[Xenomai-help] National Instruments PCI-6711

[Xenomai-help] National Instruments PCI-6711

[Ettore Pedretti]

Dear All,

Thanks for all the help getting the PCI-6711 running under xenomai. I
have a question about the functionalities implemented by the analogy
driver,

With the NI linux driver we can programme  some waveforms in and let
them run without further intervention. We do not need to change the
waveforms in real time. Is it possible to do this with the analogy
driver?

I guess you should use analogy commands and an asynchronous
acquisition command. All the examples I have seen on comedi/analogy
put the acquisition in a while loop. It seems as if you need to fill
the buffer each time.

What is the closest functionality I can get to programming the card
and letting it run without further intervention?

All the best
Ettore

Re: [Xenomai-help] National Instruments PCI-6711

[Alexis Berlemont]

Hi,

Ettore Pedretti wrote:
> Dear All,
> 
> Thanks for all the help getting the PCI-6711 running under xenomai. I
> have a question about the functionalities implemented by the analogy
> driver,
> 
> With the NI linux driver we can programme  some waveforms in and let
> them run without further intervention. We do not need to change the
> waveforms in real time. Is it possible to do this with the analogy
> driver?

I remember having read in the DAQ STC manual that such a feature was
available for the digital output subdevice. So, it may not be
surprising that it is also available for the analog output device.

> 
> I guess you should use analogy commands and an asynchronous
> acquisition command. All the examples I have seen on comedi/analogy
> put the acquisition in a while loop. It seems as if you need to fill
> the buffer each time.

Precisely, even if the hardware is able to work like you expect, the
command API does not support it. The command structure
(inherited from Comedi) defines via the stop_src and stop_arg fields
the length of the asynchronous acquisition and all the data are to be
going through the application. 

On the other side, that is just the theory. I would not be surprised
to find a Comedi driver which proposes the behaviour you ask thanks to
a driver-specific hidden cheat flag somewhere in the structure.

> 
> What is the closest functionality I can get to programming the card
> and letting it run without further intervention?
> 

I have no answer so far. That would need an extension of the framework.

> All the best
> Ettore
> 
> _______________________________________________
> Xenomai-help mailing list
> Xenomai-help@domain.hid
> https://mail.gna.org/listinfo/xenomai-help

-- 
Alexis.

Re: [Xenomai-help] National Instruments PCI-6711

[Ettore Pedretti]

Hi,

> On the other side, that is just the theory. I would not be surprised
> to find a Comedi driver which proposes the behaviour you ask thanks to
> a driver-specific hidden cheat flag somewhere in the structure.

How can I find out the "hidden cheat" and how can I exploit it?

>> What is the closest functionality I can get to programming the card
>> and letting it run without further intervention?
>>
>
> I have no answer so far. That would need an extension of the framework.

How difficult would that be? Is it a major change to the framework? It
seems to me that this is a nice option. You would not use any of your
bus bandwidth and could just change the waveform when you need it.

Ettore

>> All the best
>> Ettore
>>
>> _______________________________________________
>> Xenomai-help mailing list
>> Xenomai-help@domain.hid
>> https://mail.gna.org/listinfo/xenomai-help
>
> --
> Alexis.
>



-- 
Ettore Pedretti, SUPA
School of Physics and Astronomy
University of St Andrews, North Haugh,
St Andrews, Fife, KY16 9SS, Scotland
Ph: +44-1334-461666, Fax: +44-1334-463104
The University of St Andrews is a charity registered in Scotland : No SC013532

Re: [Xenomai-help] National Instruments PCI-6711

[Alexis Berlemont]

Hi,

Ettore Pedretti wrote:
> Hi,
> 
> > On the other side, that is just the theory. I would not be surprised
> > to find a Comedi driver which proposes the behaviour you ask thanks to
> > a driver-specific hidden cheat flag somewhere in the structure.
> 
> How can I find out the "hidden cheat" and how can I exploit it?
> 

That was just a supposition. I have not seen such a thing in the
Analogy NI driver (inherited from the Comedi one). Have you had a look
at the Comedi list ?

I just read again your first mail. You said the NI Linux driver
provides the requested option. Do you have any source code ? I
remember just having found a binary driver.

