I have a Lochinvar boiler for hydronic heating in my home. It has a Hall Effect sensor that triggers domestic hot water production when it senses hot water flow from a water tap in the house. Unfortunately, the Hall Effect sensor triggers instantly and fires up the boiler even for 1-2 seconds of water flow (such as rinsing my hands in a sink). The system assumes that even a 1 second long request for hot water means someone wants to take a shower. There is no way to program the boiler to ignore short/errant requests for hot water at a tap. I would like to insert a circuit delay that would hold off firing up the boiler for 15-30 seconds. This would limit cycling of the boiler to times when someone REALLY wants hot water delivered to a tap. I've attached a photo of the sensor. I think it's a 12v sensor. I don't have a wiring diagram. Any suggestions on how to do this?

 

The service manual diagram for the boiler shows a water flow switch. That may be what you described above. Is the switch accessible to testing, as in checking for a voltage change as soon as the water starts flowing?

 

There are two types of flow switch, either of which could be Hall-effect operated:
Type 1, is just a switch, which is on when water is flowing and off when it isn't, generally operated by a flap-valve with a magnet on it.
Type 2, gives a train of pulses, depending on the flow, generally a paddle wheel with a magnet on it.
I can't tell from the photo which type it is.

 

The service manual diagram for the boiler shows a water flow switch. That may be what you described above. Is the switch accessible to testing, as in checking for a voltage change as soon as the water starts flowing?

Let me check.

 

This Circuit will provide a ~50-Second Pulse-Train-Delay.
It makes the 2 assumptions that the Power going to the Hall-Sensor is ~5-Volts,
and that the Hall-Sensor has an "Open-Collector" Output, ( which is extremely common ).

Unless the Voltage is substantially different that 5-Volts, it will work as-is.
.
.
.

 

Here's what I've measured. Three wires coming off the flow sensor are red, white, blue. When measured to ground Blue is 9v at all times (flow or no flow). White is zero volts with no flow, 9v with flow. Red is zero volts at all times (flow or no flow). Blue-Red difference is 9v at all times (flow or no flow). White-Red difference changes from zero (no flow) to 9v (flow). Blue-White difference changes 9v (no flow) to zero volts (flow). Apparently, Blue is the 9v power supply. White is the load. Does that make sense?

 

My interpretation is that the blue wire is powering the flow sensor. When flow occurs the sensor puts 9v onto the white lead which then tells the boiler to fire up. I have no means to build the circuit myself. Is there an off-the-shelf device that I can splice into the white lead that would delay the 9v signal for about 45 seconds. The device would need to auto reset if the voltage on the white lead drops below a threshold of say 3 volts.

 

Those numbers indicate that the white wire is the flow signal, blue is the supply voltage, and red is circuit ground.
The value of the white wire flow signal being equal to the supply voltage may imply that the circuit is a magnetic switch of some sort, rather than an actual hall sensor. .
... Not absolutely certain what the switching mechanism looks like.

 

Blue is +9V to Ground,
Red is Ground.
( backwards from my guess, and the usual Color-Code convention )
White is switched.

You need to do a test ........
Measure from Blue to Red with an LED and an appropriate Resistor, ( ~750 to ~1K Ohms ).
Turn on a Hot-Faucet to a "fast-drip",
see if this changes the status of the LED, you are looking for whether the LED
stays on Continuously, or Flashes very quickly, ( too fast to see on a Digital-Meter ),
if no change, increase the Water Flow slightly,
Test again, if no change, keep increasing Water-Flow in small increments until you
get either a Flashing-LED, or a Steady-On-LED.

If you get a Flashing-LED at any point in the testing,
the solution will have to be something like the Circuit that I provided above.
The only changes to the Schematic are switching the Red and Blue Wires and
changing the LED-Resistor to 750-Ohms.
I'll change the Schematic when I get the time.

If you can't make the LED Flash, and it is either steady-On, or steady-Off,
under all various Flow-Rates,
then you can use any one of a thousand different "Delay-Modules".
.
.

