Hello David,

TM4C129 series has EPI controller for interfacing to a 256 Mbit SDRAM. We do not have a EMIF controller (but instead something called EPI).

Now selecting which TM4C129 device you would required would depend on what additional features like Ethernet. USB, Flash, SRAM, etc.

Regards

Amit

Hello Amit,

Has the EPI controller been tested with the SDRAM described in the attachment?

David Robinson

@ Mr. Robinson,

Mon ami - are not "pioneers" usually recognized by their arrow-strewn corpses - littering the pristine/unpaved trail?

Unless you (really) must have that exact SDRAM - is it not far safer (and less puncturing) to employ devices tried/proven by esteemed vendor?  (i.e. on their board - or at least upon their "approved/tried" list...)

David Robinson said:

Has the EPI controller been tested with the SDRAM described in the attachment?

 Hi David, from first post is unclear if you wish test ram for functionality or use to expand internal memory.

Selected ram at first glance can match bandwidth of TIVA 60MHz EPI channel but not sure it can work on, after parallel search due to site failure these are Infineon brand so provider is solid not as his leaky site!!!

 Also your application is unclear and absolutely no data come from you than usage of EMIF so suppose you can come from DSP or C2000 series and goal is to test pattern sensitivity or parameter, if this is true TIVA is not fast enough to exploit full bandwidth of chip to overstress it.

 Here how I checked who is your silicon provider but from my perspective who use IIS and M$ server is far from solid!!!

http://www.qimonda-licensing.com/
HTTP/1.1 404 
Connection: close
Date: Mon, 03 Nov 2014 11:21:54 GMT
Server: Microsoft-IIS/6.0
MicrosoftSharePointTeamServices: 12.0.0.6421
X-Powered-By: ASP.NET

http://www.qimonda-licensing.com/default.aspx
404 FILE NOT FOUND

http://www.qimonda-licensing.com/usageterms.html
HTTP/1.1 404 Connection: close Date: Mon, 03 Nov 2014 11:28:14 GMT Server: Microsoft-IIS/6.0 MicrosoftSharePointTeamServices: 12.0.0.6421 X-Powered-By: ASP.NET

http://www.qimonda-licensing.com/imprint.html
HTTP/1.1 404 Connection: close Date: Mon, 03 Nov 2014 11:28:55 GMT Server: Microsoft-IIS/6.0 MicrosoftSharePointTeamServices: 12.0.0.6421 X-Powered-By: ASP.NET


 From Digikey:

Qimonda

- Qimonda is the second largest global supplier of DRAM memory products. Following the carve out from Infineon Technologies AG on May 1, 2006, Qimonda went public at the New York Stock Exchange on August 9, 2006.  Qimonda has access to five 300mm manufacturing sites on three continents and operates five major R&D facilities, including its lead R&D center in Dresden, Germany. The company is a leading supplier of DRAM products to PC and server manufacturers and is increasingly focusing on products for graphics, mobile and consumer applications as well using its power saving trench technology. 

 My compliment to Infineon.. or ex Siemens to choose that poor tech!!

Hello,

The task is to use a processor to test an SDRAM device at an elevated temperature. The devices to test are as follows:

INFINEON HYB39S256160FE-7

INFINEON HYB39S256160FT-7

QUIMOND HYB39S256160FE-7 

QUIMOND HYB39S256160FT-7

 These devices will then be installed on PCBAs and used with an Atmel AT91RM9200 processor.

Thank you,

David Robinson

David Robinson said:

 These devices will then be installed on PCBAs and used with an Atmel AT91RM9200 processor.

 David, as you can figure Atmel is not a TI product, so answer to your question is not so simple at all nor can be done on a firm forum, maybe someone other know this processor, IMHO I bought 10 pcs some year ago but never I discovered how to use it and I abandoned Atmel due to lack of support.

 From your word I gotcha the purpose of your request, answer is more close to NO than yes, testing these ram require a lot of bandwidth and I think better way is thru an FPGA and support proocessor or multicore DSP to be sure have enough bandwidth to spread pattern and analyse them on parallel.

 TIVA on the counterpart has some Hi temp part but I don't know nothing about and I fear this low volume can be much better handled by just IPcore processor and specialized hardware (again on IPCORE) built around an sdram controller core to do just this task.

 I suggest you try Altera forum or multicore DSP on TI e2e, the first as best fit.

Hello David,

We have only tested Micron SDRAM parts MT48LC32M16A2TG (limited) with the EPI @ 60MHz. So if the new data sheets match up with the specification, you may be able to have the SDRAM tester with a TIVA Part.

Regards

Amit

Amit Ashara said:

We have only tested Micron SDRAM parts MT48LC32M16A2TG (limited) with the EPI @ 60MHz. So if the new data sheets match up with the specification, you may be able to have the SDRAM tester with a TIVA Part.

 Hi Amit, I think the Infineon RAM can fit better TIVA but from initial writing of David I inferred the desire to test ram than match with EPI, so as I wrote a test unit is more close to his need. I don't think EPI is enough to test stress and measure reliability of sdram chip as David wish to do. I definitively recommend FPGA and specialised IPCore but some TI multicore DSP can do quite same job but I don't figure how to measure parameter.

 Best fit require as usual more and more data from initial request as habit of this forum.

 After a full working day Qimonda site still is locked out.....

Hello Roberto,

I agree, the requirement of the Tiva as "Tester" is not very clear as has been the case in the past.

