State of MAX7000S I/O pins before programming

In the MAX7000S datasheet, it says the I/O pins are tri-stated and (weakly)
pulled-up, to avoid board conflicts, during programming.  What should the
state of the I/O pins be _before_ programming?  I have a sample EPM7128S
device here which appears to be pulling them low - but it also fails the
blank test, so I think it must have been programmed.  How should a brand-new
part behave?

Hi Andrew Holme,

> In the MAX7000S datasheet, it says the I/O pins are tri-stated and
> (weakly)
> pulled-up, to avoid board conflicts, during programming.  What should the
> state of the I/O pins be _before_ programming?  I have a sample EPM7128S
> device here which appears to be pulling them low - but it also fails the
> blank test, so I think it must have been programmed.  How should a
> brand-new part behave?

There's some internal bit in the MAX7000A and later that says that the
device contains a valid configuration. If this bit is not set, the device
remains in tristate. In the MAX7000S, this bit is not present, but if the
device is blank, it should still stay in slightly-pulled-up-tristate.

Only if the device is not blank (this does _not_ mean that it has a _valid_
configuration), it will go out of tristate and do something.

Blank that sample and see what happens ;-)

Hope this helps.

Best regards,


Ben

Ben Twijnstra wrote:
> Hi Andrew Holme,
>
> > In the MAX7000S datasheet, it says the I/O pins are tri-stated and
> > (weakly)
> > pulled-up, to avoid board conflicts, during programming.  What
should the
> > state of the I/O pins be _before_ programming?  I have a sample
EPM7128S
> > device here which appears to be pulling them low - but it also
fails the
> > blank test, so I think it must have been programmed.  How should a
> > brand-new part behave?
>
> There's some internal bit in the MAX7000A and later that says that
the
> device contains a valid configuration. If this bit is not set, the
device
> remains in tristate. In the MAX7000S, this bit is not present, but if
the
> device is blank, it should still stay in slightly-pulled-up-tristate.
>
> Only if the device is not blank (this does _not_ mean that it has a
_valid_
> configuration), it will go out of tristate and do something.

Thanks.

> Blank that sample and see what happens ;-)

How do I do that?  Is there an option in Quartus for this?  I couldn't
see anything ....

Hi Andrew,

>> Blank that sample and see what happens ;-)
> 
> How do I do that?  Is there an option in Quartus for this?  I couldn't
> see anything ....

Wups... I don't have an actual EPM7128S device lying around here, but from
the command line try "quartus_pgm --operation=R". If you need more help,
type "quartus_pgm --help=operation".

Best regards,



Ben

Ben Twijnstra wrote:
> Hi Andrew,
>
>>> Blank that sample and see what happens ;-)
>>
>> How do I do that?  Is there an option in Quartus for this?  I
>> couldn't see anything ....
>
> Wups... I don't have an actual EPM7128S device lying around here, but
> from the command line try "quartus_pgm --operation=R". If you need
> more help, type "quartus_pgm --help=operation".

This is what I'm getting:

C:\>quartus_pgm -c "ByteBlasterMV [LPT1]" -m JTAG --operation=R
Error: Programming option string R is illegal. Refer to --help for legal
programming option formats.

The help says:

<options> must be one of the following combinations:
P, V, B, S, E, L, PVBL, PBL, PVB, PVL, PL, VL
where
P = Program         V = Verify
B = Blank-check     L = Lock/Security Bit
S = Skip/Bypass*    E = Examine*
(* Cannot be used in combination with other options)

There doesn't appear to be an R option.

I'm using Quartus II Programmer Version 4.1 Build 181 06/29/2004 SJ Web
Edition

Does this version have an erase option?

Hi Andrew Holme,

> Ben Twijnstra wrote:
>> Hi Andrew,
>>
>>>> Blank that sample and see what happens ;-)
>>>
>>> How do I do that?  Is there an option in Quartus for this?  I
>>> couldn't see anything ....
>>
>> Wups... I don't have an actual EPM7128S device lying around here, but
>> from the command line try "quartus_pgm --operation=R". If you need
>> more help, type "quartus_pgm --help=operation".
> 
> This is what I'm getting:
> 
> C:\>quartus_pgm -c "ByteBlasterMV [LPT1]" -m JTAG --operation=R
> Error: Programming option string R is illegal. Refer to --help for legal
> programming option formats.
> 
> The help says:
> 
> <options> must be one of the following combinations:
> P, V, B, S, E, L, PVBL, PBL, PVB, PVL, PL, VL
> where
> P = Program         V = Verify
> B = Blank-check     L = Lock/Security Bit
> S = Skip/Bypass*    E = Examine*
> (* Cannot be used in combination with other options)
> 
> There doesn't appear to be an R option.
> 
> I'm using Quartus II Programmer Version 4.1 Build 181 06/29/2004 SJ Web
> Edition
> 
> Does this version have an erase option?

