Very poor Xilinx experience

I'm hoping someone at Xilinx reads this, because I can't find any
other way to get through to anyone to help me.

Short version: I've bought an SP605 board, it looks as though it's
broken - there's no video output from the built-in-self-test program.
I have to complete a webcase to get an RMA, and the <expletive-
deleted>-ing useless Xilinx mailing-list software won't let me send
any information to my webcase consultant.

The details of the problem, along with screenshots, are at
http://forums.xilinx.com/t5/Xilinx-Boards-and-Kits/sp605-won-t-boot-into-the-reference-design/td-p/224717.

Having had to re-set my email address from my personal address to my
corporate address (why should I need to do this to get an RMA ?), it
seems I can no longer log into the webcase system, so I can't append
any pertinent information to the case, and my case consultant seems to
be trying to get me to set DIP switches / set header pins according to
the "SP605 Evaluation Kit" hardware setup guide, which is directly
contradicting the advice in my "Spartan-6 FPGA Embedded Kit" hardware
setup guide pamphlet that came with the kit.

I have repeatedly (4 times now) tried to send mail asking for
clarification on this, and every damned time, the mailing software
rejects my email as being badly formed. My case consultant probably
thinks it's all sorted out, and there's no way for me to tell him I'm
just getting frustrated with the lack of response! There's no Xilinx
forum for "help, the support system is broken", and I got no traction
in the 'Xilinx boards and kits' forum (see the above link).

So, guys, what the hell is going on ?  I hate to resort to using a
public forum to detail individual problems, but I can't think of
anywhere else to go! If anyone at Xilinx reads this, the webcase
number is 915564, and I would dearly like to get some progress on it.
The board is just sitting here, and I don't want to change the DIP
switches and possibly break it...

Simon.

On Tue, 03 Apr 2012 14:43:05 -0700, Simon wrote:

> I'm hoping someone at Xilinx reads this, because I can't find any other
> way to get through to anyone to help me.
> 
> Short version: I've bought an SP605 board, it looks as though it's
> broken - there's no video output from the built-in-self-test program. I
> have to complete a webcase to get an RMA, and the <expletive-
> deleted>-ing useless Xilinx mailing-list software won't let me send any
> information to my webcase consultant.
> 
> The details of the problem, along with screenshots, are at
> http://forums.xilinx.com/t5/Xilinx-Boards-and-Kits/sp605-won-t-boot-
into-the-reference-design/td-p/224717.
> 
> Having had to re-set my email address from my personal address to my
> corporate address (why should I need to do this to get an RMA ?), it
> seems I can no longer log into the webcase system, so I can't append any
> pertinent information to the case, and my case consultant seems to be
> trying to get me to set DIP switches / set header pins according to the
> "SP605 Evaluation Kit" hardware setup guide, which is directly
> contradicting the advice in my "Spartan-6 FPGA Embedded Kit" hardware
> setup guide pamphlet that came with the kit.
> 
> I have repeatedly (4 times now) tried to send mail asking for
> clarification on this, and every damned time, the mailing software
> rejects my email as being badly formed. My case consultant probably
> thinks it's all sorted out, and there's no way for me to tell him I'm
> just getting frustrated with the lack of response! There's no Xilinx
> forum for "help, the support system is broken", and I got no traction in
> the 'Xilinx boards and kits' forum (see the above link).
> 
> So, guys, what the hell is going on ?  I hate to resort to using a
> public forum to detail individual problems, but I can't think of
> anywhere else to go! If anyone at Xilinx reads this, the webcase number
> is 915564, and I would dearly like to get some progress on it. The board
> is just sitting here, and I don't want to change the DIP switches and
> possibly break it...
> 
> Simon.

Perhaps Xilinx wants you to design in Altera parts.

