Section of MSTP Bacnet controllers not discoverable/communicating

[madperry]

I am attempting to add 7 Distech ECB-203 controllers to an existing network. Network is on a Loytec MSTP router. There are 10 devices already on this network(almost all are ECB-203's). Having zero issues communicating with them. To add to this network, I broke it(after the 3rd existing device) and continued the daisy-chain through my 7 new devices, then re-established a connection on the already existing remaining devices. All wire terminations and polarity verified countless times. Through this Loytec, I can still communicate/discover the original 10 devices. I can not discover the 7 new devices. They are only indicating a persistent, rapid blinking orange transmit led. If I isolate the daisy chain of new devices, I can easily discover/communicate with these devices via laptop/N4 and a different router.

I have confirmed baud rate on the Loytec router is 38400, as well as the original 10 devices. When I am using the router on my laptop, it also is at 38400, and I can discover the new devices. Have confirmed no conflicts with MAC addresses or Device ID's. Out of desperation, I have tried reversing the polarity at the beginning and ending of my new daisy chain section of network. Same result. Original 10 communicating. New 7, not.

Any suggestions are greatly appreciated.

[ABBCylonAmerica]

It sounds like you didn't tell the network that you were adding new units.

Go into the highest addressed existing device and tell it to offer the token to higher numbered devices on the MSTP wire.

For example, if you have devices 0 through 9 on the network right now, go to device 9's device object and set the max master to 127 so that any and all new devices you attach are offered the token to join the network. If the highest numbered device on the network is 15, then go into it and set the max-master to 127.

Assuming your devices are at the same baud rate this should be all you need to do to fix your issues.

[JCIman]

It sounds like you didn't tell the network that you were adding new units.

Go into the highest addressed existing device and tell it to offer the token to higher numbered devices on the MSTP wire.

For example, if you have devices 0 through 9 on the network right now, go to device 9's device object and set the max master to 127 so that any and all new devices you attach are offered the token to join the network. If the highest numbered device on the network is 15, then go into it and set the max-master to 127.

Assuming your devices are at the same baud rate this should be all you need to do to fix your issues.

I agree with what your saying. The only exception I would take is that I think you should change the max master to 127 on all the devices. In your example, if controller 15 went offline, 14 would not be able to pass the token to 16.

[orion242]

Ditto.

Max Masters, the setting that shouldn't be...

[madperry]

It sounds like you didn't tell the network that you were adding new units.

Go into the highest addressed existing device and tell it to offer the token to higher numbered devices on the MSTP wire.

For example, if you have devices 0 through 9 on the network right now, go to device 9's device object and set the max master to 127 so that any and all new devices you attach are offered the token to join the network. If the highest numbered device on the network is 15, then go into it and set the max-master to 127.

Assuming your devices are at the same baud rate this should be all you need to do to fix your issues.

Thanks for the suggestion. Will try it out and post back results.

[bolt950rider]

I agree with what your saying. The only exception I would take is that I think you should change the max master to 127 on all the devices. In your example, if controller 15 went offline, 14 would not be able to pass the token to 16.

Sounds like this is the issue as well.

I’d make sure the routers set to 127 and then change them all from 127. If you do need to mess with the max masters to reduce timing I’d do the last 3 active addresses, F doing them all.
But really if you’re changing max masters from 127 to get a better stable bus, you likely have bigger issues at hand and it will likely not make a lot or any difference.


Sent from my iPhone using Tapatalk

[JCIman]

If you do need to mess with the max masters to reduce timing I’d do the last 3 active addresses, F doing them all.

What happens if the last 3 controllers are on the same transformer and they lose power? Why wouldn’t you set them all to 127?

[madperry]

I set all devices and the router to 127. No difference. Same issue. They were all at odd #'s. 13 on one. Router was at 24. I rebooted all devices after changing, and power cycled just because. Confirmed 38400 baud rate.

I feel like there must be a wire issue, but can't understand why all original devices before and after the new chain are perfectly communicating. As stated before, if I seperate the chain of 7 new devices, I can talk to them on laptop.

[JCIman]

I set all devices and the router to 127. No difference. Same issue. They were all at odd #'s. 13 on one. Router was at 24. I rebooted all devices after changing, and power cycled just because. Confirmed 38400 baud rate.

I feel like there must be a wire issue, but can't understand why all original devices before and after the new chain are perfectly communicating. As stated before, if I seperate the chain of 7 new devices, I can talk to them on laptop.

Time to break out wireshark, and see what the last original device is doing with the token. You will definitely get a sense of what’s going on looking at the messages. I would wireshark on the MSTP side , then the ip side. That will tell you who’s dropping the ball. It sounds like it’s the router.

[jschulze]

When I am using the router on my laptop, it also is at 38400, and I can discover the new devices.

Please expand on this statement. Does this mean that you used the router with your laptop and BACnet discovery software (YABE, BQT, BDT, etc.) and could discover all 7 new devices AND all 10 old devices?

If so, there is no problem with your BACnet network, the problem is with your front-end client. Perhaps a configuration issue or license limitation?

