[werk_ag]

What current draw is expected at 9V for the Pro2+ RX, with backlight on?

The current consumption is variable, just the ESP8266 module when transmitting takes near 200mA
A switched 9V 1A to 1.5A power supply is enough.

With the TFT backlight OFF the 3.3V regulator should run cold to touch. Where on the board have you connected the TFT VCC pin? To 5V or to 3.3V?

[clubjoker]

The TFT Vcc pin is connected to the PCB 5V pin, top right of the group of (8x2) 16 holes in the PCB. I've just double checked, and there is continuity between TFT Vcc and 5V on the RX RF module.

I will report back when I've tried different power supplies, hopefully I'll have a chance to do this tomorrow.

Thanks again.

John

[clubjoker]

Another test:

Conditions:
I am running the RX+ off the same 7809 linear regulator, but now with a 12V lead acid battery instead of the cheap 12V switched mode PSU I had been using up till now for both TX and RX+. On the 7809, I've got a 0.1uF ceramic capacitor on the input side (12V lead acid battery), and 22uF electrolytic on the output regulated 9V side, each on the breadboard, about 9cm of wire from the TO220 package of the 7809 - perhaps not close enough?. The 7809 has a small heatsink, so should cope with 1A draw.
Timeout to big clock display set in software config to 1 minute.

Observations:
- 3.3V reg on the RX+ gets hot as the system starts up, and hot with the main weather summary screen showing all the sensor readings at once.
- When the big clock screen kicks in, the 3.3V reg cools down so it's just a bit warm to the touch even though the TFT is on (this was a surprise to me!)
- Press the push button PB1, back to the summary screen, and the 3.3V regs gets HOT

Mysterious...

Buck step-down regulator should arrive tomorrow, to test as 9V supply for RX+ instead of the 7809 breadboard hack that I've been using up till now. Hope I'll have time to try it

[werk_ag]

- When the big clock screen kicks in, the 3.3V reg cools down so it's just a bit warm to the touch even though the TFT is on (this was a surprise to me!)
- Press the push button PB1, back to the summary screen, and the 3.3V regs gets HOT

If you did that experience after 8PM, the explanation is simple: Between 8PM and 8AM (or based on the Solar Radiation value, if available), the backlight level is automatically reduced when the Big Clock is shown, so less current is draw from the 3.3V regulator.

What about the voltage regulator on the Meduino board. Is it hot too? It shouldn't.

[clubjoker]

- When the big clock screen kicks in, the 3.3V reg cools down so it's just a bit warm to the touch even though the TFT is on (this was a surprise to me!)
- Press the push button PB1, back to the summary screen, and the 3.3V regs gets HOT

If you did that experience after 8PM, the explanation is simple: Between 8PM and 8AM (or based on the Solar Radiation value, if available), the backlight level is automatically reduced when the Big Clock is shown, so less current is draw from the 3.3V regulator.

That makes sense. Yes, it was after 8pm in the evening, and I can see that the backlight is dimmer with the big clock display than when the weather screens are displayed. Strange how it seems to make such a big difference to the regulator temp when the TFT backlight is dimmed slightly - the temp of the regulator seems to be sensitive to the extra power draw.

What about the voltage regulator on the Meduino board. Is it hot too? It shouldn't.

I've just checked and the regulator on the Meduino board is hot too. That one is hot when the TFT backlight is on and also when the TFT light is off.


I've also today received the step-down buck converter, set it to 9.0V and tested with that supplying the Pro2+ RX. Same issues with the regulators running hot:

Conditions:
buck converter to 9V

Observed:
TFT light on: RX+ 3.3V reg hot, Meduino reg hot
TFT light off: RX+ 3.3V reg a bit warm, Meduino reg hot

Now this is strange because otherwise everything seems to be functioning fine, except for the display corruption, which I haven't seen yet today.

Any ideas on what to check next?

[werk_ag]

I don't know what, but something is draining more power than it should. The suspect now seems to be the TFT display. For a while try to switch off the TFT backlight by pressing the push button, and check if both regulators start to cold down.
Normally, even keeping the TFT backlight always ON, both regulators should run almost cold.

[clubjoker]

Hmm.

The regulator on the RX+ board is only hot when the TFT light is on. The regulator on the Meduino is always hot (whether the TFT is on on not). What does that tell us?

[clubjoker]

OK, so I've come back to this to try to solve the problem, after a bit of a break. No luck yet, but I've got some more information.

I cut the ribbon cable to the TFT shorter, so it's now about 15cm long. Result is that there is no more corruption / garbling of the display, so that's fixed. Except for the hot regulator issue below, the system seems to be working perfectly.

However, the hot linear voltage regulator problem remains. I've tried with a 20x4 LCD as an alternative to the TFT to get more information to try to diagnose the problem.

Results:
TFT backlight on: Pro2+ reg hot, Meduino reg hot
TFT backlight off: Pro2+ reg slightly warm, Meduino reg hot
TFT cable disconnected: Pro2+ reg slightly warm, Meduino reg slightly warm
20x4 LCD backlight on: Pro2+ reg slightly warm, Meduino reg hot
20x4 LCD backlight off: Pro2+ reg slightly warm, Meduino reg slightly warm

In all cases when the backlight is switched from on to off, the regulator cools down rapidly from hot to warm, within about 5 seconds.

I thought the heat issue could be related to the RTC battery charging / drawing too much current due to being faulty. So I unplugged the RTC board from the main Pro2+ board, but still both regs hot when the TFT is connected with backlight on.

Reading around on the web, hot to touch probably means it's in the range of 60-70 degrees celsius. I suppose that putting my finger on the regulator for a few seconds might contribute to the higher temperature by stopping convection.

Do the above results tell us anything? Are you 100% sure that the regulators on your boards are not hot to the touch when the backlights on the TFT and LCD are on?

[werk_ag]

I've also today received the step-down buck converter, set it to 9.0V and tested with that supplying the Pro2+ RX. Same issues with the regulators running hot:

I really don't know why your voltage regulators are running so hot, but as you are using a step-down buck converter you can try lower the output voltage to around 8V and see if it makes any diference. Doing it will not affect the functionality of the system.

[clubjoker]

I've also today received the step-down buck converter, set it to 9.0V and tested with that supplying the Pro2+ RX. Same issues with the regulators running hot:

I really don't know why your voltage regulators are running so hot, but as you are using a step-down buck converter you can try lower the output voltage to around 8V and see if it makes any diference. Doing it will not affect the functionality of the system.

I will try that, reducing the voltage to 8V.

But I would like to find and fix the root cause of the apparently excessive power draw. I wonder if there is some static electricity damage.

From the diagnostics tests I ran previously, and the fact that the TFT display is working fine, can we be sure that the 3.3V<->5V buffer level-shifter 74AHC244 chip is fine / not damaged? I did remove the power diode on the Meduino before powering the system up, but wonder if chips / transistors might have been damaged by static electricity or (less likely?) by overheating when I soldered.

Is it worth me trying by replacing the BMP280 sensor with a BMP180 sensor I have spare? (Pressure readings seem to work OK, but it would be good to eliminate that as the cause of the problem).

Intending to minimise the risk of future failure, I have soldered all the components except the RTC to the PCB, so it's a bit of a pain to swap things out to test them. Maybe I would be best to try to build another PCB up, but $$$