-- 
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com

On Apr 3, 5:34=A0pm, Tim Wescott <t...@seemywebsite.please> wrote:
> On Tue, 03 Apr 2012 14:43:05 -0700, Simon wrote:
> > I'm hoping someone at Xilinx reads this, because I can't find any other
> > way to get through to anyone to help me.
>
> > Short version: I've bought an SP605 board, it looks as though it's
> > broken - there's no video output from the built-in-self-test program. I
> > have to complete a webcase to get an RMA, and the <expletive-
> > deleted>-ing useless Xilinx mailing-list software won't let me send any
> > information to my webcase consultant.
>
> > The details of the problem, along with screenshots, are at
> >http://forums.xilinx.com/t5/Xilinx-Boards-and-Kits/sp605-won-t-boot-
>
> into-the-reference-design/td-p/224717.
>
>
>
>
>
>
>
>
>
>
>
> > Having had to re-set my email address from my personal address to my
> > corporate address (why should I need to do this to get an RMA ?), it
> > seems I can no longer log into the webcase system, so I can't append an=
y
> > pertinent information to the case, and my case consultant seems to be
> > trying to get me to set DIP switches / set header pins according to the
> > "SP605 Evaluation Kit" hardware setup guide, which is directly
> > contradicting the advice in my "Spartan-6 FPGA Embedded Kit" hardware
> > setup guide pamphlet that came with the kit.
>
> > I have repeatedly (4 times now) tried to send mail asking for
> > clarification on this, and every damned time, the mailing software
> > rejects my email as being badly formed. My case consultant probably
> > thinks it's all sorted out, and there's no way for me to tell him I'm
> > just getting frustrated with the lack of response! There's no Xilinx
> > forum for "help, the support system is broken", and I got no traction i=
n
> > the 'Xilinx boards and kits' forum (see the above link).
>
> > So, guys, what the hell is going on ? =A0I hate to resort to using a
> > public forum to detail individual problems, but I can't think of
> > anywhere else to go! If anyone at Xilinx reads this, the webcase number
> > is 915564, and I would dearly like to get some progress on it. The boar=
d
> > is just sitting here, and I don't want to change the DIP switches and
> > possibly break it...
>
> > Simon.
>
> Perhaps Xilinx wants you to design in Altera parts.
>
> --
> Tim Wescott
> Control system and signal processing consultingwww.wescottdesign.com

It certainly seems that way. To Xilinx, I'm small potatoes, not even
on the radar, but to a hobbyist like me, dropping $2k on a board, add-
on modules, and PC to host the Xilinx software (I only had Macs, and
parallels doesn't work with the fpga programming cable under Linux)
only happens every two to three years. It's seriously disappointing to
then get what appears to be broken hardware and encounter an
impenetrable wall around the company...

Simon.

On Apr 3, 5:47=A0pm, Simon <goo...@gornall.net> wrote:
> On Apr 3, 5:34=A0pm, Tim Wescott <t...@seemywebsite.please> wrote:
>
>
>
>
>
> > On Tue, 03 Apr 2012 14:43:05 -0700, Simon wrote:
> > > I'm hoping someone at Xilinx reads this, because I can't find any oth=
er
> > > way to get through to anyone to help me.
>
> > > Short version: I've bought an SP605 board, it looks as though it's
> > > broken - there's no video output from the built-in-self-test program.=
 I
> > > have to complete a webcase to get an RMA, and the <expletive-
> > > deleted>-ing useless Xilinx mailing-list software won't let me send a=
ny
> > > information to my webcase consultant.
>
> > > The details of the problem, along with screenshots, are at
> > >http://forums.xilinx.com/t5/Xilinx-Boards-and-Kits/sp605-won-t-boot-
>
> > into-the-reference-design/td-p/224717.
>
> > > Having had to re-set my email address from my personal address to my
> > > corporate address (why should I need to do this to get an RMA ?), it
> > > seems I can no longer log into the webcase system, so I can't append =
any
> > > pertinent information to the case, and my case consultant seems to be
> > > trying to get me to set DIP switches / set header pins according to t=
he
> > > "SP605 Evaluation Kit" hardware setup guide, which is directly
> > > contradicting the advice in my "Spartan-6 FPGA Embedded Kit" hardware
> > > setup guide pamphlet that came with the kit.
>
> > > I have repeatedly (4 times now) tried to send mail asking for
> > > clarification on this, and every damned time, the mailing software
> > > rejects my email as being badly formed. My case consultant probably
> > > thinks it's all sorted out, and there's no way for me to tell him I'm
> > > just getting frustrated with the lack of response! There's no Xilinx
> > > forum for "help, the support system is broken", and I got no traction=
 in
> > > the 'Xilinx boards and kits' forum (see the above link).
>
> > > So, guys, what the hell is going on ? =A0I hate to resort to using a
> > > public forum to detail individual problems, but I can't think of
> > > anywhere else to go! If anyone at Xilinx reads this, the webcase numb=
er
> > > is 915564, and I would dearly like to get some progress on it. The bo=
ard
> > > is just sitting here, and I don't want to change the DIP switches and
> > > possibly break it...
>
> > > Simon.
>
> > Perhaps Xilinx wants you to design in Altera parts.
>
> > --
> > Tim Wescott
> > Control system and signal processing consultingwww.wescottdesign.com
>
> It certainly seems that way. To Xilinx, I'm small potatoes, not even
> on the radar, but to a hobbyist like me, dropping $2k on a board, add-
> on modules, and PC to host the Xilinx software (I only had Macs, and
> parallels doesn't work with the fpga programming cable under Linux)
> only happens every two to three years. It's seriously disappointing to
> then get what appears to be broken hardware and encounter an
> impenetrable wall around the company...
>
> Simon.- Hide quoted text -
>
> - Show quoted text -

