topic Re: XEP100 EEEPROM in S12 / MagniV Microcontrollers

XEP100 EEEPROM

johnthompson001 — Wed, 07 Mar 2018 09:31:29 GMT

I notice in the EEEPROM_Quick_Start.mcp project in the main.c code there is the line..

#define PARTITION_D_FLASH 0x20 /* M48H mask only */

As I am using a different mask (1N35H) could someone tell me how the PARTITION should be modified.

We are using the DEMOAX9S12 evaluation board with the XEP100 device and using the EEEPROM_Quick_Start.mcp project from the downloaded AN3743SW.zip files. I am able to verify the EEEPROM memory functionality by stepping through the code using Codewarrior when connected to the Demo board but when trying to repeat the process on our design of system board with a XEP100 then the FDIV error is displayed whilst executing the step through. I've confirmed the

'#define FCLK_DIV 0x2F // Flash clock divider for 50MHz crystal'

in the CLOCK DIVIDE SECTION is correct and the board/device functions correctly with the rest of the project. This is the second system board with an XEP100 device showing the same problem so I'm confident I've ruled out a device with faulty memory.

Any ideas?

Re: XEP100 EEEPROM

danielmartynek — Wed, 14 Mar 2018 09:29:49 GMT

Hi,

The difference between 1N35H and 5M48H is only in the fab site not in the silicon.

You can refer to errata of these masksets, they are identical.

https://www.nxp.com/docs/en/errata/MSE9S12XEP100_1N35H.htm

https://www.nxp.com/docs/en/errata/MSE9S12XEP100_5M48H.htm

Do you really use 50MHz crystal?

The specification of max f_osc is 16MHz with loop-controlled and 40MHz with full-swing Pierce oscillator.

Regards,

Daniel

Re: XEP100 EEEPROM

RadekS — Wed, 14 Mar 2018 11:21:44 GMT

Hi John,

The AN3743 is now quite old application note.

It seems that “M48H mask only” note is rather related to the preproduction masksets (like xM22E, not produced anymore) insted of 1N35H. The 1N35H is newest version produced in TSMC11 fab. It exactly refers to 5M48H maskset from TSMC FAB3 (dual source for avoiding delivery issues).

As correctly mentioned by Daniel, the maximum oscillator frequency for S12XE is specified as 40MHz (Full Swing Pierce oscillator connection, see Table A-4. Operating Conditions in RM).

What oscillator you use on your design of system board?

See Table 24-9. FDIV vs OSCCLK Frequency in RM for obtain correct FCLKDIV register value.

Please look at attached short S12XE EEPROM description and example codes.

I hope it helps you

Best regards

Radek

Re: XEP100 EEEPROM

johnthompson001 — Wed, 14 Mar 2018 12:16:37 GMT

Hi Radek,

Many thanks for your reply.

I’m using a SMD crystal oscillator fed into the Extal input on the XEP100. Interestingly, I’ve installed a 50MHz crystal oscillator on the DemoAX9S12 board and tried the same EEEPROM Quick Start and it worked OK… but it is not the same oscillator! I’ll try a 20MHz oscillator now I’m armed with your information.

Regards

John Thompson

Testometric

materials testing machines

Unit 1, Lincoln Business Park

Lincoln Close, Rochdale

Lancashire, England OL11 1NR

Telephone: +44 (0) 1706 654039

Facsimile: +44 (0) 1706 646089

Website: www.testometric.co.uk<blocked::http://www.testometric.co.uk/>;

Reg. No. 1112862 England

P Please consider the environment before printing this email.

Re: XEP100 EEEPROM

johnthompson001 — Wed, 14 Mar 2018 12:17:49 GMT

Hi Radek,

Did you mean to attach something?

Regards

John Thompson

Re: XEP100 EEEPROM

johnthompson001 — Wed, 14 Mar 2018 12:23:13 GMT

Many thanks Daniel.

Re: XEP100 EEEPROM

RadekS — Wed, 14 Mar 2018 12:47:46 GMT

Hi John,

I think that MCU can run with 50MHz external CMOS oscillator, but it is still beyond operating characteristics. So, no guarantee.

Please be aware, that crystal oscillator may be connected in two oscillator connections (Loop Control/Full Swing).

