topic Re: USB Multilink Programmer fails to connect to the HCS08 target in 8-bit Microcontrollers

USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Sun, 06 Apr 2014 16:37:29 GMT

Hi everybody,

I'm trying to program/debug/erase a Freescale MC12311 transceiver IC (which has an HCS08 MCU embedded in it) via the 6-pin BDM header, using a P&E USB Multilink Universal, and CodeWarrior 10.4 IDE, on Win7 64-bit platform. The following "power cycle" error occurs when I try to program/debug/ease the target IC:

But unfortunately, turning off/on the power doesn't help, and it fails to connect to the target.

The following picture is the schematic of my circuit, in which the red circles depict the BDM header and connections:

The following is a list of the related issues I've tried:

I've added 10k pull-up resistors to both BKGD and RESET lines; but no success! The same error occurs and it fails to connect to the target.
I try to turn off/on using either the main power supply, power jumpers, reset switch, etc.; but it doesn't matter; again no success!
I checked the pins on the pcb, and everything seems OK. I mean, it seems that it has been soldered without any significant problems.
I monitored both BKGD and RESET pins: some activities (going from low to high, and high to low) are monitored after I press the "erase" button on CodeWarrior IDE.
I tried to monitor the external 32 kHz and 32 MHz crystal resonators, using a 10x probe, with both an analog oscillator and a digital counter. However, no activity is observed, nothing at all! It seems that either the crystals or the respective internal circuits are dead!
The datasheet mentions that "the MCU has an internal crystal oscillator, which is enabled in every power-on-reset, without need for external components". What I understand from this part, is that the program/debug/erase step should be done without need for external crystals. This could mean that even if both of the external crystals are dead, I should theoretically be able to erase the MCU with the help of the internal oscillator. But I don't understand why it fails!
I tried to program/debug/erase this IC using another version of CodeWarrior, such as 6.3, and also on another operating system, such as Win XP, or Win7 32-bit. But again no success!! The same error occurs, and it fails to connect to the target.
There's another circuit board which uses the same MC12311 IC with almost the same components, configurations, and schematic. I tried all the mentioned steps on that board. But unfortunately, no success again! The same errors and the same failures!
I checked the power: I provide 5-6 volt external power supply; it passes through a 3v3 regulator; then it passes through a schottky diode, and the final regulated 3.1 volt powers the IC. It is mentioned in the datasheet that the accepted input voltage range is approximately 2.5 ~ 3.7. So I don't think this "3.1 volt" would cause any problems.

I really don't know what to do next to solve this challenging problem!

What should I do? I would be really grateful if you give any helpful comments or replies.

Thanks a billion for your time.

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Thu, 10 Apr 2014 19:46:49 GMT

I wonder why nobody has replied to this question.

Just one point:

Could it be the clock problem? I mean, I can't see any activities on either pins of the external crystals.

And besides, I don't know how to check if the internal oscillator of the MCU is working properly or not.

Does anyone have any idea how to solve this challenging problem? :smileyconfused:

Re: USB Multilink Programmer fails to connect to the HCS08 target

al_muir — Thu, 10 Apr 2014 23:03:03 GMT

Can you please confirm the state of the BKGD pin (PTA4) when reset is released (rising edge of RESET) and also at POR?

Is BKGD connected to anything else other than the debug connector?
Regards,

Alistair

Re: USB Multilink Programmer fails to connect to the HCS08 target

TomE — Fri, 11 Apr 2014 06:18:29 GMT

(Vahid sent my mail asking for help, I wouldn't see this otherwise).

> I wonder why nobody has replied to this question.

Because nobody has a good answer for you most likely. Nobody else may be using that chip, or at least nobody else on these forums.

You've done most of what I would have triied, including different board and different computers (non-64-bit OS).

I'd suggest testing your pod on an evaluation board for that chip. If that works, monitor all the CPU and debug pins you can get to with an oscilloscope, and then compare all of them with the non-working boards.

You should also look for documents on what the debug protocol looks like so you can recognise what it is doing and work out why it is failing.

Tom

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Fri, 11 Apr 2014 06:23:14 GMT

Thank you very much alistairmuir for your reply. I do really appreciate it.

