Hello wonderful forum folks. I'd like to ask for some assistance and review of a circuit I'm working on.

My parents have an indoor pet waterer that runs 24/7 and that seems a little unnecessary to have it running all the time. My goal is to have it only run when one of the pets comes near it and triggers a motion sensor. (And to learn some more about electronics and have a fun project!)

The pet waterer has a little submersible pump that is powered by a 12VAC 200mA power supply. In the middle of the power cord there is a male/female barrel plug which is where I would like to insert my circuit, thus making it very easy to hook up, and we won't need to cut any of the original products cords that way. [I would be fine cutting/splicing into them, but my mom would rather not.]

For the motion sensor my dad gave me a little unit originally from Harbor Freight that he was not using. It is normally powered by a 9V battery.

Attached is the circuit I've been working on. We can think of it as having four parts, and I have a few questions for each area.

**AC to DC, and voltage reduction section:
1) If I can find the AC to DC bridge rectifier part, is there anything about that is is much better then soldering up my own with 4 diodes?
2) I've seen variations on the number of capacitors before/after a voltage regulator. I've included 4 in my current circuit but would like to learn a bit more about this and if they are all necessary.

**Motion sensor and trigger signal section:
I plan to use the 9VDC from the regulator to power the motion sensor and not use a 9V battery any more.
3) I'm hoping to use the ~1.6VDC I'm measuring at the LED in the motion sensor to send a trigger signal to a 555 timer. I'd like recommendations as to if my current design in this section makes sense, and if I need both R4 and R5 resistors.

**Timer circuit section:
I'd like to have a few more pairs of eyes check my 555 circuit. I'll calculate the R1 and C5 parts to provide around 60 to 120 seconds or so of pulse time and that will hopefully satisfy the pets!

**Relay:
I've never actually worked with relays before so am having some trouble figuring out what my by appropriate. I was looking at NTE's website and saw two that I thought would work, but I'll have to see if my local Fry's Electronics actually carries them. [The Fry's near me seems to carry mostly NTE parts, but probably not /all/ of them.]

Let me know if you would like any more info on anything or if I've totally missed something with this design.
-marc

 

Look at Digikey for parts bridge relay etc.
Depending on the motor current, you may only need a small relay, you also need a diode reverse connected across the relay coil, 1n4005 etc.
The problem with NTE is as they specialize in semi part No. substitutes and they tend to be a bit more expensive that the original part No.
Max.

 

Overall, your circuit looks good to me. A few points:

You can make your own bridge rectifier or buy one pre-built. Only advantage to the later is you're less likely to wire it wrong. Not much difference in cost either - to my knowledge.

The number of capacitors before and after the regulator can probably be cut in half. Refer to the datasheet of the regulator for recommendations.

Yes, keep R4 and R5. If the motion sensor ties its output to GND when off, then you don't need R5, but for the cost, keeping R5 is safest bet. Definitely need R5 to limit current to the base of the transistor.

Add a 1N4001 diode across the relay coil to protect the 555 from coil voltage spikes. See here for more info.

Most wall warts don't have much in the way of voltage regulation. I suggest using capacitors rated 50VDC to be safe. 20V is cutting it a bit too close.

I don't think triggering the 555 on the motion sensor LED is going to work. The transistor will want to see close to 9VDC to conduct. I'd open the motion sensor up and see if can find a true output. Most PIR sensors I've played with only need about 5VDC and output a 5V signal. I imagine you'll find a PIR circuit connected to an RF circuit (since the HF item you called out is a wireless driveway detector). If you can find the output signal between the PIR portion and the RF portion and measure the voltage, we can offer some suggestions. You might get away with using a logic level MOSFET for Q1 as is. They typically accept a little less than 2VDC to activate.

Any small relay should be fine. I imagine a 1A relay would be fine. You just don't want to go too big so the coil current required by the relay pulls more than the 555 can handle (~200mA).

Note, as shown, the 555 will stay on when the motion sensor is triggered. For your application, this seems ideal. If, however, you only want the 555 triggered when motion is detected initially, take a look at this. If you add an edge-trigger, the 555 will only re-trigger once the motion sensor is off then triggered again and only after the 555 completes its current timing cycle.

Good luck.

 

I can speak with some authority on this topic, having owned one of these things for ~20 years of continuous service to 2 cats. Early on we used to have ours connected to a Clapper, which we would activate when we saw a cat approaching. We eventually learned the motor is incredibly reliable and just left it on 24/7. I'll bet we now have well over 10 years of always-on service from that little pump. Incredible. It still pumps like new.

A better problem to solve is the plugging of the intake. The flow rate drops quickly - within days - due to some combination of cat hair and bio-slime. It's easy to remove but accessing the pump is a real nuisance.

 

You will find the motion sensor will have a set of normally closed contacts, and will open on detection of a body, so this circuit will trigger your timer

 

MaxHeadRoom- Thanks for these notes.

elec_mech- I'll see if my local store has a pre-built bridge retifier, but if not I'll be happy so make my own and pay attention to wiring pattern. Good to know there's not too much difference besides convenience and probably a smaller footprint for a pre-built one. Thank you for your guidance and suggestions.
I think I did see that edge-triggered version of the 555 monostable circuit and figured wasn't needed just for a pet waterer. But good to know about for sure.

Dodgydave- thanks for your diagram. I'll take a more in-depth look at my sensor.

Will upload a revised circuit once I study up a bit more with the info you guys have provided.

wayneh- Aaaah man are you suggesting to not bother with the motion sensor? You're taking the fun out of this project!
I'm sure you're right since these pumps are designed to run 24/7, but I'm just going to have to build this anyway to see if I can make it work. Then if we don't want the sensor we can simply unplug it out of the system due to it's in-line plug-in design. The intake plugging hasn't come up yet luckily. Thanks for your insights wayneh.

 

The motion sensor is a fine project, go for it! I just know that as soon as I finish typing this I have to go clean my pump inlet again. I think you could argue that flowing water is attractive to the cat and that with cats you want to stimulate drinking, so maybe it's not good to turn it off. But as long as the water comes on whenever the cat gets close enough to even see the water, it should do the job.

I use a motion detector light for my cats in the basement, to light their feeding and litter areas. It works great - I'd say it's very rare for the cat to not trigger the light. I just bought the cheapest security light I could find at the hardware store and adapted it.

 

Hi all. Here's a new circuit diagram with these updates:

1. Only one cap before and after voltage regulator now. C1 also specified as 50V.
2. Added D2 (1N4001) inline from 555 Q to Relay.
3. Added D3 (1N4148) across relay -/+
4. Picked R1 and C5 to have the 555 pulse be about 220 seconds.

For the Relay I'll just have to see what available at my local Fry's. Will look for something small with a 1A rating. If I don't find anything I'll mail order one.

Here's what I'm hoping will be interesting info. I took the board out of the motion sensor. Attached voltage meter neg probe to the ground wire off the LED and then probed around various places with the positive probe. (While I kept waving it about some now and then so it would be triggered. Probably would have looked funny to anyone walking by, heh heh.)

First, on the front of the board I found two spots where touching the leg of a cap gave 2.2V when sensor was triggered. I'm guessing probably not so useful.

Then on back of board I found a large area of solder that doesn't seem to be soldered to anything on the front of the board. Measuring voltage here I got 4.4V. So I'm hoping that's enough to work with to power the base of Q1, a NPN transistor (if that's the way I should still approach this).

See blue arrow in attached photos for where I measured 4.4V.