News:

  • October 03, 2026, 04:03:06 AM

Login with username, password and session length

Author Topic: Two BRXs one process  (Read 15532 times)

Ingeniero

  • Full Member
  • ***
  • Posts: 45
Two BRXs one process
« on: July 22, 2019, 10:51:53 PM »
I work on a very fast process that uses math integration. in other brand PLCs this means you have to keep the scan time constant, I imagine the same applies for a do more system?

In the past several companies used two plcs, one to handle communications, ethernet and alarms and sequencing fans and pumps, the other to run the main process.  The slow one could run at 50 milliseconds or so and the fast PLC 10 milliseconds. This architecture was used back in the 1990s when PLCs were slower and had less memory. I frankly forget how the Two PLCs communicated, been a while, there has to be communications between the PLCs.

I understand that domore has a "bump" in scan time for ethernet communications. This is OK for the Slow communications PLC above but might not work for the fast process PLC. Could using two PLCs that communicate with mod bus work. Does mod bus also cause a scan time Bump? Is there some other serial port or USB port communications protocol that would work.

BobO

  • Host Moderator
  • Hero Member
  • *****
  • Posts: 6176
  • Yes Pinky, Do-more will control the world!
Re: Two BRXs one process
« Reply #1 on: July 22, 2019, 11:12:20 PM »
I work on a very fast process that uses math integration. in other brand PLCs this means you have to keep the scan time constant, I imagine the same applies for a do more system?

In the past several companies used two plcs, one to handle communications, ethernet and alarms and sequencing fans and pumps, the other to run the main process.  The slow one could run at 50 milliseconds or so and the fast PLC 10 milliseconds. This architecture was used back in the 1990s when PLCs were slower and had less memory. I frankly forget how the Two PLCs communicated, been a while, there has to be communications between the PLCs.

I understand that domore has a "bump" in scan time for ethernet communications. This is OK for the Slow communications PLC above but might not work for the fast process PLC. Could using two PLCs that communicate with mod bus work. Does mod bus also cause a scan time Bump? Is there some other serial port or USB port communications protocol that would work.

Serial comm doesn't affect the scan very much. You can also use a timed interrupt to run math at a very precise interval.
"It has recently come to our attention that users spend 95% of their time using 5% of the available features. That might be relevant." -BobO

Ingeniero

  • Full Member
  • ***
  • Posts: 45
Re: Two BRXs one process
« Reply #2 on: July 22, 2019, 11:49:43 PM »
mod bus via serial comms?

BobO

  • Host Moderator
  • Hero Member
  • *****
  • Posts: 6176
  • Yes Pinky, Do-more will control the world!
Re: Two BRXs one process
« Reply #3 on: July 22, 2019, 11:55:04 PM »
mod bus via serial comms?

Serial in general, but yes, Modbus also.
"It has recently come to our attention that users spend 95% of their time using 5% of the available features. That might be relevant." -BobO

ATU

  • Internal Dev
  • Hero Member
  • ****
  • Posts: 2126
  • YKPAIHA
    • ATU, Inc.
Re: Two BRXs one process
« Reply #4 on: July 23, 2019, 08:11:09 AM »
I have 2 BRX's running one process. I use Modbus  to send data that needs fast updates. Its sending 16 Modbus registers and I would guess the transmission time is under 8 ms or less. However, that really depends on scan time. My scan time is around 3-6 ms, but its a huge program. I am sure it would be less if the scan time was less. Depends on how you structure the program. What's nice about the Modbus instructions is that I can send the data exactly when I want to,rather than wait on polling.   Peerlink is also good to use for configuration data when a 100 ms update is acceptable.

Ingeniero

  • Full Member
  • ***
  • Posts: 45
Re: Two BRXs one process
« Reply #5 on: July 23, 2019, 09:50:09 PM »
thanks ATU.

100 ms might work. The faster PLC would just run the 10MS process with an interrupt to trigger and communications to the Second slower PLC. This PLC would have only inputs needed for directly in the process alogrithm.

The second slower PLC would have sequencing, alarms and trips and the ethernet . There has to be some way to trip the PLCs when the Mod bus is down. Each PLC's trip output would be hard wired to the other PLC. Some critical inputs can be wired to both PLC's.

Cool, my dream project is to get away from allan bradley and reduce the project costs by $30,000 to $40,000.


