R.I.P. Compact FieldPoint

[ThiCop]

Hello Mads,

As long as they are aware of it and the issue is known to them it is good for me.

I honestly don't know what information I can publically disclose concerning the dual ethernet support.

I don't want to get in problems for disclosing non-public information, so I am afraid that I might have to currenly keep to what I can find in public documents: 

http://www.ni.com/pdf/embedded/NI_System_on_Module_Flyer.pdf

This Flyer only officially mentions GigE support, but doesn't seem to state explicitly if it is dual GigE.

This morning I was discussing this with a colleague and this might have brought up some other possible paths.

Our CVS System for GigE Vision (http://sine.ni.com/nips/cds/view/p/lang/nl/nid/211839) has more than 2 ethernet ports and does have a dedicated RS-485/422/232 serial port.

In the Digital IO section there also are 2 Bidirectional RS-422 Channels (http://www.ni.com/pdf/manuals/374217a.pdf).

Also it has a user-programmable FPGA that might be used to accomplish some tasks.

However, I don't know if it is worth further research since it is using a different form factor.

[Timmar]

Found this whilst trying to find out why LV2014SP1 couldn't see my cFP's

fp 2014 drivers

[Timmar]

We are moving to to cRIO 9064/5/6 family of 4 slot chassis. Surpasses cFP in every way but footprint.

[Mads]

Yes, the driver for the cFP is out, that buys us some time.

But we are moving on... (the cFP-2220 is one of the models that are continued longer than most, but it is still going out of production by 2018 as far as I remember...and with no more support in future versions of LabVIEW it gets less interesting even faster) . We currently use cRIO-9030 as a replacement when the footprint and power usage is not a problem. It has dual networking with identical functioanlity on both, which is a must have for us.

However, the footprint and power usage *is* a problem for us in many cases, and right now the only (!) NI-product able to solve that and the networking requirement is the sbRIO-9651 (System on Module). The downside to that though is that there are no compact carrier boards available for it with dual networking implemented...So we have to get that developed for us first.

[Timmar]

We looked into the SB-RIO as well but were vary wary of its EMC compliance issues.

NI didn't actually say it was a problem but removed the CE tick approval recommended that you put it in a metal box (Shield) and purchase the heatsink tophat for it.

Reading between the lines it gets hot and the RF screams like a banshee.

RoboRio is cute but only has 1 com port.

We looked into the SOM, the built in DAq is handy.  ordering by the 100's was a bit of a turn off. and needing to design a base bord (with PSU) and backup battery for RTC made the cFP look really good.

[Spex]

Hi Timmar,

I just want to chime in on the topics 'read between the lines,' and dispell a couple myths.  

The recommendations made for a heat spreader and enclosure and your subsequent concerns about heat and shielding are valid, but I'd like to add context.  

The components used in Single-Board RIO are shared with CompactRIO controllers.  The recommendations for a heat spreader and metal enclosure are essentially what you get off the shelf when buying a CompactRIO controller.  We recommend these solutions for applications that need to acheive similar environmental capabilities as CompactRIO (-40 to 70 ambient, no fans/passive convection cooling, etc.).  Every application will exercise the processor and FPGA within the Single-Board RIO differently, and our recommendations are generally based on what NI considers worst case scenario, where an application exercises the hardware at its full capability.  Many applications based on Single-Board RIO do not require additional thermal design (depending on actual utilization and environmental requirements), but the heat spreader in available to make advanced thermal design easier.

Regarding the assumption that the RF emissions 'scream like a banshee,' that just isn't the case.  Board level products are rarely certified for CE/emissions because the testing doesn't make sense.  Having CE on a board level product doesn't prevent (or even reduce) the end product (+RMC and enclosure) from requiring its own dedicated CE testing.  Our internal emmisions pre-scans for many of our board level products indicates that they could pass, or be very near passing levels, without any special enclosure or design.  They wouldn't be able to pass the ESD tests or conducted immunity tests that an enclosure would help resolve.  

Clearly you see the value of complete product design (with thermal design and certifications), and those are a couple of the many reasons we believe the CompactRIO controllers have more value than the board-level products.  Many companies do not have the ability to implement and validate the mechanical design surrounding a board-level product and meet the quality and reliabiltiy standards that NI already puts into our packaged systems.  

Regards,

[Timmar]

Spex,

Thanks for your respnse, that is the most information I have been able to recive about this.

When I first enquired, an NI applications enginneer made a specific point that the product was not EMC approved and specificaly refered me to the shielding design recommendations in the application note. and that it would require compliance testing.

It recommends small cable apertures and was most certainly metal, and not plastic.

Once you put anything in a box, shielded or otherwise, you lose convection cooling. 

The black body form of the heat spreader encourages heat to spread and radiate.

I get that it loses its ESD rating, but that is a seperate compliance matter.

Can you publish the CI/RI/CE/RE results to back your statement.

Our product is already in a closed stainless steel box, we can handle RI and RE at higher frequency.