First, let me say that both BKGD and RESET pins have external pull-up resistors (10k and 15k, respectively).

The RESET pin has also a decoupling (pull-down) capacitor too (1000 pF).

Both of these pins are connected only to 6-pin BDM header and the IC itself, and nowhere else.

For the exact states of the BKGD and RESET pins, I should probably use a digital oscilloscope to monitor the toggling of their states. But for now:

Both the BKGD and RESET pins are "High" after POR (probably due to their pull-up resistors).

However, after I press the debug/erase button at CodeWarrior, the BKGD pin goes "Low", and the "power-cycle" dialog appears.

I turn off/on the MCU, as the "power-cycle" dialog suggests, and press OK; then the BKGD pin goes "High" again.

After a few seconds, the CodeWarrior gives the following error:

If I press "Retry", the same procedure happens again: the BKGD pin goes "Low", and the "power-cycle" error occurs!

During all these procedures, the CodeWarrior logger console writes the following logs:

Frequency change to ~0hz.

Error: Can not enter background mode .

Frequency change to ~0hz.
Frequency change to ~0hz.

I guess the problem could be related to the internal clock generator (ICG) of the MCU. How should I know if it's working properly?

I hope somebody helps me find the solution to this challenging problem. :smileysad:

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Fri, 11 Apr 2014 06:31:45 GMT

Thank you very much TomE for your response.

Actually, as I've mentioned in parts 7 and 8 above, I've tested this on different computers with different operating systems, and also on different boards. But the result is the same on all occasions.

Unfortunately, I cannot afford to buy an evaluation board; so I'll probably test the P&E USB multilink universal pod on an HCS08 based MCU to see if it works or not.

Besides, I've studied all the documents, application notes, etc. to see how the BDM works, but no success!

Thanks anyway for your time and reply. I do really appreciate your help. :smileyhappy:

Re: USB Multilink Programmer fails to connect to the HCS08 target

pavel_sadek — Fri, 11 Apr 2014 07:57:11 GMT

HI Vahid

I would bet your device is in reset state for whatever reason, check all reset connections (disconnect all except the BDM pod).

Pavel

Re: Re: USB Multilink Programmer fails to connect to the HCS08 target

Robinwithu — Fri, 11 Apr 2014 12:10:00 GMT

Hi Vahid,

Here is Application Note on " HCS08 SG Family Background Debug Mode Entry " please go throgh it really very useful.

Best Regards,

Robin

Re: Re: USB Multilink Programmer fails to connect to the HCS08 target

Robinwithu — Fri, 11 Apr 2014 12:21:31 GMT

Hi Vahid,

can you remove the Pull-up Register as well as Decoupling Capacitor from BKGD pin. I am also using 8 bit controller (MC9S08RN8), Codewarrior 10.5 and PE micro USB multilink Programmer and for me it's working fine .

Please see the Attachment.

Please make sure while programming only your programming pin should get power supply, & not other pins of micro controller.

Please let me know the result.

I hope this will help you.

Thanks and Kind Regards,

Robin

Re: USB Multilink Programmer fails to connect to the HCS08 target

karina_valencia — Fri, 11 Apr 2014 12:59:56 GMT

alistairmuir can you continue with the follow up>

Re: USB Multilink Programmer fails to connect to the HCS08 target

al_muir — Fri, 11 Apr 2014 23:18:09 GMT

Vahid,

Today I confirmed no issues in communicating with a standalone S08QE8 device using Codewarrior 10.5 and a P&E USB Multilink. I simply created a new bareboard project with just minimal initialization code, so it just has an endless for loop that sevices the COP. Using this I was able to connect and program flash. I am not sure what kind of project you have created but you may want to try duplicating this on your board. Make sure that the debug configuration for the project is selecting the right debug interface (P&E universal multilink in your case).

Are you able to confirm the part is truly soldered to the board correctly? can you "curve trace" the pins relative to Vss to check you see the ESD diodes and make sure there is a good connection to the board?

When you are trying to commmunicate with the debugger is RESET getting pulle all the way to Vdd and is the BKGD pin swinging feom Vss to Vdd?

