Testing/usage Modbus 3.7.x

[John245]

In pretty sure when a value changed I was updating the device twice which was causing your event to fire twice. I hope that was it.


Sent from my iPhone using Tapatalk

Beerygaz,

To be clear the issue that the event is fired twice is solved.

I hope that this issue also solved the issue that pointed me to this minor bug.

---

John

[John245]

Disabling gateway will not close COM port

Hi Beerygaz,

The manual stated "Gateway Enabled - By default the gateway is disabled, allowing you to specify the correct parameters before polling starts. This also allows you to temporarily suspend polling without the need to remove the device from HomeSeer completely."

This could be helpful and I understand why you don't close the COM port in this case. However I have devices that use the RS-485 connection also to update the FW. In that case it is not handy that the COM port is still open as I need to disable the Modbus PI to get the COM port available (which disables also my other devices).

I would appreciate it if the COM port can be closed when the gateway is disabled.

---
John

[beerygaz]

This is quite a big change in the design and I don't want to make a quick fix and introduce new issues. I'll put it on the feature list and implement it as soon as I have a little more time to implement this properly.


Sent from my iPhone using Tapatalk

[John245]

This is quite a big change in the design and I don't want to make a quick fix and introduce new issues. I'll put it on the feature list and implement it as soon as I have a little more time to implement this properly.


Sent from my iPhone using Tapatalk

Hi Beerygaz,

As you can see in the first post of this thread I indicated it as feature request.

I appreciate it when you implement it, but normally I should not have that much FW upgrades. And I can also plan these upgrades in ta maintenance time slot of HS3.

---
John


Sent from my iPhone using Tapatalk

[John245]

Beerygaz,

To be clear the issue that the event is fired twice is solved.

I hope that this issue also solved the issue that pointed me to this minor bug.

---

John

Hi Beerygaz,

This fix solved also the issue that pointed me to this bug.

Thanks again.

---
John

[John245]

Slave ID

Hi Beerygaz,

The manual stated: "Slave ID - Each module or device on the bus has a unique slave ID, up to 254 devices can be connected to a single bus, this ID uniquely identifies the specific device on the bus with which you wish to communicate."

Can 255 be set as Slave ID?

---
John

[beerygaz]

A little further research suggests slave IDs can be 0-247. I'll update the docs accordingly.

So your answer is no. Up to a max of 247.


Sent from my iPhone using Tapatalk

[John245]

Hi Beerygaz,

I have a Modbus RTU over IP device and see the following issue after a hour of 12 running


Jul-13 7:41:37 AM Modbus Error Error in Read: Response was not of expected transaction ID. Expected 17570, received 17569.
Jul-13 7:41:37 AM Modbus Error Error in Read: Response was not of expected transaction ID. Expected 17566, received 17565.
Jul-13 7:41:37 AM Modbus Error Error in Read: Response was not of expected transaction ID. Expected 17562, received 17561.
Jul-13 7:41:37 AM Modbus Error Error in Read: Response was not of expected transaction ID. Expected 17558, received 17557.
Jul-13 7:41:37 AM Modbus Error Error in Read: Response was not of expected transaction ID. Expected 17554, received 17553.
Jul-13 7:41:37 AM Modbus Error Error in Read: Response was not of expected transaction ID. Expected 17550, received 17549.

What could be the issue?

---
John

[beerygaz]

How often are you polling? It looks like the device can't keep up with your polling schedule.


Sent from my iPhone using Tapatalk

[John245]

I'm polling every 2 minutes. I also polled for every 30 seconds for 1 hour without any issue.

I can try to change the poll period to 5 minutes but I don't think that will solve the issue as the polling went well for 12 hours.

Once the error occurred there is no correction by the PI. A restart solves the issue.

---
John


Sent from my iPhone using Tapatalk

[beerygaz]

That's not the issue the. If you were polling every second or less it might have been, but not minutes.


Sent from my iPhone using Tapatalk

[John245]

That is clear. Should the plug in (modbus master) not take an action to synchronize (for e.g. A read with the same transaction ID)? As after this error all reads are asynchronous. With as result no read out of the data.

---
John


Sent from my iPhone using Tapatalk

[John245]

Hi Beerygaz,

For test purposes I increased the delay between each address poll to 100 ms (instead of 5 ms) and decreased the poll interval to 10 s (instead of 2 minutes).

For your info 41 addresses of this device are polled.

---
John

[beerygaz]

5ms between each address is too short for the device to respond and the packet to arrive and be processed. Did the increase to 100ms help?


Sent from my iPhone using Tapatalk

[John245]

5ms between each address is too short for the device to respond and the packet to arrive and be processed. Did the increase to 100ms help?


Sent from my iPhone using Tapatalk

Hi Beerygaz,

I need to test for a longer period. As with 5 ms between each address poll the system was successful for over 12 hours.

---
John