Cannot Program PIC, invalid Device ID

[dluu13]

I just got a couple of prototype PCBs with a PIC24FJ128GA308 on them.

I am unable to program them, with this problem:

[dluu13]

Ah, it looks like VCAP pin has 3.3V instead of 1.8V. This is different from my old working board.

The only thing that changed was that a couple of IO pins was swapped, so I don't imagine that could have fried the PIC.

Looks like I have a dead PIC anyway. Any ideas what could have caused this problem? I really can't think of anything besides maybe some ESD... It is really dry here lately.

[dluu13]

I have just read on some old threads about VCAP. Apparently wrong AVDD and AVSS can cause this problem I appear to be seeing with VCAP.

The thing is that my AVSS and AVDD range are within spec according to the datasheet. (VDD and AVDD = 3.3v). I am just using different regulators to generate each one.

One thing that I just remembered, though is that I had the regulator for AVDD hooked up wrong on my previous board. So I had taken that off and jumpered AVDD to VDD. I can't imagine that being the issue, but I will bypass the AVDD regulator and jumper it to VDD on my new board. We'll see what happens...

[gaugeguy]

According to the data sheet AVdd must be within +/- 0.3V of Vdd. This would include during power up. If the two supplies do not rise at exactly the same rate the chip could be damaged.

[dluu13]

I couldn't find that rise time reference in the datasheet. But I did get it to work by removing the separate regulator I was using for AVDD and jumping AVDD to VDD.

That seems to be in line with what you're saying.

I am using 3V3 for both regulators, just that the VDD regulator is giving something like 3.31 V and the AVDD is giving 3.300. So I'm within the voltage tolerance there. I'll check the rise times.

Edit: looks like AVDD is coming up to 3.3V about 500 us after main VDD comes on... I guess that's a problem?

Ok so it's solved. In the future, make sure that AVDD and VDD come up "at the same time". Although I could not find in the datasheet what that tolerance is. I was originally using my second regulator to power AVDD, and I wanted to use that as my "VREFH" to save on pins. But it looks like I'll have to use my second supply to connect to VREFH instead.

[Ttelmah]

Yes. This was met years ago in another thread:
<https://www.ccsinfo.com/forum/viewtopic.php?t=53052>

The basic key is that if AVdd differs by more than a FET 'drop' from Vdd,
you can trigger a lock up condition in the circuitry, and stop things from
working. There isn't a 'time' delta here, just a voltage limit.
AVdd, experiences some quite nasty current spikes as the ADC works, so
is still likely to not be a terribly 'good' source for Vref, even if you feed it
off a separate regulator. Adding extra filtering to it, is the 'limit' of what
you can really do. :(

[Gabriel]

i usually solve this by unplugging the pic-kit from the ICSP on the target board and reconnecting.
_________________
CCS PCM 5.078 & CCS PCH 5.093

[dluu13]