I have a christmas light display with a 30W RGB LED run by an official Uno R3 that I set up outside. I have tuned the sketch indoors, and when I take it outside it appears to run at a slower speed, about 25% to 1/3 slower. Is this expected?
Further info - the temperature is about 75 degF indoors and 30 degF outside. The Uno is providing 4 PWM outputs to control 3 NPN mosfets and a power op-amp, and taking 2 analog input readings. The other components are a 50 kohm resistor and 10uF electrolytic capacitor that form a LPF on the PWM input to the OP amp (as a non-inverting amplifier). Some connections are on a soldered breadboard and others are on a solderless breadboard with jumper wires. The LPF frequency is ~.3Hz to filter the PWM really well. I am surprised but it really seems that the Arduino is slower but I don't understand if this is expected. EDIT: Forgot to mention the battery is a large 12V 7Ah LI-ion battery back and there's a boost converter operating at 34V for the LED. Thanks.
If you're using software on the arduino it's timing stability will depend on it's reference - a crystal or ceramic resonator. Neither should change significantly with a 40F temperature range, so long as it's within specs.....
crsytals would be maybe 100ppm, ceramic resonators worse, but not terrible......
Could condensation around freezing be upsetting things?
perigalacticon:
I have a christmas light display with a 30W RGB LED run by an official Uno R3 that I set up outside. I have tuned the sketch indoors, and when I take it outside it appears to run at a slower speed, about 25% to 1/3 slower. Is this expected?
Further info - the temperature is about 75 degF indoors and 30 degF outside. The Uno is providing 4 PWM outputs to control 3 NPN mosfets and a power op-amp, and taking 2 analog input readings. The other components are a 50 kohm resistor and 10uF electrolytic capacitor that form a LPF on the PWM input to the OP amp (as a non-inverting amplifier). Some connections are on a soldered breadboard and others are on a solderless breadboard with jumper wires. The LPF frequency is ~.3Hz to filter the PWM really well. I am surprised but it really seems that the Arduino is slower but I don't understand if this is expected. EDIT: Forgot to mention the battery is a large 12V 7Ah LI-ion battery back and there's a boost converter operating at 34V for the LED. Thanks.
No electronics likes 100% relative humidity - I think you have this issue. Circuitry outside should either be
hermetically sealed from the atmosphere or in a box generating enough heat to keep the relative humidity down
within the enclosure.
I'm saying a dc incandescent light, more like a small automotive type. A 60W bulb in a small enclosed space might push the temp problems too far the other direction
The electronics are in a acrylic box that is 12" cubed, the light from the 30W LED illuminates the box from the inside, it is near the electronics in the box. The LED is attached to a heatsink that has a cooling fan, which simply moves the air around in the box but it stays cool. Are you saying you can paint the electronics with a coating to waterproof them?
There are various types of "conformal coating" for electronic applications that can make them more moisture resistant.
Given that your circuit is apparently inside a box with a fairly powerful heating element (the high powered LED) and a fan, I'm skeptical that moisture or modestly low temperature is directly a problem unless there's obvious condensation in the box.
The solderless breadboard probably isn't a great idea as they can do sketchy things even in a benign environment.
Without a schematic, I'm not clear on what you're doing with the analog inputs and the filtered PWM output.
Ok I figured it out! It was the timeout for pulseIn on the ultrasonic sensor readings. It was default to 1 second, and when I took it outside it was timing out on every reading due to nothing to reflect from. It works fine now. Thanks for your help! Happy New Year!
P.S.
BTW I once had a circuit on a solderless breadboard I set up on a window sill with a light sensor to trigger an automatic light, and it never worked right, I troubleshooted it for days before I found condensation in the breadboard at night! Worked great after fixing that.
perigalacticon:
I actually have the system in a acrylic box that I made and sealed with silicone weather-stripping, although it's not a hermetic seal. Thanks.
Is the inside significantly warmer than the outside in use? If not you will have moisture problems for sure.
Another strategy is good thermal insulation around the box so its temperature doesn't vary much over the
day/night cycle, but that's usually a lot of insulation or needs a big thermal mass to help stabilize temperature.
People often pot a circuit that's going to be deployed outside, although obviously that's an irreversible process.