Support is a two-way street and resolution comes a lot faster with
good communication and using the right channels.  The Xilinx forums
are a user community driven forums and while myself and other Xilinx
employees do read and respond to posts it is not an official support
channel.   The same goes for comp.arch.fpga.

You did use an official support channel by opening a WebCase on 4/1 at
12:11am and recieved a response when a US support engineer arrived in
the morning and had time to look at your initial comments and to route
the case to an appropriate support engineer who responded to you
within 12 hours of opening the case.

Some of the confusion is due to a lack of specifics from you.  Your
comments in the case notes include generic references to the "SP605
evaluation kit", "Spartan-6 fpga embedded kit" and "hardware and
demonstration setup guide".  The SP605 is sold in a several different
variations including:

EK-S6-SP605-G - Base Evaluation Kit
DK-S6-EMBD-G - Embedded Evaluation Kit
DK-S6-CONN-G - Connectivity Kit

Each of these have different demonstration/reference designs and setup
instructions. Providing the exact kit and doc number of the material
that you are using will eliminate confusion and allow the support
engineer to be able to resolve your problem faster.   Since you are
using an official support channel and the case is still open you
should continue with that path to resolve your problem.

BTW: Changing a DIP switch would not harm a board.

Regards,

Ed McGettigan
--
Xilinx Inc

On Apr 4, 9:16=A0am, Ed McGettigan <ed.mcgetti...@xilinx.com> wrote:

> Support is a two-way street and resolution comes a lot faster with
> good communication and using the right channels. =A0The Xilinx forums
> are a user community driven forums and while myself and other Xilinx
> employees do read and respond to posts it is not an official support
> channel. =A0 The same goes for comp.arch.fpga.

Indeed it does. I would have *loved* to have actually continued to use
my web-case to talk to Xilinx's support-engineer and resolved the
issue. BUT I CAN'T. I don't know how to make this any clearer. There
is no communication channel open to me other than

  - (a) spamming a Xilinx forum that doesn't cater for this type of
issue (I read the informative types of things to post in each of the
Xilinx forums, the closest seemed to be general technical support, but
this isn't a technical issue),  or

  - (b) Going "public" in an open forum.

In case it's still not clear:

 - I *cannot* log into the web-case system. My username/password no
longer works
 - I *cannot* send email replies to my case-officer / xilinx support -
the mailing system refuses to accept my response, and gives no
direction as to how to format the email correctly. From memory the
last email back from it said something like 'CASE_ID: xxxxxxx is
insufficient information to resolve the object'.

> You did use an official support channel by opening a WebCase on 4/1 at
> 12:11am and recieved a response when a US support engineer arrived in
> the morning and had time to look at your initial comments and to route
> the case to an appropriate support engineer who responded to you
> within 12 hours of opening the case.
>
> Some of the confusion is due to a lack of specifics from you. =A0Your
> comments in the case notes include generic references to the "SP605
> evaluation kit", "Spartan-6 fpga embedded kit" and "hardware and
> demonstration setup guide". =A0The SP605 is sold in a several different
> variations including:
>
> EK-S6-SP605-G - Base Evaluation Kit
> DK-S6-EMBD-G - Embedded Evaluation Kit
> DK-S6-CONN-G - Connectivity Kit
>
> Each of these have different demonstration/reference designs and setup
> instructions. Providing the exact kit and doc number of the material
> that you are using will eliminate confusion and allow the support
> engineer to be able to resolve your problem faster. =A0 Since you are
> using an official support channel and the case is still open you
> should continue with that path to resolve your problem.

