External Oscillator Stops when Vtest Voltage Applied on IRQ with MC908AP32ACBE

[Byar]

I have a MC908AP32ACBE chip that I am trying to get running. Whenever I apply the Vtest voltage to the IRQ pin in order to enter monitor mode the external oscillator stops working. I have removed all the IRQ circuitry and applied the voltage directly to the pin and it is still happening. If I lower the Vtest voltage to 6.0V the oscillator continues to run but then I'm not entering monitor mode. I'm using the schematic shown in the monitor section of the user manaul except the circuity connected to the CGMXFC pin matchs what is shown in the CGM section. In general the oscillator will restart after the Vtest voltage is removed.

I have probed all of the pins as I apply and remove the Vtest voltage and nothing looked wrong.

I plan on changing the CGMXFC to circuitry to see if this fixes anything but I won't be able to do that until this afternoon.

Does anybody else have any ideas?

Thanks,
Lucas

[sungam]

Hello Lucas,

se note 1 on page 116 in databook.

Normally the oscillator stops when using hivoltage entry to monitormode.

You can use the forced monitormode instead (reset vector blank).

To use incircuit programming you have to include code that erases the flash whenever you need to reprogram.

You can also use the debugging features of codewarrior vith this method (as long as you have a breakpoint). If you loose contact vith the µC during debug you can reset it vith the resetpin (no POR).

Sungam

[Byar]

I see what you mean. I missed that in the book.

Do you have any idea why the crystal oscillator cannot be used to reprogram the flash?

Thanks for the help. It's obvious now that I see the text that I missed the first time.

[Alban]

Hi Lucas,

Power sequence is also an important point in Flash programming.

When you apply Vtst, do you do a Reset or POR (Power On Reset).

I enclosed a MON08 entry graph I borrowed from the HC908GZ60 datasheet.

I was using the older AP64 (not A) with a 4.9153MHz XTAL for programming.

Hope this helps,

Alban.

[Byar]

I am doing a power on reset. From what the data sheets says the crystal oscillator is bypassed in Vtest monitor mode and is not usable. I am going to change to using an off-chip clock oscillator since a crystal oscillator can only be used to program the chip when the reset vector is blank.

This brings up another question. It's easy to find 3.3V and 5V clock oscillators. What is the maximum voltage that can be applied to the OSC1 input? I can't find a spec in the electrical spec.

I plan on using a 4.9152MHz oscillator to get the baud rate I need for monitor mode and use the PLL to run the bus clk at a higher frequency. I know the oscillator circuitry runs at the Vreg level (2.5V) but will a 3.3V or 5V oscillator work?

Lucas

[bigmac]

Hello Lucas,

If you already have a crystal of a frequency suitable for programming, an alternative to purchasing a "canned" oscillator is to construct your own oscillator using a HCMOS gate or inverter package.  One section would be wired as a Pierce oscillator configuration, and a second section of the package would be used to buffer the oscillator output and "square up" the output waveform that is fed to the MCU during programming.  The package would operate with the same Vdd as the MCU.

Regards,

Mac

[Alban]

Hi Lucas,

I downloaded the datasheet and saw this weird change.

It is easy to find 5V (TTL) canned oscillator.

When applying a square wave to the MCU as a clock, you can apply 5V when Vdd is 5V also.

Cannot you use a 4.9152MHz crystal in your application and any frequency canned oscillator for programming ?

I don't see anything forcing you to have the same clock for both, as you will not execute in MON08 but just program the Flash. The only impact is a change of baud rate.

Alban.

[Byar]

---

Alban wrote:

Cannot you use a 4.9152MHz crystal in your application and any frequency canned oscillator for programming ?

I don't see anything forcing you to have the same clock for both, as you will not execute in MON08 but just program the Flash. The only impact is a change of baud rate.

Alban.

---

I'm using the actual board for code development and may have to make changes later in the field. Is it possible to use some kind of slide switch to overdrive the crystal oscillator with the canned oscillator when I need to program and use the crystal oscillator for run mode?

The other item is that I want to reduce the amount of current consumed in stop mode. Currently I have a power fail circuit that triggers the software to configure the I/O ports and go into stop mode. If my oscillator is powered from the standby current source the current draw is much higher. If I power it from the normal power I would have to rely on capacitors to keep the oscillator running while prepared for stop mode. I'm looking into the options on the canned oscillator.

Do you know why is the external crystal oscillator can't used in monitor mode?

Thanks,
Lucas

[Alban]

Hi Lucas,

Yes, this kind of method is acceptable. You can see that you may be able to save the cost of the switch and just put a test point to apply the canned osc. Check but as the crystal does not work, this method should do.

No, I don't know why XTAL is not accepted in MON08, that's not the best idea to have removed that.

Cheers,

Alban.

[Byar]

The problem is that I want this to be flash upgradeable in the field with just a PC and nothing else.

Lucas