• State TI Thinks Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 56 subscribers
  • Views 436 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

CC2530: 32768Hz Oscillator Problem

amir mirbagheri
Prodigy 120 points

Part Number: CC2530
Other Parts Discussed in Thread: MSP430G2553, , CC2540

Dear Sir,

I am using two different micro controllers CC2530 and MSP430G2553 in my design.

The 32768KHz oscillator is shared between the two mico controllers. 

A 32.768 KHz crystal P/N:ECS-.327-12.5-1.3FLX-C and two 15pf capacitors P/N C0603C150J5GAC7867 are used to create the oscillator circuit on MSP430 per device manual.

I connected the XOUT (P2.7) of MSP430G2553 to the XOSC input of CC2530 (P2_3/XOSC32K_Q2) . 

The other Oscillator pin of CC2530 (P2_4/XOSC32K_Q1) is not connected to anywhere.

The circuit works fine at room temperatures in almost all of them.

But sometimes at high temperature (above 110F) the oscillator stops working in some of them. 

It seems P2_3/XOSC32K_Q2 pin should be the input of 32.768KHz oscillator from CC2530 side but I could not find any explicit information in datasheet or user manual regarding it.

Please let me know if I am correct and if this pin is the input of 32768 oscillator on CC2530?

Is there any biasing or level shifting or decoupling capacitor is necessary to connect the oscillator output of MSP430 to CC2530? (currently, I do not use any of these on my design)

I checked the waveform at P2_3/XOSC32K_Q2 of CC2530 both in TI evaluation board and in my design with oscilloscope.

In TI evaluation board (which is a standard desgin), it is a sinusoidal with minimum amplitude of 450mv and maximum of 620mv while in my design (with MSP430) it is a sinusoidal with minimum of 100mv and maximum of 650mv. In other words, in my design the amplitude of sinusoidal is much higher than standard TI design.

Please let me know why most of the times the circuit is working perfectly and the problem happens only  in few of them (probably 30 percent) at high temperatures. 

Thank you,

  • 0
    TI__Guru**** 220447 points

    Hi amir,

    According to this old E2E post, Q2 is the 32-kHz OSC input.  However, neither the datasheet or TRM specifically state that an external clock input is allowed.  You should monitor the CLKCONCMD and CLKCONSTA registers to make sure the the XOSC is being used instead of the RCOSC, and be sure to reference the clock-loss detector as well.  You can also set the CKOEN bit of the PMUX register to monitor both the input/output to see whether either fail at higher temperatures.  How do you verify that the oscillator stops working during these conditions?

    Regards,
    Ryan

  • 0 in reply to Ryan Brown1
    Prodigy 120 points

    Hi Ryan,

    Thank you for your response. I have already checked CLKCONCMD and CLKCONSTA registers in debugger and both of them are always zero meaning 32K XOSC is selected. 

    Is there any flag in CC2530 registers that indicates faulty 32.768KHz oscillator (or if it is stable or not)? 

    It is good to use PMUX for testing and debugging purposes but does it solve the problem?

    I found a document from TI about using external clock signal for CC254X. Given the similarity between CC2530 and CC2540 (CC254X use bluethooth) I thought it is OK to use external clock for CC2530 as well. Here is the link to this document:

    https://www.ti.com/lit/an/swra409a/swra409a.pdf?ts=1645054203136&ref_url=https%253A%252F%252Fwww.google.com%252F

    The MSP430 program uses the 32.768KHz and goes to sleep mode. It blinks an LED with a timer interrupt. Whenever the oscillator does not work, msp430 stop working and there is no blinking LED. So I visually can confirm whether it is working or not.

    Best Regards,

    Amir

  • 0 in reply to amir mirbagheri
    TI__Guru**** 220447 points

    Hi Amir,

    Section 5.2 of the TRM describes how the Clock-Loss Detector can be enabled and monitored as a reset event condition.  PMUX was recommended for debugging, not resolving the issue.  Thank you for sharing a link to the document, this confirms that Q2 is a viable external clock input with a 100 nF capacitor in series and a 100 pF capacitor to GND at Q1.  As the XOSC from the MSP430 does not work during this behavior, I recommend you review the MSP430 design and consider creating a new thread in the MSP430 forum.

    Regards,
    Ryan

  • 0
    Expert 2170 points
    amir mirbagheri said:
    Is there any biasing or level shifting or decoupling capacitor is necessary to connect the oscillator output of MSP430 to CC2530? (currently, I do not use any of these on my design)

    Yes, see table 1 and figure 1 in SWRA409. Note that you amplitude of 550 mV is outside the requirements outlined in table 2. 

    Are you sure the MSP supports connecting other circuitry to P2.7/XOUT? From your description it seems the problem lies with the MSP430 and not with the CC2530 as Ryan mentions. 

  • 0 in reply to fredrikgk
    Prodigy 120 points

    Hi fredrikgk/Ryan,

    Thank you for your response.

    Ryan, I used the exact method mentioned in the SWRA409A on a faulty module that did not work under any condition and it solve the problem. For doing it, I output the 32.768Hz osillator on P1.0 of msp430 on my firmware and then use it as an external signal to the oscillator of CC2530.

    Previously, most of the problems happened at high temperatures (around 120F or higher) but this module has been working fine continously at 120F (inside a chamber) for more than a week without any problem. We need to implement this method on more module at high temperatures to confirm it actually solve the problem but so far, it seems it is an effective solution.The only thing is SWRA409A is suggested for CC254X but I use CC2530 inmy design. I assume the oscillator part of these two micro controllers should be the same.

    Fredrikg, in my firmware, P2.7 is defined as output per datasheet direction to work as 32.768KHz crystal oscillator output.

    Thanks,