DAQ E-1608 restarting after activity high?

[Bas de Wit]

Hi, 

Currently I am working with the E-1608 in a project where the DAQ is connected to a Linux computer over ethernet. 

When continuously running, it looks like the DAQ is resetting itself. The resetting happens randomly, given that it sometimes may take >30 minutes, and other times may occur multiple times with a minute in-between.

I observed that prior to a "reset" the activity LED of the DAQ is static on, and also the ethernet activity led looks like its close to static on. After a reboot? the DAQ continues as intended. 

I have tried to vary the update rates that I am sending towards the DAQ, however this issue still persists. Even when I create a very high update rate on the DO request, the issue does not occur. So my initial guess, that the buffer is overflown, doesn't seem to be the case. When disconnecting the LAN during operation, the DAQ also continues without any issues (so the output buffer also does not seem to be the issue?).

Did someone see this issue before? Does anyone know what may be causing this issue?

[Fausto]

Hello @Bas de Wit.

3 hours ago, Bas de Wit said:

I have tried to vary the update rates that I am sending towards the DAQ, however this issue still persists. Even when I create a very high update rate on the DO request, the issue does not occur. So my initial guess, that the buffer is overflown, doesn't seem to be the case. When disconnecting the LAN during operation, the DAQ also continues without any issues (so the output buffer also does not seem to be the issue?).

Seems contradicting.  Please clarify.

Is the network used dedicated to this project or is there other network traffic external to this project?

Do you have another E-1608 to swap in and verify the hardware and network configuration and behavior?

Regards,

Fausto

[Bas de Wit]

15 hours ago, Fausto said:

Hello @Bas de Wit.

Seems contradicting.  Please clarify.

Is the network used dedicated to this project or is there other network traffic external to this project?

Do you have another E-1608 to swap in and verify the hardware and network configuration and behavior?

Regards,

Fausto

hi @Fausto,

The network consists of a Linux computer, an ethernet camera and an E-1608, all connected via an unmanaged switch. 

The issue so far only occurred on a single E-1608 in our "final" system. I tried to reproduce the issue by running a bench-top equivalent system for a while, and here I did not notice the issue. This could be due to minor differences between the benchtop test setup, and the real system, however, these also are two different E-1608. 

In the final system the issue, where the activity led at a random interval becomes static on, followed by a short change of DO states (off, then on again), is present. This behaviour was not found in the bench-top test setup. 

To give some more background. The E-1608 is used as a sensing and control device on a motor setup. the DO are used as enable/disable signals for the motor controller. We use the analog input to monitor various values of the motor. I noticed on the system that once every while, the motor speed dipped for a short period. I checked the control values on the Linux PC, and there we saw that the connection to the E-1608 was lost and restarted again. At first I thought this was some networking issue between the Linux system and the E-1608. When diving somewhat deeper, I found that having an internet connection once the system is running, is not important. The E-1608 continues operation as per last command (unplugged the ethernet cable, system continued fine). What I did notice however was that at some intervals the activity led of the E-1608 became static on, followed by a dip in motor control and a message that the connection was lost on the Linux system. Once the activity led started blinking again the system continued as intended. This feels to me like there is some sort of overload on the E-1608, followed by a restart. 

Please let me know if you need more information or insights. 

[Fausto]

Hello @Bas de Wit.

On 9/10/2024 at 2:49 AM, Bas de Wit said:

I tried to reproduce the issue by running a bench-top equivalent system for a while, and here I did not notice the issue. This could be due to minor differences between the benchtop test setup, and the real system, however, these also are two different E-1608.

Please describe the configuration and wiring used for the bench-top system.

Please test the E-1608 module used for the bench-top system with the "final" system.  Does the issue continue with the replacement module?  A malfunctioning E-1608 would be the simplest resolution, while network communication breakdown not so simple.

 

On 9/10/2024 at 2:49 AM, Bas de Wit said:

When diving somewhat deeper, I found that having an internet connection once the system is running, is not important. The E-1608 continues operation as per last command (unplugged the ethernet cable, system continued fine).

As you noticed, disconnecting the network cable will cause the E-1608 to stop functioning.

 

[Bas de Wit]

11 hours ago, Fausto said:

Hello @Bas de Wit.

Please describe the configuration and wiring used for the bench-top system.

Please test the E-1608 module used for the bench-top system with the "final" system.  Does the issue continue with the replacement module?  A malfunctioning E-1608 would be the simplest resolution, while network communication breakdown not so simple.

 

As you noticed, disconnecting the network cable will cause the E-1608 to stop functioning.

 

Dear @Fausto, 

 

The wiring for both DAQ (benchtop and system) are equal. I make use of CH0L, CH0H, AGND, CH2H, AGND and some DI/DO and the +V0 for an NTC. I don't measure any out-of-boundary values on these. 

When disconnecting the network cable, the E-1608 continues in its last known state, which is perfectly fine for me at this moment.

We are indeed looking into swapping out the DAQ to verify that the non-working one is also malfunctioning on bench-top, and the other functions fine in the actual system.

 

Do you perhaps know if noise on the DI/DO/AN cause trouble with the DAQ?

[Fausto]

I'm not aware of noise on the DI/DO/AN channels disrupting the communication of the E-1608.