Keep getting "st=fail" despite both 3.3V regulator and decoupling capacitors

[nanotube]

Folks, I have benefited a lot from reading this forum but have to come up here to ask for one more help. This problem has baffled me for nearly 2 months:

Ethernet Gateway on VeraLite is made of a Uno with ethernet shield.

Problem is with a DHT22 sensor I am trying to make. The sensor radio is the correct NRF24L01+. Using a USB Arduino Nano. DHT22 code is working fine. But the radio keeps giving "st=fail" error, see below.

ensor started, id 1
send: 1-1-0-0 s=255,c=0,t=17,pt=0,l=3,st=fail:1.4
send: 1-1-0-0 s=255,c=3,t=6,pt=1,l=1,st=fail:0
send: 1-1-0-0 s=255,c=3,t=11,pt=0,l=8,st=fail:Humidity
send: 1-1-0-0 s=255,c=3,t=12,pt=0,l=3,st=fail:1.0
send: 1-1-0-0 s=0,c=0,t=7,pt=0,l=3,st=fail:1.4
send: 1-1-0-0 s=1,c=0,t=6,pt=0,l=3,st=fail:1.4
send: 1-1-0-0 s=1,c=1,t=0,pt=7,l=5,st=fail:74.3
send: 1-1-255-255 s=255,c=3,t=7,pt=0,l=0,st=fail:
T: 74.30
send: 1-1-0-0 s=0,c=1,t=1,pt=7,l=5,st=fail:55.8
H: 55.80
send: 1-1-0-0 s=0,c=1,t=1,pt=7,l=5,st=fail:55.5

Now every once in a couple weeks I will get "lucky" and get a few minutes of "st=ok" and connect to Vera successfully. But after a few minutes the data stream would just stop and when I checked debug on the sensor Arduino serial output it was "st=fail" again. I have put the radio power on the output of a 5 V to 3.3V regulator; have put 4.7uF and even 10uF capacitor across the radio power pins but kept seen the same results. Why didn't it work?

[BulldogLowell]

@nanotube

So you have decoupling capacitors, what type of power supply are you using?

Are you sure your problem is the node and not the gateway? Ethernet shield uses a lot of Power. try looking at your serial output from the gateway

[nanotube]

Honestly I'm not sure whether gateway problems would also lead to the sensor radio "st=fail" -- is that one of the reasons? Thanks.

For the sensor I have tried both 5V brick power supply to USB and just plugging in to the USB on my desktop computer;
for the gateway it's mostly on a 6V brick power supply, plugged directly to the power port on the Uno.

I will try debug mode on the gateway and see if there is a message.

[m26872]

Try a "delay(100)" before every send message.

[hek]

@nanotube said:

Honestly I'm not sure whether gateway problems would also lead to the sensor radio "st=fail" -- is that one of the reasons? Thanks.

It's the gateway that acks(replies) your sensor transmissions. This is the problem you're seeing. You must ensure gateway have enough power and is correctly decoupled.

[RJ_Make]

Just a note, I have to use 47uf capacitors to get my radios working with V1.4.

[nanotube]

Happy new year folks. Long overdue update but still no cigar...

Thanks to everybody above's input I now understand the fail problem comes from my gateway. I have tried the following steps:

Gateway (Arduino Uno) powered via USB by a high quality charger (iphone charger);
5V from Uno goes to an external down converter to get stable 3.3V;
Radio powered by 3.3V in parallel with 40 uF capacitor.

Sensor radio is also powered in parallel with 20 uF caps. I measured after all these steps I get 3.300 +/- 2 mV to the radio. Well problem solved or so I thought.

Only after these steps would I be able to get stable communication (verified by serial debug from the sensor), but it only lasts two hours max, after which I would get st=fail again until I unplug power to the gateway, replug it in and reload Vera.

Resetting gateway didn't help; simply reloading Vera didn't help. It had to be a hard reset by unplugging it. I am starting to suspect the radio I bought off eBay is from a bad batch (I bought a 5 pack but switching between them didn't help - they all communicate for a few minutes to 2 hours max then drop dead). Any ideas ? Thanks.

[Dan S.]

@nanotube My ethernet gateway has a separate 5 volt power supply via the down converter to power the radio and another power supply for the uno/ethernet boards. The ethernet board Uno combination consumes quite a bit of power. Also I don't know if I would call the iPhone charger high quality. Bottom line is I think you still have a power supply issue using one charger for powering everything. Certainly it is good to have the separate down converter rather than tapping off the UNO but you are placing a lot of demand on a single power supply. I think it is best to power the radio separately (but make sure all grounds are common).

[griffinsaic]

While not involving an Uno, I thought I would share my experience.

I was having a similar issue with my nano serial gateway; not accepting messages or incomplete messages from sensors and/or nothing was working at all. Instead of adding an external power supply, I used a Nano Expansion Sensor shield. This cleaned up the power issues, and makes it easier for me to attach several components.

http://www.ebay.com/itm/Nano-O-Expansion-sensor-Shield-Arduino-UNO-R1-Nano-3-0-duemilanove-2009-/310546482805?pt=LH_DefaultDomain_0&hash=item484e034675

While not ideal for small for factor, it works. I have not had any further communication issues with the serial gateway. However, I do have issues when the sensor battery voltage drops below a certain level. While I have a multimeter, I have not checked at which voltage. Instead I change the batteries whenever the LEDs get dim.

From my limited experience, power is most likely your issue.

[daulagari]

I agree with @griffinsaic, power is most likely your issue but another thing to check is interference, see the Issues with multiple gateways + choosing frequency / channel thread.