Bug Filed on 03-30-2016: GPIO DO not toggling

Yes I will check tonight. I have another Pi Model B that I can swap out and compare apples to apples.

What is the typical latency for activating outputs when on the wifi network? The WAF (wife acceptance factor) needs to be very high or the project will be relegated to the junk drawer. :joy: I can picture my wife repeatedly hitting the button if the output isnā€™t near instantaneous!

The output of the LED should be instantaneous. However, the widget will likely spin for 2-6 seconds and then youā€™d be able to turn the LED back off. The reason the widget spins is itā€™s waiting for a confirmation that the LED indeed did turn on before it reflects a change in the widget state. Weā€™re improving this method so that there is not such a long time between toggling a single widget.

Note that the app (Android or iOS) functions differently and will be much faster response rate of the widget.

-B

Iā€™m interested in hearing more about how this is handled. Iā€™ve been involved in DDC building controls and weā€™re always very interested in confirming that an action did in fact take place. For example, weā€™ll install a CT donut on the power leads of a fan to confirm that when the output is triggered, weā€™re seeing confirmation that the fan did in fact turn on.

When I used WebIOPi in the past with I2C expanders, I would actually take the output of an expander and connect it to a separate input to compare. If they were not equal, it would generate an alarm. I was very surprised at how quickly the IO responded using the I2C expanders. Near instantaneous!

Iā€™ll get some more info and get back to you :slight_smile:

Well switching out the Pi did the trick. I moved the SD card into another Pi (along with the previous Pi 2) and can confirm it is working. Something is wrong with my hardware it seems. The Android App is very responsive and the LED toggles immediately in real time. Feedback via the widget in the app is very quick as well. Web interface lags by a few seconds as you indicated. Sorry for the trouble!

BTW, my new Galaxy S6 had the battery drained to 15% (from full charge in 5 hours) with the Cayenne app accounting for nearly all of the battery usage. It was running in the background and my phone was sitting idle on my desk for most of the morning not being used. Is it not shutting down in the background? Is this worthy of a bug report? Iā€™ll probably just Greenify the app to be safeā€¦

Good news!

Bad news!
Not being super familiar with the older RPiā€™s, I do know the pin out has evolved over the years, aside from the 26 to 40 pins, some of the assignments have been fiddled with. Iā€™m guessing youā€™re aware and were doing things correctly. Always sad to hear of broken hardware.

Let us know if you find out what went wrong.

Hopefully your travels will be a little less bumpy,

Ian

p.s.
I like the new avatar, doesnā€™t look like a nut though :grin:

In a nod to backwards compatibility, the 26 pins on all Pis are the same. The additional 14 new pins on the Pi 2 and 3 (2 rows of 7) are adjacent to the 2 rows of 13 that existed on the original. I believe the intent was to keep hardware somewhat backwards compatible.

The software is really slick so Iā€™m excited to come up with some projects. This Pi will be retired now. It gets really hot when itā€™s being used.

So that could be clearer. :astonished:

I was thinking of the difference between the Rev 1 and Rev 2 boards, as outlined here-
http://raspi.tv/2013/rpi-gpio-basics-2-how-to-check-what-pi-board-revision-you-have

Iā€™m new enough to the game that all I have are 2bā€™s, 3bā€™s, and Zeroā€™s. This stuff is great fun, and Cayenne keeps me from having to futz with too much code, thank goodness.

Looking forward to a full report on what gets built around your RPiā€™s, and donā€™t forget the contest, you could win a trip to Mustang country!

Ian