2
\$\begingroup\$

1I'm using class 2 medical graded SMPS in our device. So earth connection not recommended. When connecting SMPS to load (contains capacitive touch LCD), touch function is affected by SMPS noise. When connecting SMPS earth point to protective earth, LCD touch function working properly.

I don't want to use earth connection because of class 2 constraints. I want to bypass that noise. I have added few decoupling ceramic capacitor at touch input supply to avoid high frequency noise to touch input. Still am facing issue.

Can anyone give me some idea to overcome above issue ?

SMPS Datasheet

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Why do you think the noise is from the SMPS? Have you put the capacitor across the SMPS GND pins as recommended for most modules? What type of SMPS is it you're using? \$\endgroup\$
    – Puffafish
    Commented Apr 8, 2019 at 13:11
  • 2
    \$\begingroup\$ What type of SMPS? Mains in and low voltage DC out? Add a link to a datasheet. It is possible that the SMPS has too much (capacitive) coupling to the mains which then disturbs the capacitive touch. Filtering will not fix that. A "better" SMPS or a 2nd isolated DCDC converter or a "better" touch screen controller might fix it. \$\endgroup\$ Commented Apr 8, 2019 at 13:17
  • \$\begingroup\$ If SMPS body is not earthed, LCD touch signal affecting. I checked with battery supply, LCD function working fine with battery mode. That's why am doubting on SMPS. Unable to find capacitor recommendation for GND pins in datasheet or any design notes. Can you tell me between where we need to connect capacitor ? \$\endgroup\$
    – Boopathi
    Commented Apr 8, 2019 at 13:32
  • 1
    \$\begingroup\$ Per page 2 or 3 of the SPEC, there is 300uA of leakage current. We may interpret that as 3,333 ohms of leakage per volt, or about 390,000 ohm of impedance between primary and the output pins (VDD and GND/RTN pins). Use a DVM, set on AC, to measure the level of trash on your touch-screen RTN. This will be the potential between the screen and the fingers. \$\endgroup\$ Commented Apr 8, 2019 at 14:20
  • \$\begingroup\$ @Bimpelrekkie How about using LDO after the SMPS would PSRR eliminate such noise? Would isolated DC DC converter after an SMPS reduce 50Hz related noise? \$\endgroup\$
    – user16307
    Commented Apr 19, 2019 at 18:37

2 Answers 2

1
\$\begingroup\$

It is unlikely that the SMPS is what is causing your issues. This is even what you're observing: the touch screen works fine if the SMPS is powering it as long as there is a protective earth ground connection.

Therefore, the SMPS is not the problem. If it was, the issue would be there regardless of any earth ground connection.

The touch screen and everything else in the device is still grounded to the SMPS, any noise that is due to the SMPS is still there with or without a protective earth ground. So attempts to filter out noise that is not actually causing a problem in the first place is not going to be particularly helpful.

The real issue is that capacitive touch screens lose a lot of sensitivity when they are isolated from earth ground, as this creates a parasitic capacitance between the panel and earth ground/everything else. In effect, this acts like a capacitance that is always in parallel with the screen.

As far as the touch screen is concerned, humans and their fingers are little more than meat-based capacitors, and this capacitance acts as a secondary path for a tiny AC signal (~200kHz typically), and is diverted through the meat capacitor when the screen is touched.

If there is a small capacitance in parallel with this meat capacitance, this results in there being less change in the signal when meat capacitance (a touch) is added. The capacitances involved are both in the range of single digit picofarads, so the presence of that extra screen-to-earth ground capacitance can have a large impact on the change in the signal seen when a user touches the screen. Thus, sensitivity is reduced.

There are ways to help mitigate this. If the panel has a metal frame, bonding that to your device's ground via a screw and wire lug can increase sensitivity a little. However, this is typically something that you would fix using the touch screen controller itself. Most controllers have registers that allow you to adjust sensitivity thresholds for exactly these situations, or have a 'training mode' where one holds a finger on the touch screen for some number of seconds while it calibrates the sensitivity.

I can't really provide more information than that as you have not provided any information about the touch screen or the controller. I would suggest looking up the part number for the panel and/or the touch screen controller and finding the data sheet or other documentation. It should have the information you need.

\$\endgroup\$
0
\$\begingroup\$

Do you NEED to use a medical grade supply there?

You can get class II supplies that are not medical (low leakage) designs. The advantage is that they have a class Y cap across the primary/secondary isolation barrier to reduce high frequency common mode nose at the cost of higher line frequency (and harmonics) leakage.

Unless you need it, medical grade is NOT better here.

\$\endgroup\$
1
  • \$\begingroup\$ Yes, I want to use medical grate SMPS. i checked with common mode choke and Ferrite core(6 turns) at SMPS output side still the issue is reproducing. I connected Copper plate at SMPS earth point, but there is no improvement. \$\endgroup\$
    – Boopathi
    Commented Apr 10, 2019 at 13:45

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.