Hi. i'm designing some kind of touch key with lora antenna and ttp223 touch ics. the problem is that when i start to transmit or touching one of the touchpads , other ttp223 ics start to false detect. i read the lora ra01datasheet and i found out that the lora ra01 consume about 120mA of current when transmitting. is it my false detection because of this 120mA? and when i measure the voltage on lora ra01 with my multimeter, the ics start to act normal and doesnt detect false touch. maybe there is a huge capacitor in my multimeter?my design is attached, thanks for your time. Lora antenna is on bottom layer and touchpad and touch ic is on top layer. i tried to add 470uF and 680uF on vcc pin of lora ra01 but still same result. i tried to remove the vcc trace on pcb below lora ra01 but only first touch ic act normal.(the one is far from antenna) ismy false tiuch sensing because of inrush power supply or antenna TX power cause this problem? and when the antenna is receiving everything work fine. the problem is when the antenna start to send data.

 

What is the MCU? Are there pull ups on the MCU’s inputs?

 

The MCU is Atmega32 and yes there are pullup inputs on MCU, but ithink the problem is with touch ics(TTP223) the detect touch while i'm not touching anything. maybe the antenna Tx power cause this problem?

 

You may need to put low pass filtering on the inputs to the touch ICs. Do you have a ground plane between the antenna and the other side of the board?

If the touch pads are acting as antennas, filtering might help. I don’t think it is current related. I think it is RF related. Try turning the LoRa power way down and see it is changes. Of course that will change both current draw and RF out but at least it isolates the problem if it stops happening.

Have you put a scope on the input to the touch IC? If the problem is RF, you should see a strong signal at LoRa frequencies. Remember, though, LoRa is spread spectrum to is might look strange.

 

You may need to put low pass filtering on the inputs to the touch ICs. Do you have a ground plane between the antenna and the other side of the board?

If the touch pads are acting as antennas, filtering might help. I don’t think it is current related. I think it is RF related. Try turning the LoRa power way down and see it is changes. Of course that will change both current draw and RF out but at least it isolates the problem if it stops happening.

Have you put a scope on the input to the touch IC? If the problem is RF, you should see a strong signal at LoRa frequencies. Remember, though, LoRa is spread spectrum to is might look strange.

My Tx power on LoRa was dBm, i reduced it to 4 dBm the pcb now is very much stable, but still the pad closer to LoRa ic detect touching sometimes. what can we conclude from this behavior?

 

My Tx power on LoRa was dBm, i reduced it to 4 dBm the pcb now is very much stable, but still the pad closer to LoRa ic detect touching sometimes. what can we conclude from this behavior?

How much were the current spikes reduced? Try putting a really large capacitor (~1000mF) across the power supply rails and see if that changes anything. It doesn't seem to be current related, though.

Do you have a ferrite of the RFI elimination type? If you could cut the trace going to the flaky IC's input and rewire it temporarily with a bodge wire through a ferrite (looped a few time) that might tell you if using RFI suppression is an answer.

 

Do you have a ferrite of the RFI elimination type? If you could cut the trace going to the flaky IC's input and rewire it temporarily with a bodge wire through a ferrite (looped a few time) that might tell you if using RFI suppression is an answer.

No, i have no idea what is that. can you explain more of what is this ferrite is?(sorry but i'm newbie . )

 

Ferrite beads and rings are used to suppress RFI. Usually on power connections but in some cases to keep RF out of one section from another. You can get then for THD and SMD parts.

It is probably the case that you should redesign your PCB but an RF choke (in the form of a ferrite bead) might be a workaround.