BobO

  • Host Moderator
  • Hero Member
  • *****
  • Posts: 6176
  • Yes Pinky, Do-more will control the world!
Re: Two BRXs one process
« Reply #6 on: July 23, 2019, 10:06:07 PM »
How fast do you need? Pretty hard to get a 10ms scan on BRX.
"It has recently come to our attention that users spend 95% of their time using 5% of the available features. That might be relevant." -BobO

Ingeniero

  • Full Member
  • ***
  • Posts: 45
Re: Two BRXs one process
« Reply #7 on: July 24, 2019, 02:10:01 AM »
Bob O

"""How fast do you need? Pretty hard to get a 10ms scan on BRX."""

ms is milliseconds. Hard to get 10 millisecond scan rate? For the program or for IO?

Actually I want to use the Th1 because the analog input module scan rate is less than 10ms. The BRX analog imput  is like 45ms.

The terminator 8 channel IO card updates all 8 channels every scan, which I am assuming could be 5 milliseconds. The conversion time in fast mode is given as .5 milliseconds.  What exactly is conversion time? see page 1 of below.

https://cdn.automationdirect.com/static/specs/t1f08ad1.pdf

I would have to use conversion modules from other vendors to measure signals like speed and temperatures and than these would be read on the fast analog card. The whole loop, process to control device should operate at a steady scan rate of 10 Milliseconds. The conversion modules scan rate is not coordinated by the PLC but the systems operates well if these are similar to the PLC's scan rate.

I am guessing and generalizing but the time constant of the process is around 1 sec.

Am I barking up the wrong tree thinking the th1 with the terminator IO modules could do this? 

thanks Bob O.





ADC Product Engineer

  • Hero Member
  • *****
  • Posts: 270
Re: Two BRXs one process
« Reply #8 on: July 24, 2019, 08:45:51 AM »
Your time to send (both ways) and time on wire will be the limiting factor.

BobO

  • Host Moderator
  • Hero Member
  • *****
  • Posts: 6176
  • Yes Pinky, Do-more will control the world!
Re: Two BRXs one process
« Reply #9 on: July 24, 2019, 09:24:27 AM »
Bob O

"""How fast do you need? Pretty hard to get a 10ms scan on BRX."""

ms is milliseconds. Hard to get 10 millisecond scan rate? For the program or for IO?

Actually I want to use the Th1 because the analog input module scan rate is less than 10ms. The BRX analog imput  is like 45ms.

The terminator 8 channel IO card updates all 8 channels every scan, which I am assuming could be 5 milliseconds. The conversion time in fast mode is given as .5 milliseconds.  What exactly is conversion time? see page 1 of below.

https://cdn.automationdirect.com/static/specs/t1f08ad1.pdf

I would have to use conversion modules from other vendors to measure signals like speed and temperatures and than these would be read on the fast analog card. The whole loop, process to control device should operate at a steady scan rate of 10 Milliseconds. The conversion modules scan rate is not coordinated by the PLC but the systems operates well if these are similar to the PLC's scan rate.

I am guessing and generalizing but the time constant of the process is around 1 sec.

Am I barking up the wrong tree thinking the th1 with the terminator IO modules could do this? 

thanks Bob O.

If you are talking about BRX analog modules, yes, they are going to be slower. If you didn't need more than 4 channels, the BRX onboard channels are about 1.2ms.

I would be surprised if temperature readings are fast anywhere. Temperature modules update at over 1 second.

Even with Ethernet comms, BRX shouldn't be spiking over 10ms. I don't know anything about your program, but most scans are below 5ms.

We are developing faster analog cards for BRX, but they won't be available until next year.
"It has recently come to our attention that users spend 95% of their time using 5% of the available features. That might be relevant." -BobO

Ingeniero

  • Full Member
  • ***
  • Posts: 45
Re: Two BRXs one process
« Reply #10 on: July 24, 2019, 02:16:11 PM »
Yes the temperature loop is slower, the TC response is slow, the AB TC card scan rate was 100ms. The PLC portion ran at 10MS. I found some faster TC modules but dont have the link right now. I have been kicking this around for a few years now.

The speed loop  is a mag pickup on a toothed wheel, no response time from the sensor to the frequency module, the frequency modules can be 5 or 10 ms. there are enough teeth on the wheel to pickup small differences in acceleration and speed. Then the analog card scan which was fast at 5 or 10 ms and the PLC loop was 10ms.

The fuel valve actuators were hydraulic or electronic, faster then pneumatic. I am going to say the response time of the fuel system was .5 seconds.

 In some applications 4 analog inputs could be enough so the BRX could work.