Sensor systems Oil watchman, technical details wanted

I started with one of these: 

Now it's in production making pretty graphs of how much oil we have, though, so I got another one for playing with:

I don't believe you have to be particularly careful about which one you buy.

They do have different antenna sockets though. The one from Amazon is MCX, and and I've just ordered a 433MHz (17cm) antenna to help improve the signal when I migrate the receiver from my office to the cupboard under the stairs where the home automation system is actually running:

Thank you. Is the error consistent, over say a 5cm drop in oil level? i.e. if the level drops by 1mm does the gauge reflect this?

I haven't actually calibrated it very well. I spent quite a while on temperature calibration ? a discussed at the hackaday link I posted . But the depth reading was so obviously "about centimetres" that I haven't spent much time on it. Just before my tank is filled, I'll measure the reading and compare with the actual height of air. I'll then repeat that when it's full and use those points for calibration.

FWIW my initial temperature calibration isn't great either; I'm going to stick a 1wire temperature sensor out there next to it and monitor them both over time.

But it's doing the job well enough.

In answer to your question: no, when the level drops by 1mm, that usually (9 times out of 10) results in no difference in the reading from the monitor since that is only centimetre precision.

If you briefly stick a magnet by the sensor it starts sending three readings a second, instead of its usual one per 35 minutes. And when the reading is part-way between one centimetre and the next, you often see it fluctuate ? I have logs with it sometimes reading 240, somet imes reading 241. There's no pattern to it; just a slight inaccuracy in the reading.

That was happening even when it was just pointed at the wall in a room at a stable temperature. With it outside and the speed of sound also varying, it's unsurprising that the reading does fluctuate a little.

And of course the actual volume of oil is changing too. In my scripting I've attempted to compensate for that and report the volume at 10°C rather than at the current temperature. I'm not yet sure if that's adding noise or reducing it :)

FWIW here's the last few days' readings from it:

So you can see that although there's about 1cm precision, the accuracy is a little less than that. But that doesn't mean it's not fit for (my) purpose. Once converted to a "percentage full", it gives data which are perfectly sufficient: 

At an average reading of 48.7% on December 13th, and an average of

42.5% today, I've used about 6% of the (1300l) tank in 8 days and at this rate I probably have about 7 or 8 weeks worth of oil left. Not that it *will* continue at that rate, of course.

(I'm using percentage because the Domoticz software I'm using doesn't have a suitable sensor type for an absolute volume; there is a water meter which has a unit of litres (or m³) but its behaviour isn't quite what we want. I need to fix that.)

replying to David Woodhouse, efpomo wrote: Hello,

I have an Apollo oil monitor sender unit. No receiver unit. I have deployed RTL-433 exe ? i.e. Windows. I have a batch file as follows to start RTL_433.exe ____________ @echo off cd C:\ADSB\RTL 433\Daily rtl_433.exe -f 433900000 -g 40.2 -m 3 -p -39 -G -F json 1>fullscan.txt pause ____________

I have also run the batch file without the ?m 3 parameter.

The gain is set at 40.2, which receives a radio thermometer in my house, as follows:

{"time" : "2016-11-01 17:17:33", "model" : "TFA pool temperature sensor", "id" :

228, "channel" : 1, "temperature_C" : 4.100}(I put the thermometer in the fridge to ensure that it was my thermometer & watched the Rtl_433 output as the temperature dropped)

So, it seems to me that my dongle receiver works ok.

I have touched the Apollo with a magnet to get quick data transmissions ? the led pulses ? the lcd display shows lines - But, nothing received.

I have also deployed the magnet for a longer period, in efforts to change the transmitter id.

Assuming that the transmitter is not faulty, can anyone give me some pointers on how to proceed. I am also assuming that the transmitter does not need to be "paired" with a receiver

I should mention that at max gain i.e. 49.6 I can pick up the ?Watchmans? below, with a constant i.e unchanged id

I am reasonably certain that they are from neighbouring installations.

{"time" : "2016-10-30 17:00:23", "model" : "Oil Watchman", "id" : 672018617, "flags" : 128, "maybetemp" : 20, "temperature_C" : 15.000, "binding_countdown" :

0, "depth" : 66} {"time" : "2016-10-30 17:10:15", "model" : "Oil Watchman", "id" : 138674681, "flags" : 128, "maybetemp" : 21, "temperature_C" : 13.333, "binding_countdown" : 0, "depth" : 29} {"time" : "2016-10-30 17:26:19", "model" : "Oil Watchman", "id" : 138674681, "flags" : 128, "maybetemp" : 21, "temperature_C" : 13.333, "binding_countdown" : 0, "depth" : 29} {"time" : "2016-10-30 18:15:50", "model" : "Oil Watchman", "id" : 672018617, "flags" : 128, "maybetemp" : 20, "temperature_C" : 15.000, "binding_countdown" : 0, "depth" : 66} {"time" : "2016-10-30 18:30:32", "model" : "Oil Watchman", "id" : 138674681, "flags" : 128, "maybetemp" : 21, "temperature_C" : 13.333, "binding_countdown" : 0, "depth" : 29}

Could it be as simple as the the battery in the oil monitor sender going flat? It may have enough power for its local LCD display which requires next to nothing to power but insufficient power for a radio transmission.

You may have to completely remove the sender unit from the oil tank to reveal the screws to get to the innards. The battery is likely to be a (less popular) coin cell.