The Full Swing connection should contain external resistor (typ. 1MOhm) in parallel to the crystal. The PE7 (XCLKS) pin should be pulled-down during reset for Full Swing connection and External CMOS oscillator as clock sources.

If the code works on evaluation board and not on your system board, the issue may be also somewhere else. Could you please share your schematic for short review?

If it cannot be shared publicly, please create a service ticket for that.

Best regards

Radek

Re: XEP100 EEEPROM

johnthompson001 — Wed, 14 Mar 2018 16:29:52 GMT

Hi Radek,

I’ve tried the EEEPROM Quick Start with a 20MHz oscillator (remembering to alter the FCLK_DIV to 0x13) with no positive change.

XCLKS is pulled down on Reset.

I’ve also tried running the two projects in the attachment (Emulated EEPROM & Initial_data_in_Emulated EEPROM) but I’m not sure what I’m looking for to verify the EEPROM memory is working.

I can send you the schematic if you’ve time to have a quick look although our XEP100 based board works when I load other projects… just not the EEPROM.

John

Re: XEP100 EEEPROM

johnthompson001 — Wed, 14 Mar 2018 17:38:53 GMT

Hi Radek,

I feel I’m missing something obvious here.

If I connect to the Demo board (50MHz clock) and launch the EEEPROM Quick Start project I see the following memory:

When I connect to our board (20MHz clock) and launch the project on our board I see the following:

I’ve verified the FCLK_DIVIDE is set correctly.

I don’t see any errors whilst stepping through the code in Debug.

Sorry if I’m missing something obvious.

John

Re: XEP100 EEEPROM

johnthompson001 — Thu, 15 Mar 2018 10:11:56 GMT

Hi Radek,

There seems to be a problem with programming the Flash memory on the XEP100 device so please ignore the last message to you.

I have another board with a XEP100 device that I will use. This can be flash programmed via the BDM but still does not execute the EEEPROM Quick Start project successfully.

Regards

John Thompson

Testometric

materials testing machines

Unit 1, Lincoln Business Park

Lincoln Close, Rochdale

Lancashire, England OL11 1NR

Telephone: +44 (0) 1706 654039

Facsimile: +44 (0) 1706 646089

Website: www.testometric.co.uk<blocked::http://www.testometric.co.uk/>;

Reg. No. 1112862 England

P Please consider the environment before printing this email.

Re: XEP100 EEEPROM

danielmartynek — Fri, 16 Mar 2018 11:32:16 GMT

Hi John,

Could you please elaborate on what exactly doesn't work? Do you see any error flags? Which Flash operation sets the flags?

Just to be sure, the external oscillator is 1.8V CMOS compatible, and you can measure correct bus clock frequency (max 50MHz) on ECLK pin, right?

Regards,

Daniel

Re: XEP100 EEEPROM

RadekS — Mon, 19 Mar 2018 12:10:39 GMT

Hi John,

Thank you for more information.

Just idea: Do you work with the same project at both boards (50MHz/20MHz)? Or you have different projects with almost the same code?

The default project does not refresh EEEPROM content when we stop MCU. This must be enabled in menu-HC12MultilinkCyclonePro-Debugging Memory Map-global eeprom-refresh memory when halting. Please check whether this option is checked in your project.

Note: The Partition command must not be interrupted (e.g. by any reset), otherwise the D-Flash may be just partially formatted. If you want to check whether D-Flash was correctly formatted, please look at address 0x13F000. There should be two words with sector header. The 0xFFFFFFFF value means, that Partition command was probably interrupted. In that case, you must do mass erase of MCU prior next Partition command execution.

The mass erase is also necessary when you want to change any of partition parameters.

For mass erase go to menu-HC12MultilinkCyclonePro-Unsecure-select appropriate FCLKDIV.

The Full Partition command do not need mass erase, but it could be run only in special mode (=with debugger interface connection).

I hope it helps you.

Best regards

Radek

Re: XEP100 EEEPROM

johnthompson001 — Tue, 20 Mar 2018 18:04:34 GMT

Hi Daniel,

The crystal oscillator is HCMOS/TTL compatible reduced to a 2.2 v amplitude signal and confirmed at 50MHz.

