Newbie puzzled by HC908 programming

I'm a complete neophyte when it comes to hardware, so I'm sure I'm
doing something completely stupid in the way I am trying to program a
Freescale HC908QY4CPE.

I built the programming circuit shown in the manual and have tried
downloading a program using the protocol documented in the manual. I
have checked and rechecked connections, voltages, parts. The
oscillator is oscillating, the bits from the serial line are getting
to the right input, the 9V is getting to the IRQ line, and so forth.
But the device sits there like a lump. I get the backtalk echo from
the '125 hooked to the '232; I can see the security bytes and commands
arrive at PTA0, I can see the 9.8304MHz on PTA5, the static voltages
on the other pins are as per the manual; and I get no response at all
from the ROM monitor. It has to be something I'm doing wrong as I've
swapped the '908 with another with the same results.

I know I need to sacrifice a goat when I'm configuring a SCSI chain --
and DigiKey doesn't carry them! -- so what am I obviously forgetting
here?

Thanks for any advice,

Brian Hetrick

On Apr 1, 5:49 pm, "bhetr...@gmail.com" <bhetr...@gmail.com> wrote:
> I'm a complete neophyte when it comes to hardware, so I'm sure I'm
> doing something completely stupid in the way I am trying to program a
> Freescale HC908QY4CPE.
>
> I built the programming circuit shown in the manual and have tried
> downloading a program using the protocol documented in the manual. I
> have checked and rechecked connections, voltages, parts. The
> oscillator is oscillating, the bits from the serial line are getting
> to the right input, the 9V is getting to the IRQ line, and so forth.
> But the device sits there like a lump. I get the backtalk echo from
> the '125 hooked to the '232; I can see the security bytes and commands
> arrive at PTA0, I can see the 9.8304MHz on PTA5, the static voltages
> on the other pins are as per the manual; and I get no response at all
> from the ROM monitor. It has to be something I'm doing wrong as I've
> swapped the '908 with another with the same results.
>
> I know I need to sacrifice a goat when I'm configuring a SCSI chain --
> and DigiKey doesn't carry them! -- so what am I obviously forgetting
> here?
>
> Thanks for any advice,
>
> Brian Hetrick


Brian,

The most annoying problem I found with using the Motorola HC908
devices is that after a security failure you MUST turn off power and
let the supply voltage drop close to zero (<0.7V or so) or it will
remember the security failure and refuse to talk any more.  I often
had to put additional load resistors across the power supply so I
didn't have to wait for 20 seconds.  Apart from that the debugging
scheme worked well.

kevin

In article <1175474944.134982.90180@o5g2000hsb.googlegroups.com>, "bhetrick@gmail.com" <bhetrick@gmail.com> writes:
> I'm a complete neophyte when it comes to hardware, so I'm sure I'm
> doing something completely stupid in the way I am trying to program a
> Freescale HC908QY4CPE.
> 
> I built the programming circuit shown in the manual and have tried
> downloading a program using the protocol documented in the manual. I
> have checked and rechecked connections, voltages, parts. The
> oscillator is oscillating, the bits from the serial line are getting
> to the right input, the 9V is getting to the IRQ line, and so forth.
> But the device sits there like a lump. I get the backtalk echo from
> the '125 hooked to the '232; I can see the security bytes and commands
> arrive at PTA0, I can see the 9.8304MHz on PTA5, the static voltages
> on the other pins are as per the manual; and I get no response at all
> from the ROM monitor. It has to be something I'm doing wrong as I've
> swapped the '908 with another with the same results.
> 

I had similar problems with the QT4 (but interestingly, not the JB8)
during my early days with the HC08 series.

Is the QY4 programmed or unprogrammed ? If it's unprogrammed, then the
following suggestion is not applicable because it should go into monitor
mode anyway:

Try powering up the QY4 after the rest of the circuit is powered up.
The MAX232 may not have raised the voltage on the IRQ pin to the required
level by the time that the QY4 samples it during startup.

However, regardless of whether it's already programmed or not, you imply
that you have created your own program to do the downloading.

This is what I have also done, but to make sure that the programming
circuit was working first, I started by using the free download of
CodeWarrior for the HC08 to test the circuit.

Simon.

-- 
Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP
Microsoft: Bringing you 1980's technology to a 21st century world

Thank you to both replies.

I now have failures with three implementations of the circuit with two
sets of parts, two QT4, and one QY4, and both homebrew loading
software and CodeWarrior. One of the QT4s is now a small but energetic
hot plate, so I assume I fried it at some point; all the other ICs
I've used in breadboarding various diagnostic circuits act like their
datasheets say they should, more or less; and all the 908s sit there
like turnips, or in one case, a quite hot turnip.

I'm clearly doing something really stupid like counting pins clockwise
instead of widdershins, and will seek a local guru who can look at the
circuit for two seconds and tell me what's staring me in the face. Or
I suppose I could cut my losses and buy a demo board...

Again, thanks.

Brian Hetrick

I tracked this down with some dredging through the Freescale user
forums. The failed attempts at programming used a USB serial port
dongle. When I used an old box with a real serial port, everything
worked immediately exactly as intended.

The user forums say USB serial ports are a leading cause of
programming failure. Are the things half-duplex, or do they use wildly
inaccurate time bases for baud rates, or something?

bhetrick@gmail.com wrote:
> I tracked this down with some dredging through the Freescale user
> forums. The failed attempts at programming used a USB serial port
> dongle. When I used an old box with a real serial port, everything
> worked immediately exactly as intended.

Did you try to use the latest drivers for your USB-serial cable?

> The user forums say USB serial ports are a leading cause of
> programming failure. Are the things half-duplex, or do they use wildly
> inaccurate time bases for baud rates, or something?

The USB to serial dongles are trying to packetize the communication via 
the USB part. Due to that, there could be a big request - response delay 
on the serial port. This causes the RS232 handshake problems with some 
applications. What is funny this problem typically happens with the 
older embedded development stuff, such as eval boards or ICEs.


Vladimir Vassilevsky

DSP and Mixed Signal Design Consultant

http://www.abvolt.com

 Get mcu h/w that others tout as

 easy to start .

  I have many ARM 7's .

 I will ignore  all the software and simply

flatten the  pins on several old HC chips

 and SRAM chips and build

 a state machine to boot the ARM mcu's.

  Its actually faster .

  My state machine has 3 levels , so

 it has max leverage .

 First SRAM has the primatives ,

 2nd has the address lists , 3rd is

 very hi-level .

   This way , dont have to create a proceedure

  as you would in C/C++ .

    With only a few primatives , i can trial-error
 boot any ARM , if it fails boot
 , i dont have to alter  any primatives ,
 cause they are a ROM Emulator and
 there is only need to rearrange ( alter the
 addr lists ,) addresses .

  ARM thinks it is branching , but it's in
 the state machine , because there is where
 the real addressing sequence is .