Talk:Boundary scan

From Wikipedia, the free encyclopedia
Jump to: navigation, search


Untitled[edit]

  • Boundary scan is a method for testing interconnects (thin wire lines) on printed circuit boards

Please merge boundary scan info from JTAG page into here...[edit]

There's a lot more to be said about boundary scan ... and it shouldn't be getting said on the JTAG page. -- 69.226.238.251 (talk) 04:27, 28 November 2009 (UTC)

I disagree - boundary-scan IS JTAG, on-chip debug etc..is not. Remember the acronym 'Joint Test Action Group' - no mention of debuggers. The JTAG committee that devised this system in the 1980s did so with boundary-scan in mind - nothing else. Additional chip features that are acccessd by the JTAG port are not part of any perceived JTAG standard and should not form part of this generic article. In my view all such separate debug interfaces (ARM, COP, MIP, and what have you) should be listed under separate articles. 195.60.15.185 (talk) 14:53, 3 December 2009 (UTC)
This article is about boundary scan. So why do you argue that details of boundary scan should not be in this article, dedicated to that topic?? Color me puzzled. JTAG certainly hasn't been limited to boundary scan for quite a few years now ... but digital boundary scan is fairly JTAG-centric. I only suggested that content be moved here, as part of improving this (pretty weak) boundary scan article ... not that JTAG for some reason be presented as not addressing boundary scan.
The JTAG article is covering the big picture of JTAG, which certainly includes debug as well as boundary scan. The ARM11 example just fleshes out that part of the skeleton. It's unlike boundary scan for several reasons ... including the obvious one, that boundary scan already has a dedicated (but weak and needs-improvement) page in WP. --69.226.238.251 (talk) 01:06, 9 December 2009 (UTC)

I moved the details and cleaned some of the transitions. This page still needs work, it's messy; but at least you don't have to go to another page to see most of the WP information on the topic of this page. --69.226.238.251 (talk) 19:26, 15 December 2009 (UTC)

I argue on behalf of boundary-scan since you 69.226.238.251 have taken it upon yourself to redefine what it JTAG. Once again I remind you that it is structural test interface and exists to serve that purpose for many thousands of users. The fact that the port structure has been hijacked by IC vendors as a conduit to/from debugger cores does not mean that the important boundary-scan functionality be relegated to outside of this article. Your heavy handed editing will only serve to confuse the many. By all means duplicate the boundary-scan register based test functions definitions in the boundary-scan article but do not also remove them from JTAG.
Might I suggest that you transfer your recent edits to a new artice entitled 'JTAG Debug' or similar and refrain from imposing your definitions (honest as I'm sure your intentions are).
81.103.50.148 (talk) 22:34, 15 December 2009 (UTC)
Hmm, it appears rather that "you 81.103.50.148" have refused to address the issues I raised, and are rather the one who has taken upon yourself to cause problems by ignoring Wikipedia:Revert only when necessary and ignoring points raised:
  • "Once again I remind you" ... that the JTAG article still presented boundary scan as one of the reasons JTAG is significant, and included a concise summary along with the ref to this main article. Your assertion that I tried to "redefine" JTAG was thus false/dishonest. Thus a WP:REVEXP was effectively missing. Accordingly, I've undone your revert. See below about resolution.
  • "Once again I remind you" ... that for at least 3/4 of the time it's existed, JTAG hasn't been *ONLY* about boundary scan. It's kind of ludicrous to assert that "chip vendors" are hijacking a transport mechanism. They are after all the ones providing that transport! And they do that to support many things. It's theirs; no "hijacking" could be involved. Adding new instructions, TAPs, and applications is fully within scope of JTAG. (If you're obsessed with the "T-for-test" ... system debug is a harder testing problem than board integrity, as a rule. Millions more things to break in even a moderate size software stack.)
  • "Once again I remind you" ... that while I addressed every issue you raised, you ignored all the ones I raised. Then after ignoring my response (for over a week), you just reverted an edit the same day it was made, using a dishonest justification. Grmph.
Note that WP policy says that the responsibility for justifying inclusion of any content rests firmly with the editor seeking to include it (cut/paste from WP:TE). Given that you've not justified the need for all that boundary scan detail to sit specifically in the JTAG page, instead of in the boundary scan article where it's appropriately the entire focus. I specifically asked for such justification above... so, likewise from WP:TE: you ignore or refuse to answer good faith questions from other editors. You seem doubly in the wrong. Whether or not you intended to do so, you have sent me the message that your reason is akin to Wikipedia:I just don't like it, facts-be-damned.
Might I suggest ... that you first respond to the points I've raised, before starting an edit war. Rather than just ignoring them, work to build consensus by sticking to logical arguments. So far you haven't really tried, so there'd be no point in even considering any of the Wikipedia:Dispute resolution stuff. That could still be available later though.
In fact, please use your energy to improve the Boundary Scan article, instead of weakening the JTAG article or re-adding semi-commercial refs to that boundary-scan.co.uk site (yours?). If you work in boundary scan, that's the article you should be working to seriously improve.
--69.226.238.251 (talk) 08:47, 17 December 2009 (UTC)

