In my home automation system I use a Fibaro Door Window Sensor (FGK-101 https://manuals.fibaro.com/content/manuals/en/FGK-10x/FGK-10x-EN-T-v2.0.pdf) connected to a pressure mat sensor Arun PM3 (http://www.arun-electronics.co.uk/pdf/Pressure Mat Data Sheet - issue 1.pdf). The use case is to detect when someone goes up the stairs so I can turn on an LED strip to illuminate the staircase. This is a common use for the Fibaro Door Window Sensor which provides a potential-free input (see page 8 of the manual, #5 Button or external sensor). This works all fine and the sensor sends the signal to my home automation hub when the pressure mat sensor is activated. The issue I have is that I want to wire the Fibaro sensor to a PSU since I don't want to have to replace the battery when it runs out. The Fibaro sensor uses a Saft LS14250 1/2 AA Lithium Battery (DC 3.6v 1200mAh). I bought this PSU https://www.amazon.co.uk/gp/product/B00V2YFX3M/ from Amazon UK which provides 600mA at 3V. The problem I have is that as soon as I replace the battery with the PSU the Fibaro sensor starts detecting as if someone was jumping up and down in the pressure mat sensor (basically it starts detecting the pressure mat going on and off constantly). I tried with another power supply, a DC-DC Step-down PSU with the same results, even when I tried different voltages from 3.6v till 3.1v. I measured current and voltage on the Fibaro sensor input both using the battery and the PSU and they were the same in both cases: 1.5v y 0.03 micro amps. The Pressure Mat Sensor data sheet says the max carry current is 250mA so 0.03micro amps shouldn't be causing problems. Any ideas what could be happening here and how to prevent it?

Many thanks!

 

Try putting a 10μF electrolytic capacitor across the power supply lines. Make sure you connect the capacitor the correct way, -ve polarity to -ve side of the power supply.

 

Those multi voltage psus use a linear Transformer and small smoothing capacitor, which gives ripple,,
i would try a 470 to 1000uF electrolytic capacitor across the output and maybe a 100nF also.

 

Battery is pure DC. Power supplies are not pure. Not unless you over produce then regulate them way down to below RMS value of some voltage, dependent on the type of PSU. Switch mode power supplies (SMPS) don't have the ripple, they have hard ON and hard OFF periods that can occur a thousand times per second. A sensitive sensor may go absolutely nuts with such a supply. Like others are suggesting, throw some capacitance on it. But not too much, as that can have negative results too. Oh, and I'm not the guy to ask about what the right cap would be. They're mysterious devices to me. Well, not totally, but I don't know enough about them to offer any real advice. So maybe I should shut up now.

 

Those multi voltage psus use a linear Transformer and small smoothing capacitor, which gives ripple,,
i would try a 470 to 1000uF electrolytic capacitor across the output and maybe a 100nF also.

I would take that route.

Ron

 

Thanks for the info guys. I have bought several capacitors over the web and I will report back once I test them.

 

Tantalum caps are good for this. A 1uF with a .1uF in parallel should do well.

 

Hi, finally had a chance to try this out. I used 4 different electrolytic capacitors:

10v 10uF
16v 470uF
50v 680uF
6.3v 1000uF

With the first one the sensor will not power. With the other 3 it will power for an instant and then shut off. Any ideas? Thanks

 

It sounds like you are connecting the capacitors in series with the power supply to the unit. They should be connected in parallel with the power supply.

Les.

 

Hi, yes I was connecting them in series. I have now tried connecting them in parallel with the power supply, basically to +/- of my power supply the same as my sensor. The sensor now works but I still get the sensor to misfire (ie my original problem). Do I need a bigger capacitor? Is the capacitor's voltage important or does it just need to be higher than my circuit's voltage (3.6v)? Thanks

 

The capacitor voltage rating is not important so long as it is more than 3.6 volts. You should not need anything larger than 1000 uF but I think it would be worth also connecting a 0.1 uF (100 nF) mylar or ceramic capacitor in parallel with the electrolytic. Install this capacitor as close to the unit as possible. (This is because this capacitor is to filter out high frequency noise.) Another thing you could try is a ferrite ring on the cable from the power supply to the unit. Does the power supply have an earth pin to connect it to the mains. (You do not give any idea in your profile of where in the world you are located so we have no idea about eathing arangements in your country.) I either side of the DC output is not connected to earth via the mains plug you could try earting the negative output of the power supply.

Les.

 

Hi Les. I am in London, UK so here we have 220v AC at 50 Hz. The power supply is not earthed, so I will your suggestion of earthing it on the negative output of the power supply along with two capacitors (0.1 uF + 1000uF) in parallel connected in series to the output of the power supply. On my tests I connected the capacitors directly to the power supply output and then had a short 15cm cable run to my sensor. I can increase this cable run if you think that will help. Thanks for the tips!

 

If possible can you look at the PSU output on a scope, or the ac range on a good dvm. My hunch would be as the matt is designed to be battery powered hence very low power, and the caps are not working the PSU needs some more load, place a resistor across the PSU. 30R will draw 100mA just as somewhere to start. Just adding a power LED might do the job.

 

Scope is always your friend in these situations. Just make sure you do it under real conditions with the load connected. Take care if earthing Class 2 appliances that are not meant to be earthed. Some of these PSUs can float at unexpected voltages. Try the same setup with a bench lab-type PSU and work backwards. There are a lot of iffy plug top PSUs out there at too good to be true prices. Same advice, get a decent one from the big suppliers - Digikey, RS, Farnell to start with.

 

Before you go to the effort of using a scope on the power, use an accurate digital meter to check the supply voltage, since most of those small cheap low power supplies are not regulated. Your sensor device probably needs very close to a specific voltage to avoid becoming confused. Adding a capacitor to many of those supplies will cause the output voltage to increase a bit, depending on the load. So there is another thing to watch.