Hmmm... I'm using a beta of version 5.0 (only available under NDA), which
does have the option available.

Will try 4.2 tomorrow - that's the oldest workable version I have laying
around (I also have a copy of 3.0 and the initial 1996 release, but I don't
count those as current).

Best regards,



Ben

You can always search the answer at the www.altera.com.

For MAX� 7000S devices, they are shipped blank but have specific bits
set to tri-state all their I/O pins. If the part is erased, it is
blank, but the factory default tri-state bit settings are also erased
and therefore pins will no longer be tri-stated. To tri-state the
blank part again, you need to have the blank POF (programming object
file) from a factory shipped blank device that has never been
programmed. You can get the blank POF by examining a factory shipped
MAX 7000S device using the MAX+PLUS II software (Quartus II does not
support examine for MAX 7000S); the examined POF will be blank but
will have the tri-state bit settings. Programming a MAX 7000S device
with this POF will tri-state the I/Os and blank the design.

MAX 7000AE, MAX 3000A, and MAX 7000B devices differ from MAX 7000S -
if the part is blank or erased it will automatically tri-stated the
all the I/O pins and does not need tri-state bit settings. If you
erase these devices, they will tri-state all I/Os. This is because MAX
7000AE, MAX 3000A, and MAX 7000B remain tri-state until a specific bit
is set during ISP called the ISP_DONE bit. It is the last bit to be
programmed when programming the device - if that bit is not set, the
device remains tri-stated.

All MAX3000A, 7000S, 7000A, 7000B, and MAX II devices are shipped
blank from the factory. When a blank device is powered up, all of the
I/O pins are tri-stated.

If you erase a MAX 3000A, 7000A, 7000B, or MAX II device it behaves
the same as a blank new factory shipped device. The I/O pins of a
blank device are tri-stated when powered. However, this is not true
for MAX 7000S. A new factory shipped MAX 7000S device will tri-state
when powered, but if this device is erased in some manner (Jam, JBC,
SVF, etc), then the I/O pins will drive out undetermined values. This
is because a new factory shipped MAX 7000S part is blank except for
some OE control bits which are set to tri-state the I/O pins. When
erased, those OE bits are cleared and the I/O pins drive out. The only
way to get the I/O pins to tri-state again is if the customer examined
and saved the .pof from a new factory shipped MAX 7000S device.
Program a MAX 7000S with that .pof and it will behave the same as the
new factory shipped part, that is it will be blank with tri-stated
I/Os.


Here is the link

http://answers.altera.com/altera/resultDisplay.do?page=http%3A%2F%2Fwww.altera.com%2Fsupport%2Fkdb%2F2003%2F12%2Frd12082003_9642.html&result=0&responseid=45b50f58f81c41c5%3A576e70%3A10338fea6ab%3A6f47&groupid=1&contextid=5179%3A1971.2113%2C1625.1759%2C1827.1970&clusterName=DefaultCluster&doctype=1000&excerpt=Program+a+MAX+7000S+with+that+.pof+and+it+will+behave+the+same+as+the+new+factory+shipped+
art%2C+that+is+it+will+be+blank+with+tri-stated+I%2FOs.#Goto1971

http://answers.altera.com/altera/resultDisplay.do?page=http%3A%2F%2Fwww.altera.com%2Fsupport%2Fkdb%2F2004%2F09%2Frd09092004_5239.html&result=3&responseid=45b50f58f81c41c5%3A576e70%3A10338fea6ab%3A6f4d&groupid=1&contextid=5514%3A1785.1980%2C1044.1152&clusterName=DefaultCluster&doctype=1000&excerpt=MAX+7000AE%2C+MAX+3000A%2C+and+MAX+7000B+devices+differ+from+MAX+7000S+-+if+the+part+is+blank+or+erase
+it+will+automatically+tri-stated+the+all+the+I%2FO+pins+and+does+not+need+tri-state+bit+settings.#Goto1785

