Distribuited I/O via Ethernet Radio link

[RTK]

Hello everyone...

I have an Siemens IM151-8PN/DP CPU and few distributed I/O's all under Profinet.

I'm trying to make a wireless bridge between the CPU and remote modules, using a Banner Surecross 900MHz on "star" configuration (http://www.bannerengineering.com/en-US/wireless/surecross_web_dataradio)

All is pretty much good - I can ping all devices, remote connect the PLC from the slave side and get online, download, etc... BUT in the HW Config doesn't "see" the remote devices, and PLC still returning a Bus Fault ?!

I bumped the "update time" at the Profinet to 8.000ms, to make sure is higher than the wireless update cycle, and still no luck...

I think is definitely something at Step 7, since the communication is basically working. No GSD file is available from vendor..... so any other clue would be appreciated !

Thanks in advance

 

[mk42]

With most wireless systems, there are a lot of errors due to interference. I've seen more success bumping the IO rate even higher, to something like 16 or 32 ms.

I'll be honest though, I'm not 100% sure why. I would think that sending data less often shouldn't affect the error rate/timeout. I've been meaning to take wireshark traces one of these times, to dissect it and see whats going on.

 

[Firejo]

The Banner radios are Frequency Hopping Spread Spectrum radios and thus must always have a network master. Because of this, even in a “star” topology all radio traffic must go through the master to be redistributed to the remotes. This adds time and there is no way around it. 8ms isn’t going to be even remotely long enough nor would I guess 16ms or 32ms would be. I’d start with at least 100ms and work down from there (if it works). You might have the wrong technology if the application can’t tolerate the latency. A “WiFi” type radio might give you the throughput you need (the higher the throughput generally the lower the latency, but not always). However the problem with most WiFi radios is they operate in the 2.4GHz frequency band which is very crowded which can also cause throughput issues. I’d look at something like these (http://www.data-linc.com/fastlinc/fastlinc910e.htm). They have a true “peer to peer” mode (so any device can communicate with any device without having to go through a “master”) and they use “WiFi” type technology but operate in the 900MHz frequency band which is much less crowded. Keep in mind that wireless “real time” I/O is a difficult nut to crack and isn’t always workable (more often than not it doesn’t work very well) so even the FLC910E’s might not do the trick but they are going to give you a much better chance than FHSS will.

 

[Firejo]

It’s also worth noting that Siemens has a wireless product line that actually works pretty well even for wireless I/O (they designed it to work with I/O). I’ve only had a little bit of exposure to them but what I did have left me with a pretty good impression. That said, they are expensive products and you have to jump through a lot of hoops to set them up properly.

 

[RTK]

Thanks for the inputs.

Yeah... I guess stil too good to be true having wireless bridge for distribuited I/O's

I'll bump up the update time and letyou guys know the results. I'm aware the Siemens product but like you said, the price still a little too high... maybe wait a little longer until gets more popular and cheaper....