[Elphel-support] testbench 353

[Marc Reichenbach]

Hi,

thank you for your answer. I tried it with icarus and it works but
takes a very long time (~15 min) and it is, in my opion, not very
debug-friendly.

Long story short, if anyone is interested: The different behaivoral in
both simulators (modelsim and icarus) comes from the different way,
how registers are initialized. I found no option in modelsim to set a
default initialize berhaivor - so I tried other simulators. In
Synopsys VCS there exist such a option and it works fine. I got a
simulation speedup of approx. 10 - 15 times so the simulation now
takes only ~ 1 minute. If anybody wants to now all details (for
example how to fix some files to work with VCS) don't hesitate to
contact me.

Thanks again for the camera,

Marc

On Tue, Oct 25, 2011 at 8:29 AM, Andrey Filippov
<support-list at support.elphel.com> wrote:
> Marc,
>
> Thank you for the nice words about our cameras. We use Icarus Verilog plus
> GTK Wave ourselves  so I would recommend you to try the x353_sim.sh script
> that runs them, I do not have experience with modelsim myself.
>
> Currently the sensor data is provided by the sensor.dat text file, so the
> zero output from the sensor is likely caused by either this file missing or
> some difference in syntax  between Modelsim and Icarus.
>
> Andrey
>
>
> On Sat, Oct 22, 2011 at 4:25 AM, Marc Reichenbach
> <marc.reichenbach at informatik.uni-erlangen.de> wrote:
>>
>> Hi,
>>
>> after long time, I have some time to work with this great camera. But
>> I have the following problem:
>>
>> I want to simulate the x353 with modelsim and this works so far - but
>> no data is send from the sensor (in the simulation). I captured the
>> Output of the sensor and this was always 0. Is this the normal
>> behaivor? What the delivered testbench are should doing? It is not
>> documented :( Are there are a testbench which will stimulate the
>> sensor to send the "ramp" oder "random" image? The SPI Input to the
>> sensor changes a littel with some Weak-Hi Weak-Low - but no corect
>> data, I think.
>>
>> Thanks a lot,
>>
>> Marc Reichenbach
>>
>> _______________________________________________
>> Support-list mailing list
>> Support-list at support.elphel.com
>> http://support.elphel.com/mailman/listinfo/support-list_support.elphel.com
>
>