*how* ? Exactly ?

Your comments are also misleading at best. I was *quoting* the titles
at the top of the two documents in order to distinguish them. Further,
I don't believe there was *any* confusion over the board I had bought,
since the web-case engineer said (again, from memory) "I cannot find
the document you are referring to on the embedded-kit page, but I
found this other document on the evaluation kit page, why don't you
try setting the DIP switches / jumpers like it says in this other
document I found".

To further clarify matters, I had sent a snapshot of the front of the
document I had in front of me (the one that your support engineer
couldn't find) as well as a snapshot of the DIP-switch settings
instructions within that guide.

More misleading characterization: you mention the 12-hour response
from the engineer (and you know what, I was impressed by that) but you
fail to mention the subsequent 3-day impenetrable silence.

> BTW: Changing a DIP switch would not harm a board.

[sarcasm]
Right. Never. Of course. I've *never* come across a board where a DIP
switch or header is responsible for setting a voltage, say. And
there's *never* any possibility of shorting out anything by putting
jumpers across pin headers.
[/sarcasm]

If you had read the two documents that I was disputing the difference
between, you'd have seen that there are directly-contradicting
instructions for both DIP switches *and* pin-headers. It's not exactly
unknown to have to set *groups* of headers just-so in order to get a
board to work, and having to use *either* this-set-of-headers *or*
that-set-of-headers, with any-mixture-of-this-set-of-headers-and-that-
set-of-headers being potentially dangerous to the board. In any event,
I don't think it's an unreasonable question to ask! Xilinx' own FMC
XM-105 board has a set of voltage headers that control another set of
headers, for example.

When the document that ships with your kit is saying 'do X' and the
support engineer can't find that document, but finds another one
that's vaguely related that says 'don't do X, do Y', I think it's
perfectly reasonable to question that support-engineer's advice (my
last response was along the lines of 'are you sure' ? and I *think*
that one made it through the user-hostile email manager).

---
Addendum: I've just been 'phoned by a Xilinx support manager. In
contrast to Ed's passive-aggressive tone above, the support manager
listened to the problem, realized that the email system wasn't working
in this case, and has told me an engineer will *call* me today to try
and sort things out. Hopefully this will all have a happy ending, now.

So, Ed, it seems that the only way I had to resolve the situation was
to do exactly what I did. Xilinx' engineer was waiting for an email
response that could never arrive, without me raising the issue where
someone would see it, I'd still be stuck where I was at the start of
the week. The manager explained that you've recently moved to
requiring a corporate email account for webcast support, and me
changing my email to fulfill that requirement after initially filing
the webcase was the probable root cause of the email problems. I still
don't have webcase access, but I've sent an email asking why. Perhaps
if that gets restored, things will go smoother too.

BTW: I've worked in support of highly-technical products before, hell
I owned the company. The best way to resolve something is *not* to
piss off the customer by trying to shift the blame onto them when your
own support can't find one of your own documents.

Simon.

Ed McGettigan <ed.mcgettigan@xilinx.com> wrote:

(snip)
> BTW: Changing a DIP switch would not harm a board.

It would certainly be poor design if it did, but if it switched
some pins to outputs, and they were also driven from off the board,
that could harm them (or the other board components).

But that would not be usual for a system design.

I normally don't worry that dip switches will harm hardware.

-- glen

On Wed, 04 Apr 2012 19:46:37 +0000, glen herrmannsfeldt wrote:

> Ed McGettigan <ed.mcgettigan@xilinx.com> wrote:
> 
> (snip)
>> BTW: Changing a DIP switch would not harm a board.
> 
> It would certainly be poor design if it did

The few Xilinx boards that I've gotten have been good, but in my 
experience the quality of eval boards is all over the map.  Some are 
very, very good, and some are very, very bad.  I'm quite cynical about 
this, which leads me to believe that the attitude in Marketing and 
Accounting at the companies where the designs are consistently good is 
"if we do well, next year we will have more sales and make more $$$", and 
at the companies where designs are spotty or consistently bad it is "this 
eval board stuff is a bunch of hoo-ha and an expense.  Every time we ship 
one we lose $; let's minimize the loss".

