Mach X Erratic supply voltage to target

[Mississauga]

Hi,

I have a Mach X ICD attached to an 18F6490 via an RJ45 on a target board. It was working sort of OK, though EEPROM writes seemed incredibly long even for 1 or 2 bytes, but when I checked the +5 voltage on the board it was only 4V. I noticed the green LED on the programmer was also quite dim.

I checked the voltage on the USB cable, 5.07V. I reconnected the Mach X and the lights were brighter, and the target output voltage was 4.9. I reconnected the target, and the LEDs on the target were extremely bright (briefly) and the processor ran (briefly) then everything went dark on the target (the programmer box still has LEDs). (There is nothing connected to the target board other than the ICD.)

The +5V I measured at 7.4 on the target board, then I noticed the MCU was hot which it had never been. I disconnected the ICD, and checked the voltage on its output cable, 7.6V with nothing connected. I connected and reconnected the ICD, and turned on and off the target power selection from the PCW debug window and also from the separate Mach X controller. The voltage to the target bounced around from 0 to 4 to 5 to 7.6 depending on the connection but without any pattern I could figure out.

The MCLR programming voltage seemed steady any time I checked it at just over 13V, and was not affected by other changes (I don't know what it was to begin with, or whether it is always supposed to be that high - it's not outside the spec)

I reconnected the board and the +5 falls to +4 and the MCU gets very warm immediately (5 seconds).

Questions:

1. I assume the MCU is now toast. Is the Mach X also dead/dying? What might have killed them all?

2. Should the MCLR voltage always be on, and always that high? I assume not, because that voltage is not specified as required for normal operation. Is the Mach X fixable, or is it best to return it (it's only 2 weeks old)?

3. Can disconnecting/reconnecting the Mach X from the computer USB get it confused enough to supply the wrong voltage?

4. Even though the MCU was zapped with a relatively high voltage, the spec says 7.5 is the maximum and I believe the bad voltage from the Mach X was not over that by much if at all. Does anyone know if that level of overvoltage normally fries a 5V 18F PIC?

5. Everything seemed to be working pretty well with the Mach X and compiler up to this point, though now I'm nervous about trusting the Mach X with the target hardware. Perhaps it's safer not to let it supply the target?

All firmware/software was up to current versions as of three weeks ago.

Thank you for any suggestions or recommendations.