Need to "emulate" a servo

[PinGuy]

I need some ideas.
I am troubleshooting a water warmer that uses a servo to control gas flow and get the desired water temperature.
The problem is that the servo is broken, and I can't find a replacement 'soon'.

It has 4 wires, 2 go to the motor, and the other 2 to one of those "glass" magnetic switches, that gets switched on and off as one of the gears (which has magnets) spin.
If the heater does not sense the switch pulses, the display shows an error "E2" and shuts off. If I manually "emulate" the switch closures, by continuously shorting and unshorting its wires with half a second delay, the heater starts, and I can get a "manual" control of temperature by spinning the gears to the desired position. It works as a temporary solution, which is what I'm needing.
The question is how can I, with the two wires that go to the motor, emulate the 1/2 second pulses needed by the MCU?
Neither of the motor wires are positive or negative, I think this is due to the change of direction needed to raise/lower the gas.
I tried using an NPN transistor with the base on one of the motor's wires (both have power when measured against ground when the heater powers on), and the switch wires at the emitter and collector, but it still shows E2.
Sorry if this is a basic question but I'm quite stuck with this

[stj]

Re: Need to "emulate" a servo

555 timer circuit?

[Curious.George]

Re: Need to "emulate" a servo

I need some ideas.
I am troubleshooting a water warmer that uses a servo to control gas flow and get the desired water temperature.
The problem is that the servo is broken, and I can't find a replacement 'soon'.

It has 4 wires, 2 go to the motor, and the other 2 to one of those "glass" magnetic switches, that gets switched on and off as one of the gears (which has magnets) spin.
If the heater does not sense the switch pulses, the display shows an error "E2" and shuts off.

The first question you need to ask (and answer to give a better perspective on any solution you consider) is why the "system" expects this sort of feedback from the field. If it regards personal safety/fire hazard, that should seriously color any "temporary" solution you consider (ever hear of "permanent temporary fixes"? )

After all, you're just "heating water", right? Can you live with unheated water for the time it takes to acquire a genuine replacement part? (if that part won't be available for months, will you leave your "temporary fix" in place for that duration??)

The question is how can I, with the two wires that go to the motor, emulate the 1/2 second pulses needed by the MCU?

Why can't you replace the switch/sensor with a reed switch and mount a magnet on the "gears" to allow it to operate closer to as originally intended? Let the MCU command the motor to jog back and forth but let the reed switch deliver the pulses (i.e., "short the contacts") that the MCU expects by sensing gear motion conveyed by the magnet(s).

[PinGuy]

Re: Need to "emulate" a servo

The first question you need to ask (and answer to give a better perspective on any solution you consider) is why the "system" expects this sort of feedback from the field.

Upon power up, the motor spins and the MCU expect the feedback from the sensor. If it doesn't detect it, E2 is shown and stays there. No safety stuff here.

(if that part won't be available for months, will you leave your "temporary fix" in place for that duration??)

Absolutely. The only drawback is having to adjust the temperature manually, but that doesn't need to change every day so it will be fine.

Why can't you replace the switch/sensor with a reed switch and mount a magnet on the "gears" to allow it to operate closer to as originally intended?

Because the servo is unassembled in order to access the gear that moves the gas valve up and down.

555 timer circuit?

I guess a 555 timer will do the trick, but this all works with a single 1.5V battery so I will have to see where to get VCC from to avoid draining it.
That's why I want to be able to use the VCC that the servo gets to "reply back". But I don't know how to drive the polarity change it gets from the MCU.

[budm]

Re: Need to "emulate" a servo

So if the circuit detects the pulse from, the motor will then will be running at full speed or the feedback is used to maintain constant flow rate (what ever that rate is)?

[Curious.George]

Re: Need to "emulate" a servo

Because the servo is unassembled in order to access the gear that moves the gas valve up and down.

You've missed the point; mount (epoxy) one or more magnets to the "moving member" that the MCU is monitoring. Mount a reed switch proximate to them so that AS THE MEMBER MOVES, the reed switch opens and closes JUST LIKE THE ORIGINAL mechanism was designed to do. Reassemble everything (discarding the bits that have now been made redundant) and you've effectively REPLACED the defective component with a "work-alike".

Thereafter, use it as originally intended (no more manually moving gears)

[It's too bad you don't own a digital camera... ]

[PinGuy]

Re: Need to "emulate" a servo

Just wanted to follow this up, as the issue is solved. I ended building a 555 timer inside a 5V 500mA power supply, and 3 wires go out of it: Gnd, +1.5v (I added an LM317 to the PS) and the "Signal" that tricks the MCU. The C battery is not needed anymore so that's a plus. The "gas valve" gear is rotated manually to change temperature when needed.
I had an LM555 and an NE555, neither one worked at 1.5V, so I used the 5V from the PS for VCC. Just in case, to the output I added a divider (with 2 resistors) to avoid sending 5V to the MCU. So that's it for now
Thank you all for the ideas!