Regards

John Thompson

Testometric

materials testing machines

Unit 1, Lincoln Business Park

Lincoln Close, Rochdale

Lancashire, England OL11 1NR

Telephone: +44 (0) 1706 654039

Facsimile: +44 (0) 1706 646089

Website: www.testometric.co.uk<blocked::http://www.testometric.co.uk/>;

Reg. No. 1112862 England

P Please consider the environment before printing this email.

Re: XEP100 EEEPROM

johnthompson001 — Tue, 20 Mar 2018 18:27:58 GMT

Hi Radek,

Thank you for your help so far.

My project are not the same as EEPROM Quick Start but I can flash both devices with my project and they function correctly. I have captured the address from 0x13FE00 and I can see that the loaded code is not the same on my project as when I start the EEPROM Quick Start with the DEMOAX9S12 board.

The address 0x13F000 is not as you suggested when loaded into my project.

When I launch the project using Codewarrior using Debug the mass erase is performed.

I’m sorry I can’t find the Debugging Memory Map in the P&E CyclonePro settings.

John

Re: XEP100 EEEPROM

johnthompson001 — Fri, 23 Mar 2018 16:39:45 GMT

Hi Radek,

I feel to have exhausted all options here.

I’m comparing behaviour on the DEMOAX9S12 XEP100 board with our project using an XEP100 device using the EEPROM Quick Start project via Debug mode.

The crystal oscillators are the same value 50MHz and XCLKS=0.

The DEMO board works well and programs the DFLASH… see below.

Our system board with the XEP100 device using the EEPROM Quick Start shows a blank DFLASH… see below.

PS. I can flash our system board with another project and it functions as expected but this project does not use DFLASH.

If you have any more ideas what is missing?

Best Regards

John

Re: XEP100 EEEPROM

RadekS — Mon, 26 Mar 2018 07:23:10 GMT

Hi John,

When you click on Debug button, the debugger will erase Flash and EEPROM by sectors – this is not mass erase. Please, do not believe to the information from image003.png warning.

Please use unsecure command if you wish erasing Partitioning.

I would like to apologize. My mistake, that last D-Flash sector starts at address 0x107F00 and not at 0x13F000.

Please make screenshot of your P&E CyclonePro settings.

Mine is:

Radek

Re: XEP100 EEEPROM

RadekS — Mon, 26 Mar 2018 07:42:04 GMT

Hi John,

When we load EEPROM Quick Start project into MCU and MCU is not partitioned yet, the DFLASH should be blank as on image002.png.

When we start debugging = stepping through code, the D-Flash become be partitioned and after few steps we will see that DFLASH is Programmed as on image001.png.

It seems, that your system board was not partitioned yet.

So, currently I do not see any issue in described behavior.

Radek

Re: XEP100 EEEPROM

johnthompson001 — Mon, 26 Mar 2018 08:59:13 GMT

Hi Radek,

Thanks for coming back to me.

Is this the information you require below.

Regards

John

Re: XEP100 EEEPROM

johnthompson001 — Mon, 26 Mar 2018 16:45:28 GMT

Hi Radeks,

I’ve stepped through the program using the Debug function and the memory is not changing. There are no errors displayed by the Debugger.

I feel there is an issue with the partitioning but I can’t see what the issue is.

Regards

John

Re: XEP100 EEEPROM

RadekS — Tue, 27 Mar 2018 08:19:43 GMT

Hi John,

please unsecure your MCU and load AN3743 code into MCU.
Please write 0x25 into MMCCTL1 register at address 0x0013 for enable EEE IFR visible in memory map.
After that, please go to Debugging Memory Map - Modify "global eeprom" range 0x100000~0x13FFFF and check "refresh memory when halting".
After that, please go to memory "0x120000'G". There you should see 0x0074 0x0074 0x0001 0x00001 values (DFPART, ERPART and their backups) after partition command execution(/***********(3)PERFORM FULL PARTITION OF EEE RAM***********/)
at address 0x107F00'G you should see 0xFACF 0x0000 value as last sector header. In such case, the D-Flash was correctly formated.
When you will step over code, you should see EEE values change at address 0x13FF00'G or 0x0F00'L(in such case, please check "refresh memory when halting" also in "eeprom" memory range 0xC00~0xFFF).

See attached memory map for more details about local and global addresses.

I hope it helps you.

Radek