Ben Twijnstra <btwijnstra@gmail.com> wrote in message news:<BIV6e.62768$Sc7.16377@amsnews05.chello.com>...
> Hi Andrew Holme,
> 
> > Ben Twijnstra wrote:
> >> Hi Andrew,
> >>
> >>>> Blank that sample and see what happens ;-)
> >>>
> >>> How do I do that?  Is there an option in Quartus for this?  I
> >>> couldn't see anything ....
> >>
> >> Wups... I don't have an actual EPM7128S device lying around here, but
> >> from the command line try "quartus_pgm --operation=R". If you need
> >> more help, type "quartus_pgm --help=operation".
> > 
> > This is what I'm getting:
> > 
> > C:\>quartus_pgm -c "ByteBlasterMV [LPT1]" -m JTAG --operation=R
> > Error: Programming option string R is illegal. Refer to --help for legal
> > programming option formats.
> > 
> > The help says:
> > 
> > <options> must be one of the following combinations:
> > P, V, B, S, E, L, PVBL, PBL, PVB, PVL, PL, VL
> > where
> > P = Program         V = Verify
> > B = Blank-check     L = Lock/Security Bit
> > S = Skip/Bypass*    E = Examine*
> > (* Cannot be used in combination with other options)
> > 
> > There doesn't appear to be an R option.
> > 
> > I'm using Quartus II Programmer Version 4.1 Build 181 06/29/2004 SJ Web
> > Edition
> > 
> > Does this version have an erase option?
> 
> Hmmm... I'm using a beta of version 5.0 (only available under NDA), which
> does have the option available.
> 
> Will try 4.2 tomorrow - that's the oldest workable version I have laying
> around (I also have a copy of 3.0 and the initial 1996 release, but I don't
> count those as current).
> 
> Best regards,
> 
> 
> 
> Ben

Kar wrote:
> You can always search the answer at the www.altera.com.
>

[snip: text of articles]

Thanks for the lead.

For the benefit of others, those expired links are:
http://www.altera.com/support/kdb/2004/09/rd09092004_5239.html
http://www.altera.com/support/kdb/2003/12/rd12082003_9642.html

Perhaps the most useful link for me, since I don't have the old
MAX+PLUS II software, and can't use "Examine" to create a POF file from
a fresh factory-shipped device, is this link which explains how to
blank a MAX7000S device using the JAM player:

http://www.altera.com/support/kdb/1999/09/rd09131999_640.html

Andrew Holme wrote:
> Kar wrote:
> > You can always search the answer at the www.altera.com.
> >
>
> [snip: text of articles]
>
> Thanks for the lead.
>
> For the benefit of others, those expired links are:
> http://www.altera.com/support/kdb/2004/09/rd09092004_5239.html
> http://www.altera.com/support/kdb/2003/12/rd12082003_9642.html
>
> Perhaps the most useful link for me, since I don't have the old
> MAX+PLUS II software, and can't use "Examine" to create a POF file
from
> a fresh factory-shipped device, is this link which explains how to
> blank a MAX7000S device using the JAM player:
>
> http://www.altera.com/support/kdb/1999/09/rd09131999_640.html

After re-eading the above, I'm not sure if the JAM player "erase"
method is an "erase" or a "blank" as defined in the other articles i.e.
does it tri-state the I/O pins?

Andrew Holme wrote:
> Andrew Holme wrote:
> > Kar wrote:
> > > You can always search the answer at the www.altera.com.
> > >
> >
> > [snip: text of articles]
> >
> > Thanks for the lead.
> >
> > For the benefit of others, those expired links are:
> > http://www.altera.com/support/kdb/2004/09/rd09092004_5239.html
> > http://www.altera.com/support/kdb/2003/12/rd12082003_9642.html
> >
> > Perhaps the most useful link for me, since I don't have the old
> > MAX+PLUS II software, and can't use "Examine" to create a POF file
> from
> > a fresh factory-shipped device, is this link which explains how to
> > blank a MAX7000S device using the JAM player:
> >
> > http://www.altera.com/support/kdb/1999/09/rd09131999_640.html
>
> After re-eading the above, I'm not sure if the JAM player "erase"
> method is an "erase" or a "blank" as defined in the other articles
i.e.
> does it tri-state the I/O pins?

OK, having gone round the loop one last time (this blank/erased
business is very confusing!) I believe that reading a blank POF is
(quote) "The only way to get the I/O pins to tri-state again" - so I'm
hosed.