I have a problem that I can't solve with a Mega2560-based set equipped with a TFT_320_QVT_9341 screen connected via a Geekteches v2.0 shield.
All the programs uploaded to the Mega work perfectly, including the display. Only the touch panel doesn't work. It reacts to touch, but the positions are not decoded.
After many tests, I replaced the power supply provided by the USB port (which is correct, I should point out), with an external power supply via the coaxial connector.
The touch screen starts working perfectly when the external supply voltage is between 5.0 v and 5.5 v !
The recommended voltage on this connector is 7 to 12 v.
I did all my development in these extreme conditions, to be able to move forward, only plugging in the USB to upload, which instantly breaks the touch pad.
Could someone experienced or with a good inspiration help me?
Thank you in advance.
Sincerely.
There is a current thread about TFT_320QVT_9341 with a different Adapter here
Please can you confirm that UTFT v2.8.3 works with your TFT_320QVT_9341 screen and Geekteches v2.0 40-pin Adapter shield.
The TFT controller will only use about 20mA
The TFT backlight will use about 50mA-100mA
The Touch chip will use about 5mA
The Mega2560 will use about 20mA
So USB should be fine for powering the whole sandwich.
If you have high currents or unusual voltages check any jumpers or solder-bridges on screen or Adapter shield.
david_prentice:
Please can you confirm that UTFT v2.8.3 works with your TFT_320QVT_9341 screen and Geekteches v2.0 40-pin Adapter shield.
Yes, that's exactly what it is.
The TFT controller will only use about 20mA
The TFT backlight will use about 50mA-100mA
The Touch chip will use about 5mA
The Mega2560 will use about 20mA
So USB should be fine for powering the whole sandwich.
My set consumes 148 mA on the USB port (full backlight).
The port voltage drops from 5.13 v idle to 5.08 when loaded. I have the same problem with the USB ports on another machine.
I repeat : powered by USB it never works (the touch pad only).
Directly powered, it works perfectly if 5.0 < U < 5.5 volts !!!
If you have high currents or unusual voltages check any jumpers or solder-bridges on screen or Adapter shield.
bad luck, no jumpers or solder-bridges to modify !
Hello David.
Unfortunately, I don't have the shield schematics either.
The black one has only tracks, no active electronic components.
All CMS resistors are 10 KOhms, without exception (36 resistors in total).
All three, near the multiturn potentiometer, are also resistors. The last one is located near the 3.3 Volts output pin.
The potentiometer is used to adjust the intensity of the backlight.
On the white shield the buffers are HC541 (Octal Buffers and Line Drivers 3-State Outputs).
The problem with the touch panel is really weird ! From time to time (perhaps once in 50), it works correctly when powered via USB or under normal voltage, but at the first reset, the failure is back.
I will save this shield for a possible application where I only need the screen but not the touch panel...
External hardware tends to use HC family chips instead of LVC chips.
The 74HC541; 74HCT541 is an octal non-inverting buffer/line driver with 3-state outputs.
The device features two output enables (OE1 and OE2). A HIGH on OEn causes the
outputs to assume a high-impedance OFF-state. Inputs include clamp diodes that enable
the use of current limiting resistors to interface inputs to voltages in excess of VCC.
The 74LVC541A is an octal non-inverting buffer/line driver with 5 V tolerant inputs and
outputs. The 3-state outputs are controlled by the output enable inputs OE1 and OE2.
Using 5V logic with HC turns on the clamp diodes. Without series resistors the 3.3V VCC supply gets pulled above 3.3V.
So all normal SMD resistors on the Black Adapter are 10k i.e. 103
What value are the small SMD next to the Blue pot?
I would expect 47R or 22R for the pot. The backlight LED might need 50mA to 100mA.
Of course an external constant current chip would use high value resistors to "set" the required current.
If it is really wired as a 10k potentiometer. One turn would be 1k. 1/10 of one turn would be 100R.
So you would adjust the backlight brightness between 1/20 and 1/10 of one turn. Just about practical but you will be dissipating about 100mW on a tiny length of track.
I don't see any 3.3V regulator chip. If the Blue pot is fed from the Arduino 3.3V pin, the backlight will exceed the maximum 50mA from a MEGA2560's 3.3V pin
If the pot is fed from 5V pin, the dissipation will be higher but you might get up to 1/5 of one turn on the ten-turn pot.
Surely other readers have this Black Adapter with Blue pot.
I disconnected the pin that powers the backlight LEDs to power it from an external source :
same problem !
I would have liked to understand, but I now have a set that works and I'm going to move on !
Thank you so much for wanting to help me with this problem.
I am pleased that you have got one Adapter working 100%.
I am intrigued by your Black Adapter. The 10k Blue multiturn pot just does not make any sense.
Chinese designers know exactly what they are doing (or I would hope so)
Chinese shopkeepers do not know what they sell.
I can understand a misplaced copper trace. We are all human.
I can't understand a 10k pot instead of a 100R variable resistance.
I can't understand HC logic when LVC voltage tolerant chips should be used.
Hey-ho. I don't own these Adapters.
Hey-ho. I support a library originally written for Mcufriend boards. Which have always had design "warnings" and nowadays downright "ERRORS"