> >> What is the closest functionality I can get to programming the card
> >> and letting it run without further intervention?
> >>
> >
> > I have no answer so far. That would need an extension of the framework.
> 
> How difficult would that be? Is it a major change to the framework? It
> seems to me that this is a nice option. You would not use any of your
> bus bandwidth and could just change the waveform when you need it.
> 

I agree this feature would be nice but I cannot tell you that I will
add some tricky code in some corner just to let this option work. The
problem with the command system is that it expects a continuous data
transfer until the end of the acquisition. 

Maybe, a way to quickly implement it would be to use an instruction
(it may be possible, I cannot ensure you); but to do so I need the
code which properly programs the DAQ-STC. 

> Ettore
> 
> >> All the best
> >> Ettore
> >>
> >> _______________________________________________
> >> Xenomai-help mailing list
> >> Xenomai-help@domain.hid
> >> https://mail.gna.org/listinfo/xenomai-help
> >
> > --
> > Alexis.
> >
> 
> 
> 
> -- 
> Ettore Pedretti, SUPA
> School of Physics and Astronomy
> University of St Andrews, North Haugh,
> St Andrews, Fife, KY16 9SS, Scotland
> Ph: +44-1334-461666, Fax: +44-1334-463104
> The University of St Andrews is a charity registered in Scotland : No SC013532

-- 
Alexis.

Re: [Xenomai-help] National Instruments PCI-6711

[Ettore Pedretti]

Hi,

Sorry for the delay in answering. I left the observatory and went on
holiday for a while.

The standard NI driver can be downloaded from:
http://sine.ni.com/nips/cds/view/p/lang/en/nid/14480

We found this on the NI web site. I do not know if it is useful:

http://sine.ni.com/nips/cds/view/p/lang/en/nid/11737

Maybe it is the standard example that you already used to develop
comedi/analogy drivers. Please let me know if it is of any use for
you.

Best regards
Ettore

On 5 June 2010 22:40, Alexis Berlemont <alexis.berlemont@domain.hid> wrote:
> Hi,
>
> Ettore Pedretti wrote:
>> Hi,
>>
>> > On the other side, that is just the theory. I would not be surprised
>> > to find a Comedi driver which proposes the behaviour you ask thanks to
>> > a driver-specific hidden cheat flag somewhere in the structure.
>>
>> How can I find out the "hidden cheat" and how can I exploit it?
>>
>
> That was just a supposition. I have not seen such a thing in the
> Analogy NI driver (inherited from the Comedi one). Have you had a look
> at the Comedi list ?
>
> I just read again your first mail. You said the NI Linux driver
> provides the requested option. Do you have any source code ? I
> remember just having found a binary driver.
>
>> >> What is the closest functionality I can get to programming the card
>> >> and letting it run without further intervention?
>> >>
>> >
>> > I have no answer so far. That would need an extension of the framework.
>>
>> How difficult would that be? Is it a major change to the framework? It
>> seems to me that this is a nice option. You would not use any of your
>> bus bandwidth and could just change the waveform when you need it.
>>
>
> I agree this feature would be nice but I cannot tell you that I will
> add some tricky code in some corner just to let this option work. The
> problem with the command system is that it expects a continuous data
> transfer until the end of the acquisition.
>
> Maybe, a way to quickly implement it would be to use an instruction
> (it may be possible, I cannot ensure you); but to do so I need the
> code which properly programs the DAQ-STC.
>
>> Ettore
>>
>> >> All the best
>> >> Ettore
>> >>
>> >> _______________________________________________
>> >> Xenomai-help mailing list
>> >> Xenomai-help@domain.hid
>> >> https://mail.gna.org/listinfo/xenomai-help
>> >
>> > --
>> > Alexis.
>> >
>>
>>
>>
>> --
>> Ettore Pedretti, SUPA
>> School of Physics and Astronomy
>> University of St Andrews, North Haugh,
>> St Andrews, Fife, KY16 9SS, Scotland
>> Ph: +44-1334-461666, Fax: +44-1334-463104
>> The University of St Andrews is a charity registered in Scotland : No SC013532
>
> --
> Alexis.
>



-- 
Ettore Pedretti, SUPA
School of Physics and Astronomy
University of St Andrews, North Haugh,
St Andrews, Fife, KY16 9SS, Scotland
Ph: +44-1334-461666, Fax: +44-1334-463104
The University of St Andrews is a charity registered in Scotland : No SC013532