MC1321x problems

[kbu]

I am at exactly the same place with my board. I suspect though that my problem is because of the older SMAC software I'm using (it wasn't MC13213 aware), I'm going to try running the newer SMAC4.2 library, hopefully that one has the correct pin assignments for IRQ, CLK etc.

Another problem I have though is that my board uses the older folded dipole antenna (with separate tx and rx), and not the CT_BIAS for the antenna switch. Has any one used the SMAC4.2 against this configuration? (I hope its as simple as undefining the ANTENNA_SWITCH variable!)

BTW rpourzia, where did you get the KDS DSX-321G from? I've been pretty unsuccessful getting that part from the usual (avnet,future,digikey etc, in the US)

[rpourzia]

I think you might still have some solder issues. One of my boards exhibits the same symptoms that you describe. The working board, however, goes past the initialization board and IRQ gets asserted. FYI, I get 1.86V out of the regulator. Your schematic looks good to me.

[kbu]

Yep.. solder issues... finally got the board working!!

thanks much for looking through the schematics etc.

[rpourzia]

I got my first board working tonight. Here's what I did: I took a rotary tool and removed all the traces for pins 65-71 from the PCB board. Then I re-soldered the chip and sure enough, the modem initialized and all the voltages from the modem regulators look good.
Now I can go further and test the RF circuit. Watch out for those pesky pins under the chip, they are a lot easier to short than the pins around the chip.

[kbu]

Okay so I have a clean board, new chip etc. Visually all looks good.

I am able to communicate with the modem via SPI. However it's still stuck while checking if the IRQ is asserted.

After a modem reset I do see VDDD and VDDA going up to 1.4V, however I don't see a 32k signal on the CLKO line.

I've attached a schematic of my circuit.

Any ideas?

[rpourzia]

The VDDA output on the SRB is 1.86V.
I am positive that I don't have short on the external parts as I have ohmed each pin against others but I cannot speak for the pins under the chip. Since these pins are not used, one way would be to remove the PCB traces for these before placing the chip. I might try that. BTW, I have used solder paste succesully on the past witth the two chip version without any shorts ever. This new package present challenges and may require different techniques.

[kbu]

Looked below the chip, doesn't seem to be any shorts on the inner pins. but we had to pull the chip to check it and now while replacing it, it got skewed a little etc.

So back to trying this on yet another board/chip, I so wish we had a x-ray machine to see below this thing.

[rpourzia]

Here's what happens on the SRB:
In the MCUInit() function which is in the file mcu_hw_config.c, as soon as the following line gets executed, the VDDD goes from 0 to 1.86V and you see a 32.768KHz square wave on the CLKO line.

This does not happen on my system. That's why I think I have some soldering problem. I could not find the thread you were referring to regarding the soldering of pins close to the flag. I have a feeling this can be my problem althouh I was very careful.

I am going try some different soldering techniques to see if this will solve my problem.

[rpourzia]

Sorry, the line is:

MC13192_RESET = 1; /* Take MC13192 out of reset */

This is on line number 206 on SMAC V4.2 version of the file.

[kbu]

this post talks about a short.
http://forums.freescale.com/freescale/board/message?board.id=8BITCOMM&message.id=1968&query.id=4432#...
if the link doesn't work, just do a search for MC13213 and look for a post by RedMountain.

By the way what voltage do you get on VDDA?

I'm having someone remove my chip and look under the chip for shorts, my tech thinks there's definately something shorted. He also said something about the footprint of the inner pins (65-71) not really matching the layout (lets see)

[rpourzia]

I have access to an SRB board. I know that on the SRB board, the IRQ gets asserted because I go past the point where my board fails. Unfortunately I did not expose those on my own board but will do on the next rev. The question is what can cause the IRQ not to get asserted. BTW, do you see any activity on the CLKO line?

FYI, on our previous design, I had a lot of problems with crystals. The Freescale transceivers require crystals with capacitance under 9pf. When I used other crystals, sometimes the board would start and then frequency would be reset to zero. When we finally got the KDS crystals with 9pf capacitors, the problem got solved. For the MC1321x they are recommending 6.8PF capacitors. If your crystal is running fine don't worry about it but if not, you should check the crystal as the cause of the problem.

[kbu]

I tried a couple of tests, where I simply call GPIOInit() and SPIInit() and 'try' to pull the modem out of reset (I can't verify if this works though).

then I try to write something (anything) to the SPI port but I don't get a response.

At this point I'm still using the internal clock of the micro so I don't see any activity on the external crystal. (can't get to that section if the modem doesn't respond to begin with).

did you ever have any issues with solder wicking to say the reset pin (below the chip?) there are a couple of traces below the chip (on the package) and I'm wondering if some extra solder from the flag might have done some bad things there. Another thread in this group talks about that.

Per the crystal, I use 6.8pf caps on the crystal I mentioned with the 2 chip design without a problem (although I havn't checked a spectrum analyzer out put to see if its causing any issues with radio output etc) but the circuit works just fine. This is the first 13213 board so don't know about this one yet.

[kbu]

By the way what voltages do you get/feed to VDDA (hence VDDLO1, VDDLO2)
I'm confused as to why the output of the analog regulator is shorted to the local oscillator inputs?

Also curious, what do you see on VDDD?

[rpourzia]

Interesting. So you and I have exactly the same problem. We should compare notes. It would be nice if Freescale had a chip diagnostics tool or at least a troubleshooting guide for problems like these.
If I find any answers, I'll post it here.

[kbu]

The gerbers for the NCB and SRB reference board seem to show that the test pins below the chip are exposed. If you have those board you might be able to check the signals on a logic analyzer and see if atleast the SPI traffic is going through okay.
Unfortunately I don't have those boards.

[rpourzia]

The SMAC 4.1c was aware of the MC1321x. The problem is that the documentation for the SMAC has not been updated to reflect the SMAC4.2.
In my case, I use an F antenna and am pin compatible with the SRB so I can use an exiting target in the current design. If you look through the target specific files, you should be able to start from one the MC1321x targets and edit the ANTENNA_SWITCH setting to your requirements. I had to to do the same with the older chips and I don't remember it being too complicated.
Obtaining crystals for the Freescale Zigbee chips is the biggest pain. Fortunately, we bought some from KDS when we designed our first generation boards with MC13193. We still have some 20 or so left that I am using with the new MC1321x. It is a shame that this has to be such a problem to obtain these and some cases cost half as much as the whole transceiver.

[kbu]

okay so I was able to compile a SMAC4.2 app for my board, but now I'm stuck at the same place.. 'IRQ doesn't get asserted'.
gee would be nice to check the test pins below the chip to see the status of the SPI/ATTN/IRQ etc!!

BTW I use this crystal on my board '300-8206-1-ND' (digikey partno), its also quite expensive but atleast more or less available.