my finger is 4 mm away from the pad?

I played around with that IC a little while back. If I recall correctly, I was using a brass plate around 3/4in/sq and played around with the resistor values to reliably get what I was looking for.

You can experiment with sizes and pot settings to get what you are looking for.

Hi,
With some header pins and some PCB or cardboard and aluminium foil, make some test touch plates.


Tom... :slight_smile:

TomGeorge:
Hi,
With some header pins and some PCB or cardboard and aluminium foil, make some test touch plates.


Tom... :slight_smile:

I fumbled around with some aluminum foil some hours the other day. A very difficult process, but what I figured out was that the chip was still way too sensitive. It has some very good noise filtering...

tinman13kup:
I played around with that IC a little while back. If I recall correctly, I was using a brass plate around 3/4in/sq and played around with the resistor values to reliably get what I was looking for.

You can experiment with sizes and pot settings to get what you are looking for.

You mean putting resistors between the pad and the chip input?

PelleS:
You mean putting resistors between the pad and the chip input?

No. There is a divider for sensitivity on the MTSA pin of the chip. I bought the chip, not the eval board. It would correspont to the pot on your board.

From the MTCH101 datasheet

The sensitivity of the system determines how far and fast it can respond to proximity or touch. The MTCH101provides the MTSA pin to adjust the sensitivity, and the voltage on this pin will determine the sensitivity. VDD voltage will give the lowest sensitivity, while GND voltage will give the highest sensitivity.

In other words, connecting MTSA to the same power you are feeding the chip gives the least sensitivity. Coupled with a smaller pad it may be what you are looking for.

tinman13kup:
No. There is a divider for sensitivity on the MTSA pin of the chip. I bought the chip, not the eval board. It would correspont to the pot on your board.

From the MTCH101 datasheet

The sensitivity of the system determines how far and fast it can respond to proximity or touch. The MTCH101provides the MTSA pin to adjust the sensitivity, and the voltage on this pin will determine the sensitivity. VDD voltage will give the lowest sensitivity, while GND voltage will give the highest sensitivity.

In other words, connecting MTSA to the same power you are feeding the chip gives the least sensitivity. Coupled with a smaller pad it may be what you are looking for.

Yes, I know, but with the sensitivity all the way down, it can still sense my finger centimeters away. I also used an aluminum foil piece smaller than 0,5 cm without getting the sensitivity needed.

A capacitive switch will never trigger at exactly a touch. The capacitance depends on too much - Is the circuit connected to an Earth connection? Sitting on a metal table, or wood/plastic? Are you holding the device in your hand? How fat are your fingers?

A guard ring connected to circuit ground/common around the pad may help reduce sensitivity to distant fingertips, especially off-axis fingers.

You cannot just test this with pieces of foil. You =must= have the glass on top, because it has a higher dielectric constant than air. This will actually help you, as the amount of capacitance should increase more quickly as your finger comes in contact with the glass.

According to the datasheet, without ground near the sensing pad, expect a sense distance about double the diameter of the pad.

Page 9, figure 7.0

Capacitive sensor design:

polymorph:
A capacitive switch will never trigger at exactly a touch. The capacitance depends on too much - Is the circuit connected to an Earth connection? Sitting on a metal table, or wood/plastic? Are you holding the device in your hand? How fat are your fingers?

A guard ring connected to circuit ground/common around the pad may help reduce sensitivity to distant fingertips, especially off-axis fingers.

You cannot just test this with pieces of foil. You =must= have the glass on top, because it has a higher dielectric constant than air. This will actually help you, as the amount of capacitance should increase more quickly as your finger comes in contact with the glass.

According to the datasheet, without ground near the sensing pad, expect a sense distance about double the diameter of the pad.

Page 9, figure 7.0
http://ww1.microchip.com/downloads/en/DeviceDoc/40001664B.pdf

Capacitive sensor design:

http://ww1.microchip.com/downloads/en/AppNotes/01492A.pdf

Thank you for the information. I've read the last document a few times before. I've used glass of the exact thickness all the time while testing. I saw Microchip advertising with this being used specifically for switches. Why do you say "A capacitive switch will never trigger at exactly a touch."? I have a lot of devices which does exactly this, but I am unsure on how to go about this in the best way. I'll be using grounding around the pad, but I still need to be able to get an estimate of the diameter before I start ordering circuit boards.

Is there anything more I can do to figure this out? I have read all the documents I can find, but I am not able to recreate any of the touch functionality I find in other products with capacity buttons.

Hi,
Have you constructed a pad with the gnd guard?
How long is the wire from pad to the controller board?

Can you post a picture of your project so we can see your component layout?

Tom... :slight_smile:

It never triggers at exactly a touch, because what is a touch? When your finger is on the other side of a sheet of glass, it never truly touches. To be a capacitive switch, you never actually make contact with the metal.

Ideally, what you want is something that requires you press your finger down a bit into the glass. My wife has the bad habit of pointing her finger at my GPS while I’m driving, and she’ll get her finger within a few mm of the surface. At that point, my GPS capacitive touchscreen decides someone is tapping and reacts accordingly.

Capacitive coupling goes up exponentially as a finger is brought near the touch screen. However, you can modify this by changing the geometry of the sensing area and guard rings around it. You want to make that transition more abrupt.

How about all of the questions you've not answered?

Sorry for replying late. Been drowning in work. I just found it difficult to get the capacitance at the right value. That's why I wonder how other people manage to do this. I know all the formulas and the theory, but there must be some tricks I haven't picked up on. For instance, I have a kitchen cooktop from Ikea that has capacitive sensor/buttons for adjusting the heating. It works great for people with big fingers, small fingers etc., and it doesn't trigger if I hold a plate of metal right above the sensors. How do they do that? The capacitance should increase a lot when placing something with a much bigger area just above the sensors.

You have discovered that a capacitave switch is not an easy thing to design. The problem is the grounding and all the stray capacatance associated with the circuit.

Many beginners are puzzled when they find their switches will not work when they are disconnected from a computer and just run off an external power supply or batteries.

The devices you have seen have had hundreds of man hours devoted to their design and development, this is not something you can just knock up and expect it to work reliably.

You still haven't answered the questions posed of you.

polymorph:
A capacitive switch will never trigger at exactly a touch. The capacitance depends on too much - Is the circuit connected to an Earth connection? Sitting on a metal table, or wood/plastic? Are you holding the device in your hand? How fat are your fingers?

A guard ring connected to circuit ground/common around the pad may help reduce sensitivity to distant fingertips, especially off-axis fingers.

You cannot just test this with pieces of foil. You =must= have the glass on top, because it has a higher dielectric constant than air. This will actually help you, as the amount of capacitance should increase more quickly as your finger comes in contact with the glass.

According to the datasheet, without ground near the sensing pad, expect a sense distance about double the diameter of the pad.

Page 9, figure 7.0
http://ww1.microchip.com/downloads/en/DeviceDoc/40001664B.pdf

Capacitive sensor design:

http://ww1.microchip.com/downloads/en/AppNotes/01492A.pdf

The device is mounted on the wall, so most likely wood. It is to be used by anyone with fingers, so not just fat or skinny fingers. I know the dialectric constant of glass and how that affects the capacitance. The sensor was so sensitive that it didn't matter at the time. I get the impression that it's close to impossible for someone without years of experience to design such a sensor, so I'll see if I can pay someone with more experience to do it instead.