Regards

Amit

Agree w/Roberto's suggestion of Altera site - and iirc Xilinx too (perhaps in partnership w/small memory-based firm) had critical info. 

As you know - board layout - short/matched/direct trace implementation - is required for best throughput/reduced error rate. 

Should Amit's suggested SDRAM have found its way to some eval board - such would seem a, "good start" for your test/analysis.  Our findings/advice - most anything "real world" trumps data-sheet theory/claims...

Hello,

Texas Instruments has no ARM processors that you can recommend that can be connected to the SDRAM in question to perform writes and read to ensure the memory does not take fail at temperature. We need to add in an FPGA to make this happen.

I took a look at the board that this SDRAM is mounted on with the Atmel ARM processor and they did not need an FPGA.

When you communicate in a forum it is helpful if you write clearly and be professional. We should need to use our web based urban dictionary to find the meaning of

IIRC - If I Recall Correctly

David Robinson

As this reporter was "lucky enough" to be attacked - was use of hallowed "urban dictionary" really required to investigate Xilinx's potential contribution to your objective?  We think not.

And - "board layout - short/matched/direct trace implementation - required for best throughput/reduced error rate" is in violation of (your) judgment of that which qualifies as, "professional?"  May we (respectfully) disagree?

Might it be that an attack - targeted at those who freely/quickly respond with "real world" guidance - better qualifies as, "unprofessional?"  (surely qualifies as unjust/unkind/unfriendly...)

David,

Pinging other TI teams to see if they have any more input on your query.

Blake

David Robinson said:

Texas Instruments has no ARM processors that you can recommend that can be connected to the SDRAM in question to perform writes and read to ensure the memory does not take fail at temperature. We need to add in an FPGA to make this happen.

 David, if you need a simple GO - NO GO test then TIVA is suitable too..

 As I suggested to test stress and do reliable test measurement an FPGA is necessary to implement a test bed. If you just need test read write nor pattern sensitivity or other odd parameter deviation of memory reaponse then why not try use Atmel instead?  Why are you searching for a TI part to test a ram that fit an Atmel???

 Again your post was not clear of what your goal, I can suggest an OMAP series, that processor can stress test them but is not this an overkill? Try see Beagle Board series if can suit your need too, is ready and more simple to use.

If you need real professional help then follow guidelines on how to ask for and our answer can be tailored to best fit your request.

 Altera use forum like TI, xilinx also still use IIRC <<In Internet relay Chat>>, so please also update your "urban based dictionary" to what is useful too, otherwise what you find is what you wish... ;)

Blake,

As your up/pinging other TI teams - might you direct some (further & enhanced) attention to this long festering situation - too? 

http://e2e.ti.com/support/microcontrollers/tiva_arm/f/908/t/358842.aspx

Surely many more are impacted by this Exosite issue - seems strange that so singular a post (this very particular SDRAM issue) receives such, "team attention" - while "Stellaris Sai" seems (alone) in his quest to provide critically needed updates - to the many who've long sought help here - and have been, "more than patient."

Thanks & regards...

cb1,

Amit does a great job helping others and a main point of contact here for escalation. Looping in others when needed and when other teams might need to be involved to be brought in to a thread or have it moved to their forum. We do are best to help out thousands of questions coming in. You can always reach out should anything need further attention or have feedback and I'll do my best.

Amit and team are probably on top of the other issue and will post updates as they are available. We will see where things are on that so thanks for letting me know and reminding team here.

Blake

Blake,

Indeed - Amit does a great, unending job - I've acknowledged such many times - here & in private conversation.

However - he's not appeared (yet) to my best knowledge - the only TI representative has been "Stellaris Sai" (2 postings) and his promised return has been left, "vague." 

Many have requested help - such has not (effectively) been forthcoming.

If there's time for a "team" to investigate a (pardon) apparently most minor SDRAM issue (raised by just one poster) might such "team effort" be more profitably directed to the many - who have so long sought help - and remain impacted?  (somehow - seems a superior use of a rare & valuable, "team resource"... thanks...)

David - The team has reviewed and nothing more to add at this time to existing posts from Amit and others.

cb1 - Thanks again for escalating the other issue and don't be surprised to see an update soon on that thread.

Blake,

You've always gone, "Above/beyond" in all of our dealings.  Thank you.  Your focus & pooling of resources is sure to improve the results of those who've (long) hoped for comfort/resolution.

Should state that my group is neither an Exosite nor TM4C129 user - I've simply tried to be of some value to a forum which I still believe to be, "best in class" - have long supported - and with your efforts (now) may be seen as far more caring, capable & responsive...

Thanks again Blake for all you do, (and will continue to do) - very much appreciated...

Blake Ethridge said:

David - The team has reviewed and nothing more to add at this time to existing posts from Amit and others.

 Blake, thank for all support, I think sometimes Amit can be tired a lot and also other sometimes lost focus on request, we/I got offensive languages and an heavy attack of nonsense request, this case just has a vague request and appeared as a request to match an SDRAM to TIVA. Carefully weighting word appeared that goal was different, too many times request are vague and so difficult to address in a professional manner. I am refining my magic bowl.. but as you see is not fully working, when is ready I promise you some lottery number ;)

 Can I ask see also my post about library and cleaning of TIVAWARE? Can we have more examples more clean and documented commented in a good way? The only answer I got is the usual CB1 Humor..

 Thank a lot every case.