New PCB - adding wires to contacts make the voltage drop significantly?

[mortrek]

I'm very, very confused. I have a PCB that I designed/ordered and a relay which controls a 8.4V max voltage battery rail that's isolated from the rest of the board. The 8.4V rail is switched by a relay, and then goes to an output header. I've done this with previous versions of the board and it worked fine.

However, now when I measure the 8.4V rail output pads that lead from the relay output, they read the correct voltage (currently about 7.9V). When I solder two short wires to the pads, and then read the voltage at the ends of the wires, they read ~5V. I tested the wires and they are good, and I replaced them just in case and the problem persists.

Again, I measure at the pads, even when the wires are soldered, and I get the 7.9V. I measure at the end of the wires and it's ~5V. I'm very confused. I don't think there are any clearance issues or the pads wouldn't measure correctly, right?

Do you need schematics or anything? I'm sure they are embarrassing as I am a newbie, but I can post them if it's not an obvious solution.

Thank you.

edit: I played around with the position of the wires and it does make the voltage change. I've never had this problem before. I'm not sure if it'd be the pads or the solder or the wires or what... The pads read correctly, so it'd probably be the solder or wires, but I replaced the wires, so maybe the solder? Does some cheaper solder not conduct well? I've never had this problem with this solder before... the solder reads ~0ohms when tested. I'm still confused.

[Mechatrommer]

whats you measurement device? whats the resistance reading of each offending wire? showing pictures on how you measure it may give some light.

[mortrek]

Brymen BM235. Voltage range, and it's not properly powering the device that it's supposed to power due to the problem as well. I'm measuring with or without a load.

Resistance is measured from end to end of the wires, whether soldered in or loose. Both wires measure ~0.1ohms.

Hard to take pictures while measuring.

[Mechatrommer]

magic! the only conclusion i can draw so far from your description is.. your Brymen is broken. try another meter.

[alanb]

I would be interested in seeing what is driving the pads. Is there any possibility that the circuit is oscillating?

[ArthurDent]

That is a 2-sided circuit board and there is a possibility that the through hole plating is defective so there is almost no connection through the board. Try soldering the bare ends of the wire on both top and bottom of the board to see if that corrects the problem. If the plated hole is good then when solder is applied to one side of the board with enough heat, the solder should flow on the pads on both sides of the board.

[KaneTW]

That sounds like a flaky connection. Disconnect the battery and measure the resistance between the wires and the battery connector.

[mariush]

Also try cleaning the board, the solder joints, and the tips of your probes... sometimes flux residue can cause improper connection.

[mortrek]

Alright, I figured it out but I'm still confused... it must be a weird coincidence.

The power plug for the 8.4V rail was not making a good connection or something, but I moved it all around and tested the board on its back and front and the contacts were the right voltage but, without even moving the board, the voltage at the ends of the wires were not. However, once I suspected the plug, I started spinning it around and could make it 'come good' at the wires as well.

So maybe it just happened to reliably mess up only when I was testing the wires and not when testing any other part of the circuit? It should have shown the wrong voltage elsewhere as well... I'm still confused about exactly why it behaved the way that it did, but replacing the plug fixed it, so it must be the plug.

Thank you for the suggestions.

[jmelson]

I would be interested in seeing what is driving the pads. Is there any possibility that the circuit is oscillating?

Yes, good thinking.  One time, I had a 5 V regulated supply feeding some logic.  It wasn't working right, so I checked the 5 V with a meter.  I got a reading of 120 V, which I instantly knew
was absolutely impossible.  Noise on the DC supply was making the meter show a totally outrageous value.

Jon