What 'problems' exactly have I caused by adding valuable information back into the article ? You suggested this discussion from by adding a link on the JTAG page concerning the transfer of data. Since then two contributors have disagreed yet you continued nonetheless and without a fair discussion IMO.

You caused problems by reverting edits without giving a valid justification, by repeatedly refusing to answer questions, and ignoring points that disprove your assertions. Re "disagreed" ... I saw none with reasons any more substantive than "I just don't like it". That's not in the list of Good Reasons; trying to have a "fair discussion" against tissue-thin arguments is not possible, especially when the tissue-proponent doesn't provide responses with substance.
69.226.210.39 (talk) 02:21, 1 March 2010 (UTC)

Far from trying to vandalise it I have attempted to keep a balance in the JTAG article. Only you have cut information not I. Yes I am in the board test business and I am pleased to admit it. I am also keen to keep a significant portion of the JTAG article relating to the boundary-scan aspects which are not just 1149.1 but also 1149.4 ( I can contribute to this as you requested) and also 1149.6 (AC Extest) P1149.8 is still in discussion (working group stages) and I would happily add that too as and when it is ratified.

Re "keeping balance" ... I don't see that. No information was "cut". It was just moved into the article dedicated to that topic, with clear path markers leading to that location. The JTAG article presents the entire systems context for that technology ... including a clear reference to the article on your particular interest (boundary scan). That's balanced. Having any substantial discussion of boundary scan inside the JTAG article (your seeming preference) would have been imbalanced.
69.226.210.39 (talk) 02:21, 1 March 2010 (UTC)

As most of us in the test industry know, people who think JTAG think boundary-scan,therefore to major on the other uses for teh port is I believe misleading. In the design sphere then maybe people who think JTAG think debug/emulation or device programming. So you see BOTH scenarios are valid (are they not?) and hopefully in equal measure. Note also I have not sought to devalue your contributions by removing them and my use of the word hi-jack is merely a figure of speech (although quite possibly true and not in the least ludicrous)as the initial JTAG port has be re-used as an entry point for debugger tools- try not to be so sensitive.

The JTAG article says up front that there are multiple applications for JTAG, and that boundary scan is one of them. It's not in the least misleading to say that debugging is another one.
If you remove all systems developers from the equation, yes I can imagine you'd only be left with boundary scan test developers, who would then use JTAG only for boundary scan. "Misleading" would have resulted from discarding most of the product development team and then treating the rest as representative. I work with software (sometimes including Verilog or VHDL) and board developers, and we use JTAG for many other purposes beyond boundary scan. Boundary scan is usually easy to accomodate at the tail end of the process. That's not true for other JTAG applications.
Try not to be so sensitive, you say? I think that suit fits you better than me. I didn't remove any text -- just relocated details to a place where they're appropriately the entire focus. And I'm not sure how you could think reverting (without justification) a well-justified change, completely ignoring responses I've spent the time to write, and not answering my questions ... would not be "devaluing" behavior on your part Try running any relationship using those techniques, and watch the old crash'n'burn kick in right quickly. (No, on second thought ... don't try that unless you're a glutton for pain.)
69.226.210.39 (talk) 02:21, 1 March 2010 (UTC)

If you look back through the development of the JTAG article you can see how the JTAG article looked in 2004 several months BEFORE someone started a boundary-scan article.