So "it would be a poor design if it did", and "so we can assume that it 
doesn't" are phrases that do not -- generally -- belong together in the 
same sentence.

-- 
My liberal friends think I'm a conservative kook.
My conservative friends think I'm a liberal kook.
Why am I not happy that they have found common ground?

Tim Wescott, Communications, Control, Circuits & Software
http://www.wescottdesign.com

Tim Wescott <tim@seemywebsite.com> wrote:

(snip regarding jumper settings and the cause of board failure)

>> It would certainly be poor design if it did

> The few Xilinx boards that I've gotten have been good, but in my 
> experience the quality of eval boards is all over the map.  Some are 
> very, very good, and some are very, very bad.  I'm quite cynical about 
> this, which leads me to believe that the attitude in Marketing and 
> Accounting at the companies where the designs are consistently good is 
> "if we do well, next year we will have more sales and make more $$$", and 
> at the companies where designs are spotty or consistently bad it is "this 
> eval board stuff is a bunch of hoo-ha and an expense.  Every time we ship 
> one we lose $; let's minimize the loss".

I haven't tried enough boards to say.

> So "it would be a poor design if it did", and "so we can assume that it 
> doesn't" are phrases that do not -- generally -- belong together in the 
> same sentence.

Well, I would make this statement more generally than just FPGA
evaluation boards, but for boards in general. Now, there are some
the use DIP switches or jumpers to select supply voltages for 
processors, and in that case one could damage a board by setting
them wrong. Usually, though, you would know that was the purpose
for those switches. 

-- glen

On Apr 4, 8:03=A0pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
> Tim Wescott <t...@seemywebsite.com> wrote:
>
> (snip regarding jumper settings and the cause of board failure)
>
> >> It would certainly be poor design if it did
> > The few Xilinx boards that I've gotten have been good, but in my
> > experience the quality of eval boards is all over the map. =A0Some are
> > very, very good, and some are very, very bad. =A0I'm quite cynical abou=
t
> > this, which leads me to believe that the attitude in Marketing and
> > Accounting at the companies where the designs are consistently good is
> > "if we do well, next year we will have more sales and make more $$$", a=
nd
> > at the companies where designs are spotty or consistently bad it is "th=
is
> > eval board stuff is a bunch of hoo-ha and an expense. =A0Every time we =
ship
> > one we lose $; let's minimize the loss".
>
> I haven't tried enough boards to say.
>
> > So "it would be a poor design if it did", and "so we can assume that it
> > doesn't" are phrases that do not -- generally -- belong together in the
> > same sentence.
>
> Well, I would make this statement more generally than just FPGA
> evaluation boards, but for boards in general. Now, there are some
> the use DIP switches or jumpers to select supply voltages for
> processors, and in that case one could damage a board by setting
> them wrong. Usually, though, you would know that was the purpose
> for those switches.
>
> -- glen

Well, in this case neither hardware setup guide explains what the
jumpers do, the physical document I have that came with the kit (the
"Spartan-6 FPGA Embedded Kit" hardware setup guide) just says "ensure
that Jumper J60 is connected as shown, J60 should have jumper on 1-2"
in (as well as a load of other jumpers / DIP switch settings, equally
without explanation.

However, in the "SP605 Evaluation kit" hardware setup guide which was
what my webcase support engineer was suggesting I use, it says "The
following headers should not have any Jumpers installed: J45, J47, J9,
J58, J9, J13, J10, J60, J49, and J48." (along with different DIP
switch settings).

Note the conflict for J60. I didn't check what the as-shipped settings
for the others were.

I've just tried re-formatting the flash card and re-copying the
software over (which worked for some other people) but it didn't help
me - now I just see a blank screen, and the ethernet doesn't respond
to ping/HTTP either (that used to work, at least). I've put another
request for clarification into the web-case, and we'll see how it
goes.

Simon

On Apr 4, 2:46=A0pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
> Ed McGettigan <ed.mcgetti...@xilinx.com> wrote:
>
> (snip)
>
> > BTW: Changing a DIP switch would not harm a board.
>

hhehe, that's what the South Korea folks and the Japanese are praying
now

Kim Jung Il will flip the switch soon sometime this month, from his
tomb