I created a simple circuit with a 9V power supply and a on/off toggle switch for power. I had the board made and populated it. Everything worked normally for hours. No odd behavior at all. I turned the circuit off and left it for a day. When I came back to it, I noticed that the LED was on when the power switch was set to off. I unplugged the switch connector such that pin 1 and pin 2 of the power switch are not connected in any way. When the 9V source is plugged in, I read a 2V supply going to R3, C2, and the LED. For the life of me I cannot see how this is possible. I'm sure there is a logical explanation for it but I'm still a newbie with PCB design so how this is happening is stumping me. With the power switch pins not connected, the circuit should not work at all. What am I missing here. I posted a pic of the layout so you can see how things are traced out. Any ideas would be greatly appreciated.

 

I created a simple circuit with a 9V power supply and a on/off toggle switch for power. I had the board made and populated it. Everything worked normally for hours. No odd behavior at all. I turned the circuit off and left it for a day. When I came back to it, I noticed that the LED was on when the power switch was set to off. I unplugged the switch connector such that pin 1 and pin 2 of the power switch are not connected in any way. When the 9V source is plugged in, I read a 2V supply going to R3, C2, and the LED. For the life of me I cannot see how this is possible. I'm sure there is a logical explanation for it but I'm still a newbie with PCB design so how this is happening is stumping me. With the power switch pins not connected, the circuit should not work at all. What am I missing here. I posted a pic of the layout so you can see how things are traced out. Any ideas would be greatly appreciated.View attachment 165201

Do you have a schematic?

 

I'm sure this was not the most efficient way to design it, but I'm still learning how to put together custom PCBs. I appreciate the help!

 

Occasionally there are manufacturing problems with boards and you can have a narrow thin bit of copper where it shouldn't be. It can be really hard to see without backlighting the board. Solder mask and component outlines can make seeing it more difficult and components can hide it altogether. Check across the pads for the switch with an ohmmeter with the battery disconnected. If you find conductivity when there should be none, then it is a matter of trying to isolate where it exists.

 

Occasionally there are manufacturing problems with boards and you can have a narrow thin bit of copper where it shouldn't be. It can be really hard to see without backlighting the board. Solder mask and component outlines can make seeing it more difficult and components can hide it altogether. Check across the pads for the switch with an ohmmeter with the battery disconnected. If you find conductivity when there should be none, then it is a matter of trying to isolate where it exists.

Thanks! I did poke around with my multimeter and there appears to be no conductivity between the switch pins. I'll revisit and go over the entire board to try and focus in on the problem. It was just so odd that once I finished populating the board, it worked perfectly. Now it refuses to act right.

 

Thanks! I did poke around with my multimeter and there appears to be no conductivity between the switch pins. I'll revisit and go over the entire board to try and focus in on the problem. It was just so odd that once I finished populating the board, it worked perfectly. Now it refuses to act right.

There might be a short, I see that the "9V" are directly connected to the switch and "R3". I assume the power supply comes through those "9V"?

 

Thanks! I did poke around with my multimeter and there appears to be no conductivity between the switch pins. I'll revisit and go over the entire board to try and focus in on the problem. It was just so odd that once I finished populating the board, it worked perfectly. Now it refuses to act right.

The LED lighting is another measure of conductivity. It obviously won’t light without a current path of some sort. Thanks for the schematic. There’s nothing wrong with it, and your initial success shows that. Something is hiding in plain sight. Maybe it would help to show us a closeup of your build. You might spot something while trying to get a picture.

 

Thanks for the great ideas everyone. I'm going to do some testing tonight and look for problems. It also never occurred to me that putting the power supply and switch connectors on the board would be a problem. As a possible workaround, I'm going to try putting a jumper on the switch pins and then connect the power supply to the switch itself (off the board), then plug the switch into the 9V pins on the board. That should hopefully take care of any possible shorts that might be occurring on the board. Very appreciative of all the help! Cheers.

 

Do you have an antenna connected when you see this phantom voltage? Any strong RF signals nearby? You *may* be getting incoming RF rectified by the diode. Se if it goes away if you lift one leg of the diode.

 

C2 might stay at 2 volts for some time after the power is turned off. The LED current will drop to some very low level somewhere in the range of about 2 to 3 volts, depending on the color. I would have thought the amplifier would have some resistive paths that would discharge the cap within seconds, but perhaps not. I'd have to look at the internal structure. (I did that - looks like the current through it could be extremely low with only 2 V - there are no direct resistive paths shown in the schematic on the TI datasheet, only paths through semiconductor juntions.)

Certainly if the voltage across the cap stays at 2 volts without slowly dropping when your meter is connected to it, something is providing a path. A typical DMM has an input resistance of 10 megohms, so it only takes 200 nanoamps through it for a 2 V reading. With a 9 V battery, that means the resistance allowing the current would be 45 megohms, including the meter. This is assuming that no current is flowing elsewhere. There certainly will be some, but it is hard to quantify.

 

Thanks! What I noticed was that the C2 appears to discharge once I pulled the 9V supply. Given time, it discharged completely and the LED went dark. Once I plugged the 9V supply back in (with pin1 and 2 of the switch not bridged) then it charged right back up, the LED came back on, and I was reading 2V off R3 and C2. I'm hoping that taking the 9V power off the board will take care of things. As is, the circuit works as expected with the 9V supply and power switch connected and turned to the on position.

 

Well I found one point of failure. The switch was faulty. Still does not fully explain the short situation as I had the switch completely unplugged from the board at the time. Taking the power supply off the board seems to work in getting around the issue. Now I'm waiting for new switches and solder to be delivered!