Maker Pro
Maker Pro

Simplest mosfet circuit with ghosts....

punkieys17

Feb 7, 2016
45
Joined
Feb 7, 2016
Messages
45
So this is the simplest Mosfet circuit possible. Its in a shower pump I bought as a backup. The mains input (230vAC) comes to one side of the mosfet (BTA08 600b). A feed to the gate goes via a resistor (12k) and two reed switches which are triggered closed by magnetic floats in the pump. The pump died when the starting capacitor went. On replacing it the pump would come on on its own. I have replaced the mosfet (twice!) and the resistor. The pump has two reed switches mounted outside each end of the twin impellers pumps - one for hot water and one for cold. The switches are mounted in parallel so turning on either hot or cold will cause the magnetic float to rise and trigger the reed switch. One reed switch is connected with the black wires - and with just that connected the pump doesnt trigger till the float rises. However connecting the white wires, which are connected to the same PCB lands, causes the motor to trigger. The motor is not fully on its as if an intermittent connection is being made. However the circuit measures open circuit. The wires were originally the same (black) but changed them to the white to eliminate that as a potential error source. I have swapped where the white and black wires are soldered onto the PCB to ensure its not a simple PCB problem. There is plenty of space between the PCB lands and no question of something shorting - it measures open circuit. The circuit (BTA08 600B and 12k resistor) is exactly the same as the same commercial pump installed and working (Techflow 45). To ensure it wasn't picking up an ac circuit I wrapped the white wires temporarily in silver foil as shielding - with no change.

So I have 2 reed switches as triggers, both show no continuity until triggered by the magnets yet when I connect the white wires the motor stutters into life. If I put a magnetic field close to the reed switch while the motor is stuttering it fires into full life. Connecting just the black wire reed switches it works correctly. Connecting only the white the motor is triggered. I have even had the reed switch suspended in air to ensure no magnetic field -yet motor starts but no connection shows on a meter.

Any thoughts please. I have replaced the wires, switched the wires, replaced the reed switches, the mostfet and the resistor ..............I'm at a complete loss
 

Attachments

  • 20221111_191312[1].jpg
    20221111_191312[1].jpg
    177.6 KB · Views: 4

punkieys17

Feb 7, 2016
45
Joined
Feb 7, 2016
Messages
45
Blobs represent where wires attach to the pcb. Black wires run to microswitch at right hand end of twin impeller pump. That correctly operates the circuit. Attach white wires but no microswitch (so open circuit) the motor still responds to the black connected switch. Attach the microswitch to the white wires and the motor stutters into life. Measuring from each side of the microswitches is open circuit. Disconnecting the black wires makes no difference. The microswitch on the white wires has been changed 3 times and it is hanging in mid air with no magnetic field.
 

Attachments

  • 20221113_141007.jpg
    20221113_141007.jpg
    198 KB · Views: 7

73's de Edd

Aug 21, 2015
3,609
Joined
Aug 21, 2015
Messages
3,609
Sir punkieys17 . . . . . .
I would be first wanting to know the voltage /current/wattage and run capacitance used on that motor..
Let's do a test . . .
If it's going to be using that same SCR, I would swap the AC line input at the top so that you would then have all of that circuitry basically being at ground potential, with the HOT side, at least then being separated by the motors inductive and capacitive isolation.
Also that 12K gate drive resistor value choice seems realistically, extremely high.
In a NO power to the system condition, how about ohmming out its resistance, there might be a color registration/perception / error on reading out that third digit
Testee- Testee . . . . .


73's de Edd

Sound advice . . . . . for people who count on their fingers
They should maintain vocal silence, while additionally, remaining discretely out of view.


.
 
Last edited:

punkieys17

Feb 7, 2016
45
Joined
Feb 7, 2016
Messages
45
So in case anyone ever has the same problem (though I tend to have problems no one else ever has...) the solution was ...a new reedswitch. I still don't understand but in desperation I bought massively over-rated reedswitches ( by current and voltage). After replacing the problem switch it then performed as should be expected. It appears the problem switch was being closed by the voltage......though no current was flowing and it was nowhere near any source of magnetism...Despite a second set of microswitches I managed to rebuild the twin impellor shower pump with new seals both ends, new capacitor and new microswitches for less than £20.
 

ramussons

Jun 10, 2014
462
Joined
Jun 10, 2014
Messages
462
The schematic shows floating gate of the Triac. Not rtecommended. Connect a 12 or 15 K resistor between the Gate and the point marked L. That might solve the problem of false triggering.
 
Top