I’m trying to figure out what’s wrong with my bacnet setup.
I have a T3-TB as a Bacnet master and 8 TSAT10 as Bacnet slaves.
As soon as the Bacnet master is plug everything go wrong. I can no longer communicate with the RS485/usb adapter. I can still connect to the master using the Ethernet port.
I have tried many times and have not been able to replicate the problem you describe. I did find something unusual in the product list tree of T3000 on your screen shots. You can delete all the products and rescan the newtwork, you shold see something closer to what I have shown in my screen shots.
Database - all nodes -delete all
Another thing you can check is to confirm the baud rate and protocol of all devices. If even one MSTP device is at a different baud rate it will cause all communications to fail.
Thanks for the test. I tried to unplug all the TSTAT10 to keep only one connected. I checked what you mentionned above but only the T3-TB appear afer the scan. On my previous scan I was using the USB/RS485 converter to make the TSTAT10 appear on the tree. Also yesterday I tried using modbus communication instead of Bacnet and the problem was basically the same. On the same controller I have a Modbus slave which is working perfeclty on both RS485 ports which I think rule out the hardware on the T3-TB. One more point I didn’t mentionned is it did worked in the past and it stopped working when I tried to add the modbus slave on my network.
For issues which worked for a while, then dont work now, I always suggest getting an simple dual channel oscilloscope to check the signal integrity. If the cable has been badly crimped or cut somewhere it will show up with the scope.
I even have packets error when the rs485 isn’t connected to anything.
After some test theres nothing pointing in something wrong on the wiring.
Also I reset the T3-TB to factory and the problem remains.
The signal looks OK, your scope has two channels, probably it can do the math functions so it will be more informative to watch the differential, A - B plot.
Made the test with only one TSTAT10 connected to the T3-TB.
Of what I see I can only see traffic coming from the TSTAT number 17(HEX11). I also have the error malformed packets. Which I thinks is the packets coming from the T3-TB.
I had two TSTAT10 wired and configured to Bacnetslave at 115200 and one of them switch it’s baud rate to 19200. On T3000 software everything appeared at 115200 but using wireshark it wasn’t showing up until I changed wireshark baudrate to 19200. I went to the TSTAT and using the buttons I went to baudrate settings and it was displaying 115200 baudrate. I selected the 115200 baudrate again and confirmed it and then it started to appear on wireshark with my other TSTAT. I’ll continue to work on this today.
Well, I was able to connect two TSTAT10, one master, one slave, each at the extremity of the RS485 network with the T3-TB unplugged. If I try to unplug the TSTAT master and replace it with the T3-TB the network crash as before. Here’s a screenshot of wireshark with the two TSTAT10 setup in master/slave.
Chelsea will report back shortly. I asked the team to get rid of the Bacnet master and Bacnet slave settings, I see no benefit to having this as an option anyway. The baud rate setting changes after you set it at the keypad, that could be a firmware version issue. Please make sure this is the latest and greatest version by doing the firmware update procedure. You might need to switch temporarily to Modbus to do this, Chelsea will report back on this as well.
It seems that the serial port data of TB is incorrect. Please confirm the connection of the two cables of the rs485. Check the baud rate of the TB. Do not connect to any TSTAT10, and ensure that the usual Bacnet whois and token passing traffic sent by the TB on its own is normal
The network health is not normal, many lost packets. My only guess at this point is whether connection is ok, perhaps RS485- and RS485+ are reversed?
The following is normal.
I just finished reupdating everything and I still got the same problem.
Only interesting thing is I wasn’t able to update through the rs485 port of the TB and had to connect with the RJ45 cable.
I’m suspecting the RS485 ports might have be damaged …
Updates over RS485 were working over Modbus but I am not sure over RS485/ Bacnet though. CHelsea will report back on that. To test if the RS485 port is working connect with a USB converter direct from your PC to the device and scan for devices. It should show up in the device list tree at Tab1. If the USB to RS485 converter has LEDs they should flash during the scan. The communications health at Tab2 should show some successful packets. In my example I am connected over Ethernet but the status over RS485 is shown in the same way.
You mention you are doing the subnet communications at 112k baud, try reducing that by half and see if things improve.
Use the communications status line as your feedback during testing. In my example I see that we’re at 80% successful packets, could be a firmware issue going on here. Our team will be looking at all this and report back on Monday.
We do not support updating firmware via mstp now, if you want to update it by serial port, the protocol must be modified to modbus slave, and only sub RS485 port support it. TB has two rs485 ports. The main rs485 has the same functions except that it does not support updating firmware.
