MC9S08GT16 can not get Low Power mode stop3 to work.

[randenewberry]

I have followed AN2493, and think I have tried everything, but I can only get low power unless Vcc < 2V then Stop3 Icc is a few microamps.

But at 3V, Sleep current in Stop3  is 70 to 100uA, at 3.3V it's over a Ma..

I have removed every component, except the GT16 and some switches, and still get the same results.

I've tried all unused ports both as outputs low, and inputs with pull up's enabled, still get the same results.

I need to wake up with a keystroke, so can only use stop3..

Any Ideas?

[mfugere]

I don't have experience using this chip in this mode - but do you still have the BDM enabled?   That still gets clocked/remains active in stop3 mode.

Are you running the system entirely from internal oscillator, or using external reference?

[randenewberry]

Hi Michael,

I disconnect the USB Multilink after debug programming. I know with it

connected, I get 5-6mA. So I disconnect and re-power up.

I'm using the internal osc.

I do have a workaround, which is to put a diode in line with the power

to the GT16, drop Vcc by some . The application needs batteries to be at

least 2.5V for lighting the LED's, so dropping the V at the GT16 by .7V

or so I can live with the sleep current.

Only drawback is that's another spin of the boards...eeek.

I am surprised no one has had this issue with the GT16.

Thanks for the info!

Rande Newberry

H-ITT, LLC, a Hyper-Interactive Teaching Technology Company

420 Shearer Blvd.-Cocoa, FL, 32922

PH: 888-322-0089

rande@h-itt.com

[mfugere]

Is it possible that to achieve the lowest Stop3 mode current, the BDM logic in the chip itself also has to be disabled?  From the AppNote-

"

Also, it is important to note that if LVD is enabled in stop or entry into background debug mode is enabled (ENBDM bit in BDCSCR

is set), the only stop mode that can be entered will be stop3. When the ENBDM bit is set and a stop instruction is executed, the system clocks to the background debug logic remain active so background debug communication is still possible.

"

Granted, it does leave you "blind" as far as background debugging goes, but in shipping/production FW that isn't a big deal.

I am just getting into working with these newer 9s08 parts, so I'm still figuring things out myself.

[randenewberry]

It did not seem to have any affect..

I think I'll add the diode..

Thanks,

Rande Newberry

H-ITT, LLC, a Hyper-Interactive Teaching Technology Company

420 Shearer Blvd.-Cocoa, FL, 32922

PH: 888-322-0089

rande@h-itt.com

[vicentegomez]

Hi Rande

Please check that all the GPIO are set to a known state, (output in low is recommended) also check the package that you are using , if you are using the qfn 32 package, you need to set the PTG3 to know state even that this pin is not physically on the package

Regards

Vicente

[randenewberry]

Hi Vicente,

It's a 44 pin package, but that is good info.

I've found that Vcc has a lot to do with my sleep current...it's about

1.8uA in sleep3 with a Nordic nRF24LC01P also put to sleep...

ONLY when the supply (batteries) are 2.3V or less..

But at 3.3V supply the supply current can be over 1mA in sleep mode...At

3.0V Isleep can be 30-50uA....it's not linear at all..

I even removed the Nordic, same results..

I have several products that use the Nordic transceiver with MC9S08QE8

and MCF51QE32LH processors and all sleep correctly, under 2uA at all

in-spes Vcc's.

I gave up, my solution was to put a diode between Vcc and the batteries

(2 AA's), so the rest of the circuit gets full bat power (important

since I'm driving white LED's), and the processor gets less Vcc and

acceptable supply current..

Thanks for the help.

Rande Newberry

H-ITT, LLC, a Hyper-Interactive Teaching Technology Company

420 Shearer Blvd.-Cocoa, FL, 32922

PH: 888-322-0089

rande@h-itt.com