Regards,

Alistair

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Sat, 12 Apr 2014 14:02:37 GMT

Thank you very much Robinwithu for your reply.

Actually, I've already studied this application note in details, as well as others, such as AN3335, AN2497, AN2295, AN3942, AN1744, AN3305, etc.

However, there's no success yet. :smileysad:

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Sat, 12 Apr 2014 14:13:13 GMT

Thank you Robinwithu for your time and support. I do really appreciate it.

As I mentioned before, the BKGD pin is directly connected to the IC. There's no decoupling capacitor on this pin.

I've tested the device, both with and without the 10k pull-up resistor on the BKGD pin; but the result is the same in both cases: it fails to connect to the device. (The only difference is that without the 10k pull-up resistor on BKGD pin, it seems to be more vulnerable to the ambient noise; because it reports random numbers in the "Log Console" of the Codewarrior IDE as the frequency.)

Just one question: What do you mean exactly by the following comment:

Please make sure while programming only your programming pin should get power supply, & not other pins of micro controller.

Do you mean that I should not supply power to the MCU at all !!?

I would be grateful if you add more details on this. Thanks.

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Sat, 12 Apr 2014 14:54:02 GMT

Thank you very much alistairmuir for your support. I do really appreciate it.

Today I confirmed no issues in communicating with a standalone S08QE8 device using Codewarrior 10.5 and a P&E USB Multilink. I simply created a new bareboard project with just minimal initialization code, so it just has an endless for loop that sevices the COP. Using this I was able to connect and program flash.

1. Would you please provide the schematic of this simple circuit? I just want to make sure about the connections. Thank you.

Are you able to confirm the part is truly soldered to the board correctly? can you "curve trace" the pins relative to Vss to check you see the ESD diodes and make sure there is a good connection to the board?

2. I'll try to monitor the pins using a simple "octopus" circuit and an oscilloscope as "curve tracer". But just one question: Do you mean I check the pins of the IC relative to the VSS, so as to check the internal diodes of the IC? Or you mean to test the external components on the board?

When you are trying to commmunicate with the debugger is RESET getting pulle all the way to Vdd and is the BKGD pin swinging feom Vss to Vdd?

As I mentioned before, when I press the debug button, the BKGD pin goes from VDD to VSS; but after the connection error and failure, it gets released to VDD again.

The RESET pin is pulled-up to VDD with a 15k resistor, so it remains high, except when the "power-cycle" dialog asks to manually turn off/on the MCU.

Thanks a billion for your support. I look forward to hearing more from you.

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Sat, 12 Apr 2014 16:54:53 GMT

Thanks PSA for your response.

This is actually the main "victim" of the problem, I think !!! But I don't know how to test it.

The following is a selection of the AN3762 application note, regarding the reset loop of the blank devices:

Note that the initial contents of RAM, while undefined,

power up with remarkable consistency on a single part, but its contents can vary from part to part. For this

reason, when blank parts are powered up, they will exhibit self-asserted reset loop times ranging from a

few microseconds to a few milliseconds.

If a self asserted reset loop time is lesser than the time a sync command sequence needs to complete, the

tool will not be able to synchronize with the part and establish background communications. Due to the

variable period nature of the self asserted reset loop, it is difficult to guarantee that blank parts will have

loop times long enough to allow a sync command sequence to complete.

Due to this self asserted reset loop, programming of blank parts must be done by putting the part into active

BDM mode (thus stopping user code from running) at power up and then allowing a tool to perform a sync

sequence to establish communications with the part.

Understanding the reset behavior of the part is a key to designing a production line programming system.

When the power is initially applied to an MCU, an internal mechanism sets a “POR” status bit and

continues with the reset sequence. The RESET pin is then driven low by the MCU for a number of clock

cycles (refer to particular family manuals for the number of cycles) and is then released. The RESET pin

is then sampled to see if it is high — if so, then the reset is classified as a POR and if the BKGD pin is low,

the device enters active BDM mode. If the RESET pin is low at this point, the reset is classified as a normal

reset, and the reset vector is fetched. Note that even if an external pullup resistor is not connected to the

reset pin, there’s an internal pullup resistor that will pull the reset pin high.

