MC9S12E128CFUE will not oscillate with pierce

cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

MC9S12E128CFUE will not oscillate with pierce

951 Views
marshalllester
Contributor I

MC9S12E128CFUE will not oscillate with pierce

 

1) used pierce osc for 5 years on older controller pcb

2) designed new pcb with some optoisolator changes and other component changes - 2 layer

3) OCS and PLL layout unchanged

4) OCS and PLL layout unchanged exactly as specified in data sheet

5) PE7 XCLKS tied directly to VSS to select pierce osc

6) xtal is ECS 8mhz Part # ECS-80-20-4

7) 22pf caps 1.0 mohm in parallel no RS series resistor

8) tried many different xtals and caps - no success

9) tried 3 different batches of processors for DigiKey Mouser and Arrow with no success

10) checked every power supply and ground pin

11) made sure VDD1 and VDD2 are connected to only decoupling caps and nothing else

12) lifted all pins except DBUG, VDD and VSS pins and OSC and PLL pins so they are connected to NOTHING - no success

13) built new 4 layer PCB and tested as above - no success

14) I was told OSC needs NO program to run correctly and should see correct OSC on oscilloscope - nothing

15) on one pcb the OSC signal would appear around 4Vpp for 30 sec then drop to 2VPP for 30 sec then drop out entirely

16) mask sets used are all: 0M06C

 

 

2125_2125.jpgGSC PCB.jpg

 

2129_2129.jpgGSC SCEMATIC_Page_1.jpg

Labels (1)
0 Kudos
7 Replies

677 Views
marshalllester
Contributor I

I think maybe I am burning out the osc circuit in the processor. Look at my XTAL output below. 3.2VPP and it dips below ground slightly. Definitly not a sine wave.

Also - is it OK to tie PE7 direstly to ground with no pull down resistor? It is tied DIRECTLY PERMANENTLY to VSS.

DSC03076.JPG

0 Kudos

677 Views
kef
Specialist I

Did it start oscillating after you increased feedback resistance? What value it is? Try lowering, amplitude should fall down.

It is not best practice, but also not big problem to tie PE7 to GND directly. Just remember to not set DDRE bit 7 and PORTE bit 7 high.

0 Kudos

677 Views
marshalllester
Contributor I

We have been working on this.

We connected a new processor and this was the waveform at XTAL.

It ran for about 1 minute then went flat never to run again.

Please find out the answer to this question.

Can an overdriven oscillator somehow destroy or burnout the OCS drive circuit??????

XTAL Waveform.JPGWe have a new theory.

I

0 Kudos

677 Views
DPB
NXP Employee
NXP Employee

Hello

The OSC circuit can be damaged by applying voltages above the absolute max. rating of the associated pins.

The absolute max. rating for VDDPLL,XTAL and EXTAL is 3V for this device.

The voltages quoted above (3.2.V and 4V) exceed this.

Please ensure that VDDPLL is connected to a 2.5V range supply.

DPB

0 Kudos

677 Views
marshalllester
Contributor I

We dont put any Voltage to any of these pins. VDPLL is internally connected to the processors 2.5V supply. All other pins are just connected to the XTAL and caps exactly like the Freescale data sheet.


0 Kudos

677 Views
marshalllester
Contributor I

Can someone out there give me a set of part numbers/suppliers for XTAL and Caps that they know work correctly????

8MHZ?

0 Kudos

677 Views
kef
Specialist I

Does ECLK pin oscillate when you power the board up with BKGD tied low? Or when you make attempt to connect to your board with BDM debugger? Oscillator should work in self clock mode when crystal is not installed or doesn't oscillate. So entering special single chip mode you can verify that MCU oscillator circuits are working by monitoring ECLK, which should further indicate that MCU is most likely powered up properly and must work.

I don't know what's wrong with your board, but you should try also increasing feedback resistance to ~3-10 MOhm.

0 Kudos