Has anyone seen or done a project to sense when a forced air furnace filter needs to be changed? I was thinking about putting a differential pressure across the filter - by using 3.5mm bulkheads.
But I don’t know if measuring the differential pressure across a filter is a good way of determining if the filter is dirty? It seems like it would be, but the details of exactly where to place the ports to avoid turbulence impacting measurement, fan speed potentially changing, etc can really side line your readings.
I’m also wondering what kind of hardware I can use to install the port connector on the ducts. Some kind of 3.5mm tubing bulkhead connector…but I haven’t found anything that seems easy to install.
Just wondering if someone else had gone down this road and can provide some advice.
This is a good idea. I think the sensor read location would be experimental. I have done some testing using diff pressure sensors in aircraft installations mainly concerning pressure drop validation. We had good laminar airflow through our heatsink so location was not an issue. If there is turbulent air at the sensor the reading will be choppy / all over the place. With a filter I would guess it’s mostly laminar. I think I have a sensor at the shop and may try to hook it up to an ESP32 and see what happens.
I’m wondering with pressure sensors, do you need the tubing to be “facing” the air flow? Or can you just put a tubing into the aluminum ducting perpendicular to the flow? It seems like the way the tubing rests inside the duct would impact readings a lot.
I think you might get a Venturi effect at right angles. I would face it downstream. i.e. aligned with the air flow but not facing into it. I’m no fluid mechanics engineer though.
Since this was the most pressing issue in my mind for some crazy reason, I reached out to a former colleague. After what felt like a lifetime of physics lecturing, he finally answered and say’s away from the airflow. There was a caveat in there but I had already walked away as I couldn’t deal with another hour of theoretical lecture.
I’m also interested in this; have the electronics all working, just need to know how to mount the sensors.
Did anybody do it?
BTW, I think Venturi would only apply if there was a change in duct diameter, but it also seems to me like similar things would happen; hence the question/hesitation.
Also, how should I connect some tube to this:
This might help:
Filtrete has some filters with a BLE sensor that emits data about filter lifespan and has a barometric pressure sensor. They cost about $3-$5 more than the non BT ones I’ve seen at stores that carry them.
Someone did a teardown, they’re attached via hot glue (looks to be easily removable and placeable on a new filter if you have a glue gun), and have replaceable coin cell batteries.The person was also nice enough to document some of the BLE output.
Since you could buy one filter and migrate the sensor over repeatedly with a little work, that might be the move if there were an integration.
That certainly sounds like a nice feature/integration to have.
I currently have no BT/BLE devices connected to HA because I worry about the limited distance most BT/BLE devices can communicate over. Do I worry too much about this?
I have some BeeWi lights I wouldn’t mind connecting if distance wasn’t a problem.
Long ago I bought a “Class 1” BT dongle but I think it was fake; never got “too far” with it. Now I never see the class mentioned anymore and while BT or BLE 5 is supposed to allow long distance connections, it seems that’s true only for some devices.
Anyway, The original question still remains, as an ESPHome device could monitor more things in the heating room and even control stuff. Plus it already exists.
ESP32s can be set up as a bluetooth proxy, effectively ending range concerns if you have devices that aren’t in-range of your main HA host’s BT dongle.
The sensor support should be merged into esphome soon.
I’ve just installed it on my furnace. I’m happy with the results.
The drawbacks,
The values differ from my professional grade 100$ pressure meter
It needs initial calibration for the offset value
However, the price of the sensor itself (9USD) is a huge savings. Name-brand differential sensors are 40$+ just for the sensor.
My hopes is that while the absolute accuracy might be in question, for the goal of monitoring a furnace filter over time to determine if it is getting dirty, it should be adequate.
I’ve asked the manufacturer of the sensor what sort of minimum order quantity would be required to get a better suited version.
This is great to know; thank you!
I suppose I don’t even need a dongle if I do that. I wonder what happens if I add the BT proxy to ALL my ESP32 sensors…
Considering that all you need is differential, and absolute pressure is not needed or particularly useful, it seems like the cheapest/easiest implementation is to grab a MPXV7002DP breakout and couple it to something like an ESP32.
This sensor has a range of -7 to +7kPa, 15" w.c. is equivalent to roughly 3.7kPa, so well within it’s range.
A brand new filter should have a pressure drop of between 2 to 3″ w.c. (water column). Once the pressure drop exceeds 10 to 15″ w.c. it needs replacement (standard industry values). Seems like it would be pretty easy to configure an alert to trigger when the pressure reaches the threshold. The device should live on the outside of the duct with one port exposed to the ambient room air (probably put a cloth/fiber dust filter over it just so it doesn’t accumulate crap inside the filter), and the 2nd port should be attached to a hose that terminates inside the ductwork between the filter and HVAC unit.
I think a simple design where it has a housing on the outside with the controller and sensor, then a ‘spike’ sort of component that extends downward to penetrate the ductwork should do the job. Then the whole unit could be attached to the side of the duct as one piece and be able to read both pressure environments, and the body of the device would seal the penetration (make a sealing flange ring part of the housing, then tape it to the duct.
The device can likely be powered by the low voltage control power supply that’s in the main HVAC unit.