[JCIman]

Please expand on this statement. Does this mean that you used the router with your laptop and BACnet discovery software (YABE, BQT, BDT, etc.) and could discover all 7 new devices AND all 10 old devices?

If so, there is no problem with your BACnet network, the problem is with your front-end client. Perhaps a configuration issue or license limitation?

Great point. When I read what the OP said, my mind read "a router", but he actually said "the router". I would then wire shark on the IP side and see what's happening with the WHOIS. I would also try adding a controller manually and bypassing discovery.

[numbawunfela]

It sure sounddd like a Max Master setting.... the controllers are ransmittimg and the other controllers are not passing the token to them sounded like....
I would check for proper grounding and power on the new controllers. I have had ECBs that will be riding a 7vdc offset because the secondary was not grounded. Then nobody sees them. Or riding a sine wave when they are not grounded at all. Depending on how the router is powered it may also ride that offset and see the devices.
I had a guy tell Schneider their AS-P is crap because it could not see a trunk of controllers, but he could with his laptop and YABE with an RS485 adapter. He did not understand grounding. The AS-P was grounded, the controllers (Schneider 8600 stats) were not. The controllers were riding a 24v sine wave, so the AS-P had nothing. Ghe laptop was on a battery, so the USB powered RS485 adapter also was not grounded, and so could follow the 24v sine wave without issues.
Schneider responded by telling him he was the idiot and that they would be happy to fly a tech from Texas to fix his crap and charge him back for it. Fortunately it got resolved before this got further out of hand.

[digo]

How are the 7 new devices powered? Distech grounds the secondary 24VAC common and uses that as the signal reference for the RS-485. If those commons aren't all common, you've got a problem.

[madperry]

I should have updated this post sooner but have finally got this issue resolved. Unfortunately, I dont have a full understanding how I was able to do it.

To clarify, I was unable to see those 7 new devices with the existing permanent router. I was only able to talk to them when they were isolated and using laptop with my PORTABLE router.

All of the ECB controllers are by default set to auto baud rate. The router dictates baud rate to them. They simply set themselves to what is detected on the network. Or that is my understanding. Both the existing router and portable were at 38400. To resolve, I seperated trunk after 3rd device, and used laptop with portable router to view 7 new devices, and 7 existing devices. Initially, couldnt see the 7 existing. Changed baud rate to 9600, then back to 38400. ALL of them were then discoverable on laptop. Figured that was progress, so I reconnected trunk, and went to the JACE to discover. Same result. Back where I started. I then fiddled with baud rate on existing permanent router. Final setting 38400. STILL could not discover those 7 new devices. Finally, I manually entered those new devices into the JACE. After 15 minutes or less, they all, 1 by 1, came in and started working flawlessly. It is all perfect now. And I really dont understand why.

Responding to an above comment, you are 100% correct on proper grounding. I too have experienced this. Lost half a trunk once because of a common wire that had been pulled out of device. But I had carefully inspected every wire on every device in this circumstance. My wiring was sound in this instance.

Thank you all for your input. Maybe someone can explain why I was able to finally get these all talking?

[JCIman]

I should have updated this post sooner but have finally got this issue resolved. Unfortunately, I dont have a full understanding how I was able to do it.

To clarify, I was unable to see those 7 new devices with the existing permanent router. I was only able to talk to them when they were isolated and using laptop with my PORTABLE router.

All of the ECB controllers are by default set to auto baud rate. The router dictates baud rate to them. They simply set themselves to what is detected on the network. Or that is my understanding. Both the existing router and portable were at 38400. To resolve, I seperated trunk after 3rd device, and used laptop with portable router to view 7 new devices, and 7 existing devices. Initially, couldnt see the 7 existing. Changed baud rate to 9600, then back to 38400. ALL of them were then discoverable on laptop. Figured that was progress, so I reconnected trunk, and went to the JACE to discover. Same result. Back where I started. I then fiddled with baud rate on existing permanent router. Final setting 38400. STILL could not discover those 7 new devices. Finally, I manually entered those new devices into the JACE. After 15 minutes or less, they all, 1 by 1, came in and started working flawlessly. It is all perfect now. And I really dont understand why.

Responding to an above comment, you are 100% correct on proper grounding. I too have experienced this. Lost half a trunk once because of a common wire that had been pulled out of device. But I had carefully inspected every wire on every device in this circumstance. My wiring was sound in this instance.

Thank you all for your input. Maybe someone can explain why I was able to finally get these all talking?

without Wiresharking it would just be a guess as to why. I know I have better luck adding devices manually sometimes.

[lin]

All of the ECB controllers are by default set to auto baud rate. The router dictates baud rate to them. They simply set themselves to what is detected on the network.

Maybe auto baud rate cause the problem. There are 2 major problems on auto baud rate.
First problem is when to re-detect baud rate. Some only detect once when starting up. Some wait a certain time to re-detect, I know JCI FEC ddc wait about 150 seconds, but don't know how ECB handle it.
Second problem is how to switch baud rate. Changing baud rate on the router then restart it may not work. so we introduce baud rate forcing feature.

More on http://www.hvacrcontrol.com/fixedaut...rate-for-mstp/