Bridge and Bolt communication issues at 50 ft

[support]

Hello,

After receiving more information from my colleague we do not think the antenna is broken. Here is some feedback from my colleague,

" channel 7 is overlapping with 11, he should change it to 6 instead, it is only 1, 6 and 11 that should be used in 2.4GHz"

Can you please try 6?

[amromahayni]

Hi Nick
Channel 6 is used in our plant.
Can you give us the all the channel posibillities that we can run on for 2.4GHz and/or 5GHz. Or is there other recommendations
Thx

[support]

Is channel 11 also used?

[amromahayni]

yes used. and channel 1 is also used

[support]

So are all 3 of those channels heavily used? Have you reviewed all three channels with that Wi-Fi scanner software? It seems that if the channels are used but not fully loaded with traffic then they should still work.

If you do not have much 5 GHz usage in the network I think changing the access point to 5Ghz and choosing any channel would work fine.

[kent_hms]

There is a lot of confusion regarding channel usage especially in the 2.4GHz band of WLAN, the channels are 5MHz apart from each other, but each channel uses 20/22MHz bandwidth depending on modulation.
In the below example you can see 2 SSID:s using ch1, 1 using ch3, 2 using ch6 and 3 using ch11. In this setup 1, 3 and 6 will have a very bad performance with a lot of resends due to bad packages, channel 11 though should be ok.

But by changing the SSID on ch3 to either 1, 6 or 11 the problem would be gone. During normal usage the data usage in a network comes in bursts, when you open or save a file to a server or browsing the web, it is only when downloading very big files or streaming video in a high quality that the bandwidth is being used.

My point, the channels are not occupied that much in normal conditions, and as the WLAN is checking if the channel is occupied before sending the data, multiple SSID:s can share the same channel.
More info on CSMA/CA - https://en.wikipedia.org/wiki/Carrier-sense_multiple_access_with_collision_avoidance
There are numerous good sources of information regarding channel usage and other best practice example, a good one can be found at https://www.metageek.com/training/resources/why-channels-1-6-11.html they also manufacture the WiSpy spectrum analyzer that is needed for anyone working with WLAN:s to detect disturbances that are not originating from other WLAN:s but different RF sources (Motion detectors, Industrial remote control, Microwave ovens and so on)

[amromahayni]

Hi Kent
Thanks for your email and explanation
Please find attached the wifi scnning using the software you suggested. Please give advice according to what you see.

I configured the bridge and the 2 bolts on bluetooth and it looks the com is fine but we need to monitor on a longer period to see if any dropping happening.

regards
Amr

[kent_hms]

Hi Amr,

Bluetooth is a good choice if you don’t need high throughput, as it uses a very narrow bandwidth and jumping frequencies it will “always” find it’s way through.
Looking at your RF environment I would recommend changing the two SSID:s on channel 9 to either 1, 6 or 11 as they are now disturbing both channel 6 and 11. The printer using channel 10 (HP075B80) check if they are actually using the WLAN on it, if not, disable it. Check that regarding the other printers as well, if it’s not being used it should be disabled.
In the current setup, the only usable 2.4GHz channel is actually number 1.
In the 5GHz area, we only allow channel 36, 40, 44 and 48 in Access Point mode, and both 36 and 44 are available in your environment, so this is also an option.

All in all, a small cleanup in the 2.4GHz would improve the WLAN for your company, so it will be well worth doing it.

Glad that you found the explanation helpful!

[amromahayni]

Hi Kent

Thanks a lot for the details give … valuable
I will work with IT on your recommendation and see where we can reach

Regarding the bluetooth, we get a total of 128 I/O’s of data (all are discrete) going back and forth. They are not all used at the same time but I need to make further studies on their frequency, if required. Do you think that this much of data is considered high throughput and do you recommed to keep on bluetooth even if the wifi improved or your first option is wifi ??

Thanks once again

Regards
Amr

[kent_hms]

Hi,

What is connected to the different Anybus Wireless devices?
One PLC at the Bridge and I/O nodes at the bolts?
What protocol?
If Ethernet/IP have you changed the RPI (Requested Packet Interval) setting for the I/O devices, should be >64ms.

Bluetooth is best if the throughput is enough as it adapts to the environment.
WLAN is changing, a new printer is installed or a rogue router, so all of a sudden the WLAN can stop working, then you need to do an analysis of the new RF environment and correct it once again.

At least this is true regarding the 2.4GHz band, 5GHz is much better, the channels are positioned so they don’t overlap in normal circumstances.

But we do recommend Bluetooth if robustness is prioritized.

[amromahayni]

Hi Kent

One master scanner (1747-SN) connected to 4 RIO’s, the bolts are 2 of them. There is a converter (AN-X2 from prosoft technology) from DHRIO network to Ethernet to connect to the bolts through ethernet.

