# Why do you need a conductive layer with conductivity of $~10^4$ S/cm in a capacitance touchscreen?

ITO has has conductivity of $~10^4$ S/cm. A capacitive touchscreen measures the change in capacitance or electric field.

1) Why do you need a conductive layer in a capacitance touchscreen in the first place, if you just measure the capacitance or electric field?

2a) Does it really need to be as low as $10^4$ S/cm or can it also be something like $10^3$ S/cm or $10^2$ S/cm?

2b) What happens when the conductivity gets lower?

## 1 Answer

Considering that silver has a conductivity of 62 million siemens per metre, something that has a conductivity of 10 thousand siemens per cm (100 siemens per metre) isn't that great. Compare it with salt water at about 5 siemens per metre and it starts to look OK but inside a capacitive touch screen the conductors are very thin printed tracks with a print width of a few thousandths of an inch at best: -

The picture is stolen from here and as you can see, the cap screen needs an XY matrix of conductors separated by a thin layer. If you start to reduce the conductivity of these conductors you will get to a point where the XY matrix becomes ineffective and it doesn't work.

How much you can lower the conductivity is something that is difficult to say. If the overall resistance of the XY conductors starts to approach the capacitive reactance decrease due to your finger on the screen then reliable operation is going to be affected. However, it seems that even with a few pF extra when the finger is placed on the screen, to not detect this reliably would mean a conductivity that is hundreds of times lower than 100 siemens per metre.

• But why does it become ineffective and stop working when the conductivity is reduced? Is it the extra heat or noise that is produced by resistance? How do you get to that number of hundreds of times? – Per Jan 19 '17 at 12:10
• I've explained why it becomes ineffective - it becomes harder to detect a change of capacitance because resistance starts to approach a value that swamps the change in capacitance. Noise will always play a role and so will capacitive coupling between each line of the matrix and this will muddy the position of where the finger is believed to be. Hundreds of times is just an estimate based on what switching frequencies might be involved and how much capacitance is brought into play by the finger. – Andy aka Jan 19 '17 at 12:23
• @Andy_aka: A higher resistance would increase the RC time constant, but since you only have a few pF ($10^{−12}$ F ) as capacitance, this RC time constant would still be small for resistance as high as 100KΩ Since you are measuring a change in capacitance, why would you care about resistance, if the time delay would be negligible? Numerous different methods exist to measure capacitance: shift of resonance frequency, frequency modulation, amplitude modulation, charge time measurement, time delay measurement, duty cycle, etc. – Per Jan 21 '17 at 0:57
• The relevance of the other methods is? – Andy aka Jan 21 '17 at 9:21
• All of the methods are to measure changes in capacitance that can be found in the literature, or here: electronics.stackexchange.com/questions/234201/… . Question was: Since you are measuring a change in capacitance, why would you care about resistance, if the time delay would be negligible? – Per Jan 21 '17 at 14:26