'JTAG is an acronym for Joint Test Action Group and refers to the IEEE 1149.1 standard for Test Access Port and Boundary Scan. JTAG are specifications of a four-pin (plus power/ground) interface designed to test connections between chips. The interface is serial. The clock input is at the TCK pin. Configuration is performed by manipulating a state machine one bit at a time through a TMS pin. Then transferring one bit of data is transferred in and out per TCK clock pulse at the TDI and TDO pins, respectively. Different instruction modes can be loaded to read the chip ID, sample input pins, drive (or float) output pins, manipulate chip functions, or bypass (pipe TDI to TDO to logically shorten chains of multiple chips). The operating frequency varies depending on the chip, but it is typically 10-100MHz TCK (10-100ns per bit time). Note that storage (RAM, disc,etc.) is usually measured in binary thousands, i.e., 1,024 bits,K, Kb or Kilobits.' Retrieved from "http://en.wikipedia.org/wiki/Joint_Test_Action_Group"

In summary my point is that boundary-scan is one if not 'the' cornerstone of JTAG and should be represented in this article by more that a small paragraph. I personally happen to think that the ARM information, which you have inserted, is not so important. However I did not delete a single word of it irrespective of my opinion., which I happen to think is more in the spirit of Wikipedia.

The JTAG article does cover more of its Boundary scan aspects than just a small paragraph ... history, instructions, the BSR, and more. (But that's not the focus of the JTAG article ... whereas it *IS* the focus of the Boundary Scan article.
Into each article go the details specific to that article. I *still* do not have the vaguest understanding of why you think the JTAG article should have details about how to perform boundary scan testing, instead of the boundary scan article.
The ARM information is an example of how systems get structured around JTAG. FPGA and ASIC designs sometimes use the same approach (custom registers and instructions providing observability and controllability for stuff that's not accessible through boundary scan). I now drawn out that point a bit more explicitly. It's all the same piece of cloth: as a system technology, JTAG has multiple applications. But it's not very comprehensible without some kind of high level example showing how to fit the parts together.
69.226.210.39 (talk) 02:21, 1 March 2010 (UTC)

BTW FYI I have no connection to www.boundary-scan.co.uk, although I admit I found pages of it useful . On balance I agree it is possibly too commercial for wiki (along with thousands of other links).

Someone was, for a while, spending a lot of energy to make sure various commercial links kept re-appearing. That one was very active, and your IP is geographically near that site (within an hour or two by automobile). Agreed that WP would be better without such efforts.
69.226.210.39 (talk) 02:21, 1 March 2010 (UTC)

I hope that we can continue this discussion in a more consiliatory manner in future. 81.103.50.148 (talk) 20:02, 22 December 2009 (UTC)

Prof. James B. Angell at Stanford University proposed serial testing[citation needed] I'd say without a reference and time frame, this is of dubious value. Why mention him? WHat was his contribution over Goel, Turino, Konemann? I think if there is no reference, no paper, then I am moved to remove this. Jtagchair (talk) —Preceding undated comment added 19:40, 22 March 2010 (UTC).

analog boundary scan?[edit]

There seem to be IEEE 1149.4 docs relevant to mixed signal ICs and their use of boundary scan ... it's evidently no longer purely analog. Could someone who knows boundary scan add some information about that? Ditto the Analog BSDL (ABSDL) stuff. --69.226.238.251 (talk) 19:36, 19 December 2009 (UTC)

Additional topics: etymology, usage, software, photos.[edit]

A sentence or two about the etymology of "boundary scan" would be helpful. Particularly considering that this is an encyclopedia article. What is the boundary? What is scanned? Is the name as realistic now as when it was invented? My preference is to add the etymology in the History section and make the title "History and Etymology". If that is adopted, the section will fit well after the Contents rather than at the end of the article. It's a common way of structuring an article.

A few use cases would be interesting. Used primarily in industry? Would an EE undergraduate use it to troubleshoot a term project?

What software is available. Any open-source software?

At least one photo would be helpful. Some readers, me included, have never seen a boundary scan setup.

Regards, ... PeterEasthope (talk) 13:39, 7 January 2017 (UTC)