Is there any booster or repeater for the bluetooth that can be used ??

I got another question, What do you think the acceptable cycle time for capturing the data without delay ??

Regards
Amr

[kent_hms]

Hi,

We don’t have any booster or repeaters, and I would not recommend the use of it if you were to find one. WLAN and Bluetooth are half-duplex, so adding a repeater will double the response time in any wireless connection, and you don’t want to do that.

I referred to RPI before that it should be set to >64ms, and I’d say the same regarding the cycle time, in Bluetooth I would suggest around 100ms.
If you have many resends and timeouts you need to increase these values.

[amromahayni]

Hi Kent

Thanks for the reply and infromation. It is actually guiding us where to go

I don’t remember if I said before, I mounted the birdge/bolts on the machine and connect to an external PLC. I’ve also programmed closed loop function for testing purpose, sending a discrete output signal and recording the counts on the feedback input signal with an RPI of 100ms. The counts looks quite good.

We are going to try to function it with live machine by next Thursday. Hopefully it run smoothly.
I will let know about the results definitly.

Regards
Amr

[amromahayni]

Hi Kent

Good day !!

Unfortunatly the testing did not go through. We faced problems of data transmission not fast enough eventhough the Status Indication Lights are steady green.

We have total of 160 I/O’s (correction of what I said before 128 I/O’s). Partial of these I/O’s are encoder values send as discrete input and Hydraulic proportional valve send as discrete output. Those values must have correct readings all the time as position and speed need to be controlled.

What can we do to improve. apprecaite your feedback

Regards
Amr

[amromahayni]

Hi Kent

Happy new year

We are looking to have your support again … we made several modifications in our systems to improve the netowrking strength but are still facing un-stability and it is iterrupting machine functionality.

The modifications we made was:

1. The IT changed all the channels running on channel 1 to other channels and we are kept with Channel 1 for the wireless system only. Picture attached can show the details.

2. We isolate the encoder/pulses reading from machine functionality to reduce the load of data sent from/to PLC.

3. We provide heart beat (giving output signal and recieving input signal) and checked the timing between them. Attached is snapshots of these delays. We set a delay time of 2 sec and then 5 sec but without improvement

Please advice what we can do
Appreciate your feedback

Regards
Amr

[deryck_hms]

Hello Amr,

I spoke with Kent, regarding this case and we are thinking you could be transferring more data then can be supported over wireless. AWB devices can support up to 20 Mbps max. You will want to try to calculate how much data you are actually transferring and seeing if you can reduce it.

As I suggested previously try using the clients in Mac clone mode to help limit the traffic sent you send to it. You could also try using a dedicated AWB as the AP for each client.

Regards,
Deryck

[amromahayni]

Hello Deryck
I don’t know how mush of data is being transfered and if you remember I ask you how can we calculate it.
I already change the clients on Mac but did not have the chance for testing. Probably next Thursday we will be able but will do it with one client bolt and not with double bolt so we can try to reduce load traffic.
Regards
Amr

[deryck_hms]

Hello,

Calculating how much data can be tricky. The best option would be to sniff the traffic being sent though the Bridge or Bolt and add it up. This can be done with Wireshark and a hub or Managed switch.

1. Plug your PC, the AWB, and the connection that would go to the AWB into the Managed switch/Hub
2. If you are using a managed switch mirror the port the AWB is plugged into to your PC
3. Start a wireshark capture and log the traffic
4. You can then add a column that will show you cumulative bytes. You can then use that and the time to calculate the bps.

Regards,
Deryck

[amromahayni]

Hi Kent

Good Day

We need to have you feedback on our testing and what can we do to improve your system. It is still not working properly till know. We are having the system hanging once every 1 hr for 5-6 sec then going back to normal.

We made a modification on our PLC program and we reduced the amount of data transfer by 50%. We switch to wireless for the low puller only. Previously we were operating both pullers (low and high) on wireless. I attached a trend snap shot for the timer and it shows how the transmitting/recieving is dealyed by ~5-6 sec

Also, we made a calculation for the amount of data transfer every second so we can see if we are in the limit of
your system. The finiding are

PLC Scan Time = 1 scan (80 bits) every 30 ms on the low puller ==> ~ 34 scans (~2800 bits) a sec. Your system can run on 20 Mbps.
According to this calculation the system should work even on the 2 pullers with 160 bits of I/O’s

We have implemented all the recommendations from your end but still the system is not functioning as expected. My management is putting pressure for this to be finalized as it is causing us a lot of breakdowns.

Apprecaite your continuous support
Willing to hear from you soon

Regards

Amr

[deryck_hms]

Hello @amromahayni,

Are the AWB devices rebooting? Can you look at the internal web page and see if the up time counter on the over view tab resets? Can you also confirm what firmware version you are running?

Deryck