Looking for help on a truely bizzare Rx problem

I've got a main pcb board with a GPS unit where the GPS tx pin is transmitting GPS data to pin D5 on an 18F4525 which is on the same board. Ok it doesn't work. I get no data using the the SERIN2 command. Now I have a test board with an 18F4525 on it as well. So I connect the GPS pin to this test board making sure I connect it to Pin D5 as well. Works fine. Even though I'm powering the GPS unit from my main PCB board, the test board reads the data fine (both boards have a common ground). So I know the GPS is working fine so maybe there is something wrong with Pin D5 on my main pcb board. So I send serial data from the test board to pin D5 on the main pcb board. Dang if it doesn't read it just fine! To make sure I take a spare GPS unit and power it from the test board and send the tx line to the main pcb board and it reads the GPS data just fine. What the heck!!! Pin D5 on my main pcb board works fine. Now I noticed that if I power the spare GPS unit with the main pcb board and read the tx line on the test board, I get nothing. If I power the spare GPS unit from the test board and read it on the test board's PIC it works fine. The GPS unit sucks about 45 mA and I have plenty of juice for it. To test whether or not I could have a bottle neck in the power supply, I powered a small radio (which sucks down 200mA) on the same power circuit that I power the GPS unit with on the main pcb board and it works fine. I feel like the laws of physics are bending. I'm going nuts trying to figure out this problem. Any help would be greatly appreciated.
Eric

Just a thought...

How close is the GPS unit to the MCU or OSC when they are on the same board? It is pretty easy to "jam" the GPS signal. I will assume the GPS unit has a built in antenna?

The GPS is a couple inches away from the Pic and its oscillator. However I know the GPS tx stream is working because I can read it on the test board and anything I send from the test board to the PIC on the main pcb board is easily read. Really weird. I'm going to guess that there is some interference and that it's back to the drawing board and another donation to PCBExpress.

Maybe you already decided the path forward, but let's try to summarize what you have.

• GPS on its own test board (call it external GPS)
• 18F4525 on its own test board (call it external MCU)
• GPS and 18F4525 on shared PCB (call it internal GPS and internal MCU)

• Both GPS modules are identical
• Both 18F4525 have same code

• Internal GPS does no communicate with internal 18F4525 (on shared PCB)nd internal MCU?)
• Internal GPS does communicate with external 18F4525 - again, look at the connections/traces on the PCB (i.e. did you not somehow crossed the internal traces?)
• External GPS does communicate with internal 18F4525 - are the connections identical between the internal GPS and external GPS (i.e. did you not somehow crossed the internal traces between internal GPS a
• External GPS does communicate with internal 18F4525 only when power is not shared - are you sharing bypass caps between internal GPS and internal MCU? Is internal MCU operating at all - could you have brownout or reset issues?)

How are you connecting the internal GPS to the external MCU? Are they sharing the same supply, is the internal MCU still connected, is the internal MCU running?

I doubt any of these will solve your problem, but would be simple questions to hopefully "steer" you in the right direction. I would think that if the GPS is working, even if the signal is "jammed", it would still have an output.