Swapped the upper and lower traces. No DRC errors.

Was also able to get the switch in, after much trouble.

 

Is that a heatsink for the U2 regulator outlined in yellow?
I'm thinking of something like this:

 

Not using the 7806. That position will not be populated.

And yes, that was the heatsink I was going to use.

 

PCB is assembled. Time for troubleshooting.

Here is the schematic.



Not that it matters, blank boards.



Assembled board




I'm getting 5v output. When the low probe is put in the water the low LED comes on, nothing else happens. When the high probe is put in, nothing happens. When the manual button is pressed, nothing happens. The pump or pump LED never does not come on.

The Kicad checker on both the schematic and PCB came up with no errors. I verified the LED's worked, measured the resistors and checked the manual switch. Assumed other components were OK. Going to dread unsoldering anything. I still have the circuit breadboarded.

 

I would check the voltages on transistor Q1 for starters.
When OFF: base = 0 volts
collector = 5 volts
When ON: base = 1.2 volts
collector = 0.7 volts

 

Hi,
I haven't read all of the post, but have you considered that when the boat is at say 45Deg, the floats may not work properly?

If you had a permanently running peristaltic pump, this wold self prime and pump any water as it arrives. (no electronics)
C

 

In the schematic diagram it looks to me like the diode (D4) across the relay is backwards. That will prevent the relay from operating.
In addition, all of the PCB tracks are very skinny.

 

I would check the voltages on transistor Q1 for starters.
When OFF: base = 0 volts
collector = 5 volts
When ON: base = 1.2 volts
collector = 0.7 volts

When off base is 0v collector is 5v
When on base is 1.1v collector is 3.45v

I haven't read all of the post, but have you considered that when the boat is at say 45Deg, the floats may not work properly?

If you had a permanently running peristaltic pump, this wold self prime and pump any water as it arrives. (no electronics)

Not using float switches. We have wires as probes to detect the presence of water. And, the sailboat is not always leaning over.

In the schematic diagram it looks to me like the diode (D4) across the relay is backwards. That will prevent the relay from operating.
In addition, all of the PCB tracks are very skinny.

Yea they do look skinny. In the PCB editor they did not look skinny, I actually increased them from 10mil to 15mil and had to adjust them. I didn't check how wide 15mil was. This is the first time I had a PCB made from a vendor. I know better now for next time. In the breadboard that sghioto helped me with, the cathode is on pin 16 and that circuit worked as breadboarded. So not sure if i should switch it or not.

sghioto above voltage checks may help?

 

the diode (D4) across the relay is backwards.

In addition to that comment - Q1, when activated by SW1 will be drawing all the raw current to ground (or common as the symbol shows). Q1 may now be blown out.

all of the PCB tracks are very skinny.

Yes, they are. And you may have a burned out trace, though that would be easy to find.
SW2 if activated takes power from SW1 to GND. Granted, it's not likely you'll operate both switches at the same time, if Q1 is powered from either SW1 or from the relay itself through D5, what does SW2 do for you? It will ground power. I may be misreading the schematic.
The idea is to switch K1 on when the high water sensor activates it. Once activated it remains held on via the low water sensor. When the low water sensor no longer senses water the relay shuts off and the pump stops running. I'm reading this and ignoring the presence of D4 all together.
Finally, does your pump have three wires? Can you run the pump by hand on the bench by applying power to pins 1 & 2?

 

SW1/2 will not be operated at the same time.

The idea is to switch K1 on when the high water sensor activates it. Once activated it remains held on via the low water sensor. When the low water sensor no longer senses water the relay shuts off and the pump stops running. I'm reading this and ignoring the presence of D4 all together.

That is exactly how it works on the breadboard setup.

Finally, does your pump have three wires? Can you run the pump by hand on the bench by applying power to pins 1 & 2?

Two wires and the pump works fine. It's a brushed pump.

Even though this all worked on the bench following the schematic it seems there are flaws? What do I do now?

 

Is D4 installed correctly?
I suspect it's defective.

 

D4 cathode is connected to K1 pin 1 and Q1 pin 2 as shown in the schematic.

Should I pull it and replace it then? Or can it be tested somehow soldered in place?

 

From your description D4 is in backwards.
Anode to K1 pin1 is the correct connection.
I would remove and replace it.

 

OK will remove and replace. I will be out the rest of the day so will do tomorrow.

 

I suggest that you do not require D4 because Q1 will not switch off rapidly, since C2, 47 mfd, will not allow the base drive to decrease very fast. So there will be no spike voltage produced.
AND the inductive spike is only produced with the fast current drop. Which will not happen.

 

So then remove D4 altogether and not replace it.

 

Go ahead and try that.

 

That works with no D4.

The probes work as they should and the manual switch works. Have to connect SW2 with cables to try that out. The circuit draws 0.194 amps from the power supply when active.

 

You should make the correction on the schematic referring to D4.
D4 is basically optional but if deciding to use it the schematic should be redrawn correctly. ]

 

Yes will correct it, and will note that on the schematic.

I have 5 PCB's and enough components now to complete 2. So since this one is working will build a second one, although I need to order a pump. I'll be able to use the spare one in another model.

I started designing the 3D printed bracket for the PCB and pump. Will show that once I'm done. Also have to make the rig for the tubing/wires that extend into the keel.