 

According to the service manual (page 84) the DHW flow switch/sensor is a (presumably NO?) switch that closes at 0.4gpm. It requires 9 - 10v on the blue & red wires (helpfully doesn't say which is +ve) and outputs 9v when flow > set level (plus has a red LED to show flow OK). See wiring diagram (poor) on page 88 (centre bottom of page).

Not to be confused with the primary heating flow switch (wiring diagram, page 87).

The fact that John is getting 9v when it's closed suggests it's DC out and not a pulse train. Measuring the output on AC volts might confirm that assuming his test meter is true AC, or measuring frequency if his test meter can do that.

@JohnShay where are you located?

 

According to the service manual (page 84) the DHW flow switch/sensor is a (presumably NO?) switch that closes at 0.4gpm. It requires 9 - 10v on the blue & red wires (helpfully doesn't say which is +ve) and outputs 9v when flow > set level (plus has a red LED to show flow OK). See wiring diagram (poor) on page 88 (centre bottom of page).

Not to be confused with the primary heating flow switch (wiring diagram, page 87).

The fact that John is getting 9v when it's closed suggests it's DC out and not a pulse train. Measuring the output on AC volts might confirm that assuming his test meter is true AC, or measuring frequency if his test meter can do that.

@JohnShay where are you located?

I'm located in Seattle. Everything I read points to this part of the system being simple DC. The circuitry for controlling the boiler and exhaust fan is definitely pulsed... allowing the boiler to step-wise increase BTU in 10% chunks.

 

... As a preliminary design suggestion, maybe this circuit could be modified to cause a time delay when the water flow initiates a 9 volt DC signal. The idea is that the collector of the transistor goes low when turned on, and is connected to a P channe! mosfet gate, and the main mosfet transistor current goes to the original 9 volt circuit board terminal. Some value of high drain resistor could be necessary in parallel with the large capacitor. A resistor adjustment for the current flow into the capacitor would allow the timing delay to be changed.

 

I'm located in Seattle. Everything I read points to this part of the system being simple DC. The circuitry for controlling the boiler and exhaust fan is definitely pulsed... allowing the boiler to step-wise increase BTU in 10% chunks.

I agree, its DC, I've seen nothing to suggest otherwise... In the diagram on page 88, the 'sensor' connects to a 'flow control circuit' which is itself fed from the 24v AC input and presumably generates the 10v DC. However, there doesn't appear to be any other connections, arguably to the relay K1 which controls the DHW diverter valve. I suspect there's a wire missing from the diagram from the 'flow control circuit' to the main board (as implied by the text).

Anyway, here is a simple circuit that should do what you need. If you can't build it yourself, try and find a Maker workshop in your area (I'm sure there must be many in Seattle, its practically one of the 'spiritual homes' of making). I'm working on the assumption previously made by others that Blue is +9v wrt ground/red wire.



When the flow switch output (white lead) goes high Q1 is turned on, generating a -ve trigger to the 555 timer. The output of the 555 goes high, turning off Q2 thus keeping the output to the heater at 0v (it was previously off because the switch input was off - C1 removes any glitches by slowing the rise time of the feed to the emitter of Q2). If the water is turned off (ie input returns to zero) before the 555 times out then it is reset and the heater output remains off (trace 1 below). If however the 555 times out before the water is turned off then Q2 is turned on and connects the input to the output thus starting the boiler & moving the diverter valve (trace 2). The range of the trimpot is approx 3 - 15 seconds, the midpoint being around 9 sec. Obviously, if the water is only turned on for 10sec (with a timeout of 9sec) the result will be a hot water demand for just 1 second.

 

Thank you all for the great feedback. I'll let you know how the modification turns out.

 

I have no means to build the circuit myself.

An alternative would be to plumb in another tap, fed from the cold feed to the boiler.

 

An alternative would be to plumb in another tap, fed from the cold feed to the boiler.

How does that solve the TS' stated problem of unnecessarily cycling the boiler if the hot tap is only turned on for a few seconds?

 

It allows the TS to wash in cold water, rather than hot, from a conveniently placed tap.