For a production line programming, the RESET pin must be passively pulled high and allowed to operate

without external driving influences. The critical timing area is when the RESET pin is sampled after the

internal circuitry releases the pin from being driven low. It’s best to let the passive circuitry pull this high;

coordinating an active drive to correspond to this timing would be very difficult.

To successfully program the SG family parts in a production line environment, all issues described above

need to be addressed:

The S08SG family uses an internally generated POR to qualify the state of the BKGD, not the rising edge of RESET.

Blank S08 parts will tend to exhibit self asserted reset loops of variable length on initial power up that may prevent a BDM tool from using the sync command sequence to gain control over the target MCU.

Allow the RESET line to be passively pulled up and not actively driven by the programming tool.

How to do so can be summarized in the following steps:

Hold the BKGD pin low.

Apply power to VDD while leaving the RESET to be passively pulled high allowing the part to enter active background mode. Do not allow any glitches or noise on the RESET line after this sequence.

Have the programming tool perform a sync command sequence to determine the target device’s BDM operating frequency.

Program the part.

Programming tools generally do accommodate a sync sequence type of startup. They will typically hold

the BKGD pin low, then ask the user to “cycle power to the target” and then go on to perform a sync

command sequence and establish a contact with the part.

How should I test to see if my device is in a "reset loop" or not? :smileyconfused:

Re: USB Multilink Programmer fails to connect to the HCS08 target

pavel_sadek — Sat, 12 Apr 2014 20:43:25 GMT

the power cycle right prior to BDM entry usually helps,

If isolation of RESET did not help, I have another idea - I would suggest you to test the application inside your Multilink device-

as it is universal tool, it can be configured for different targets.. ARM, CF, HCS08... There is firmware updater available on PE MICROs web, update your device for S08 BDM to be sure this is not the case... if the firmware is not the latest, CW will do another update then automatically, you just need to agree ...

Pavel

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Sun, 13 Apr 2014 15:31:19 GMT

Hi everybody,

Today, I monitored both the BKGD and RESET pins on a digital oscilloscope. The following pictures illustrate the scenario:

(Note: The BKGD in the first picture is initially low, because of its state from the last debug session. In normal start ups, it is initially high.)

I would be grateful if you help me find out what the problem is, and what I should do to solve it.

Thank you very much.

Re: USB Multilink Programmer fails to connect to the HCS08 target

AlanCollins — Mon, 14 Apr 2014 02:51:14 GMT

Hello Omid.

How many of these boards you have? Do you see the problem in all boards?

Have you ever download code successfully, and the later it failed? or else, it never worked OK ?

If you previously download code to the device, does the code was using Low Power modes?

Have you tried re-soldering the device into the board ?

The BDM is supposed to take control of the RESET line during the download and debugging processes. Have you tried a different BDM multilink, Is it possible it is damaged?

See below a reference design using this device: http://cache.freescale.com/files/rf_if/doc/user_guide/12311DHRM.pdf?fasp=1&WT_TYPE=Reference%20Manuals&WT_VENDOR=FREESCA…

Please refer to manufacturing chapter.

I got your emails through LinkedIn. My apologies, I was out of the office during last week. Thanks your patience and understanding.

I will follow up this problem until it gets resolved.

Re: USB Multilink Programmer fails to connect to the HCS08 target

Robinwithu — Mon, 14 Apr 2014 07:28:45 GMT

HI Vahid,

Please make sure while programming only your programming pin should get power supply, & not other pins of micro controller.

with this line i mean when you program controller than only your programming pin should get supply voltage not other port pins of micro controller.

Thanks and Regards,

Robin

Re: USB Multilink Programmer fails to connect to the HCS08 target

vahidelc86 — Mon, 14 Apr 2014 07:35:56 GMT

Thanks Robinwithu for your reply.

What is the reason that the other port pins of the device should not get power supply?

Do you mean that I should not even supply power to VDD pins of the MCU? (Because the power supply is actually connected only to VDD, V_ref, etc. and not the port pins.)

Regards,