PLL on AP32 is not functioning

[B_Conf]

Hi,
I´m using a 908AP32CB, and I´have a problem, when I try to use a XTAL of 32.768 KHz it doesn´t work, but if I use a XTAL of 9.8304Mhz it does.
I´m using CW 5.0 and I enables by bean, the PLL in all speeds when I use the 32.768Khz Xtal, I´m using all recommended values, it means, 10Mohm, series 330K, 15pF caps, and CGM circuit 1Kohm 220nF and 10nf. I´ve got two AP32, and the results is the same.
I saw all messages refered to PLL, but...
More I probed an external oscilator of 32.768Khz and the same, The AP32 doesn´t work!
I probed with three differents Xtals.
I need a bus of 8 Mhz!
So if anyone can help me I will appreciate a lot!.

Thanks and regards!

B_Conf

[rocco]

Hi, B_Conf:

Could it be that your 908AP32 parts are really 908AP32A parts? The A part only supports a crystal in th 1MHz to 8 MHz range, which could explain why it might work at 9.8304MHz.

I made the mistake of assuming the A part was just a newer revision, to later find that the oscillator was fundamentally different.

[B_Conf]

Hi Rocco,
Thank you for your response.
I review the AP32 and it is a MC908AP32CBE, below 2L47PCTGLQ0611, I supose its mask is 2L47, I think the problem is a bit strange, don´t you?, I have the same problem when I made my own programmer and doesn´t function with the 32.768KHz XTAL.
I noticed that the reset pin with 32K Xtal does not rise never, it always at 0V, it is like the internal oscilator doesn´t count the 4096 cycles or 4096 ICLK (Chapter 7.2.2), but I don´t know why it happens.
Any suggestion?

Regards

[Alban]

Hello,

Just a small note, I think you don't use the 32kHz to program, but the 9...MHz anyway. (32kHz only works in one special case, see datasheet).

I strongly advise you to enter a Service Request (follow link in my signature).

Please do include all the details you perfectly exposed here and don't hesitate to add a link to this thread so they can see what has already been done.

The manipulation you do is really simple and the fact it doesn't work is uncommon.

As you say, it is bit strange.

Kind Regards,

Alban.

[B_Conf]

Hello Alban,
Thank you for your suggestion.
You have to know that my first circuit to program the AP32 was with the 32K Xtal, and my AP was Blank, and it doesn´t work, I think this is the "Special case" you mention, so the problem wasn´t a not blank chip, of course I use all pins with its voltages level´s appropiates.
I changed the values of CGM circuits(Are differents in Monitor circuit and suggested in Chapter 6.3.9), but the same.
I will appreciate if you send me more details of the PLL so I could analyze its Phase margin, to determinate its stability and what could be the adecuates values of the resistor and caps in CGM circuit.
Now I´m going to do some test and then I´ll be back with the results. I thinnk the problem is the mask 2L47P, but I don´t know really

Regards

Pablo.

[Alban]

Hi Pablo,

The fact that nothing is happening is not linked to the PLL, as the software doesn't even seem to execute any instruction, it would not go and set up the PLL.

When the device is blanked, it is executing some PLL setup code which has been checked by Design.

The source code of the ROM side is not available.

I searched using the mask set and it returns both the AP32 and AP32A for the same mask.

That could mean either both High and Low speed oscillators are on the same die, or it could mean the Errata Sheet for the mask has been attached to the A version by mistake.

I am curious and would like to understand as well. If you want, you could send me the SR# via Private Message.

Cheers,

Alban.

[B_Conf]

Hi Alban,
As I said I did some tests, and I found that The PLL doesn´t work in automatic mode, but in manual does, the problem is that sometimes the PLL doesn´t lock and the uC stop working and I have to reset the uC.
On the other hand, I had to put an external oscilator with 32K XTAL because of the circuit suggested (Pierce configuration conected to OSC1 AND OSC2 )doesn´t work.
The problem is really strange!!!
Now, I´m going to send a S.Request.

Cheers

Pablo.

[Alban]

Hello,
Now your problem is logged, someone from FSL will look into it.
The last point is interesting.
If the manual mode tells you that the PLL does not lock, it could be because of a PLL filter issue.
The fact that it stops working could simply be because you put an infinite loop waiting for this bit, therefore the CPU is working but waiting on this bit. Have you checked this ?

You can observe the bus speed of the MCU to see how the clock is changing.
To do so, configure a Timer/PWM channel to output the fastest square wave (50% duty cycle).
Then probe the pin with an oscilloscope and show the Frequency of that square signal (1/t).
When you set the PLL, you will see the frequency rising to the target and then oscillating towards it.

As you do the setup, you know the target frequency.
From the scope trace, you can see:
- The trace never gets close to the target: either your PLL range/multiplier selector is incorrect OR PLL problem.
- The trace goes close but oscillates a lot (greater than +-5% from the target). The PLL does its job, but the filter is too coarse, put a lower capacitor.
- The trace goes close and oscillates a little (lower than +-3% from the target). The LOCK bit is defective OR LOCK bit not read properly in software.

Cheers,
Alban.

[B_Conf]

Hi Alban,
Now the AP32 is working!, but in manual mode and differents values of CGM circuit, and I have to use an external oscilator of 32.768 KHZ, (I never thought that automatic mode couldn't work).
I could comunicate the uC with a display through IRSCI interface at 19200 baud, a small step!
I didn´t modify the code.
Thank you for all your suggestions.
Cheers

Pablo