Changing SNAP I/O Modules

Is it necessary to turn power off to the EB1 brain when changing a SNAP i/o module?

Is it necessary to restart the strategy on a PAC-S1 after changing a SNAP i/o module on a rack that uses a EB1 brain?


Short answer: There is no short answer for this one.

Looking at the rack side of things first.

First up - This answer is assuming a SNAP-PAC system. The answer is different if its a G1, G4 or Mistic system.

No, you don’t ‘have to’ turn power off when swapping a 4-channel [B]digital [/B]input or output module, ie. the same module swapped for the same module type in the exact same position.
(We recommend it, but you don’t have to).
Reason is that the digital points are ‘hard wired’ to the controller/io, so it’s a pretty straightforward event for the I/O processor to handle. Digital points are just on or off, it does not know if that’s what the module is doing, or if the module is missing.
That said, you can’t swap an input module for an output or the other way around, because the I/O processor does keep track of what its pins are configured for, innies or outies.
(That’s why you can’t add a digital module and have it work. You need to tell the I/O brain what modules are where - it needs to configure its pins as inputs, outputs or data - this is done via the configuration in the strategy).

Keep in mind that if you have any points that are ‘On’, they are going to turn off as far as the strategy is concerned when you unplug the module.
This applies for both inputs and outputs.
So, you need to think about what’s going to happen in your strategy when that digital input point(s) turns off.
What’s going to happen when that digital input turns off in your strategy? What if that input is a flow switch, and when it turns off, the strategy turns off the whole system?
When you plug in the new output module, will your strategy turn that point back on? What if the point was on before you swapped modules, but the strategy is now in a timer loop, you swap the module, but the point is not going to turn on till after the timer expires, what happens in the mean time to your process, is the tank going to run dry because the pump is not running?

Yes, you need to power down the rack when swapping analog modules AND HIGH-DENSITY DIGITAL MODULES.
Analog (and [B]all [/B]high density digital modules (input [B]or [/B]output)) are intelligent. This means they are having a little conversation over the rack to the I/O processor at all times. (EB or R - both have an I/O processor).
You can’t just jerk out a module during that conversation. So don’t.

Regarding restarting the strategy.

There are two different answers. One for digital, one for analog. (ie, one for leaving the rack on, one for turning the rack off).
Digital: As mentioned in the digital module swap answer, if your strategy is keeping on top of reads and writes, you probably don’t need to restart the strategy.

Analog: Since you can’t change the modules with the power on, the rack is going to go offline from the controller. When you power back up the rack, it will not automatically get back online with the controller.
No matter what module swapping might be going on, you should be running some sort of I/O enabler chart;
This is the chart provided and supported by Opto.
You can use that, or write your own, but no matter what, you should be re-enabling communications to the I/O units from your strategy.
You should be keeping track of what I/O racks are going on and off line.
You should be managing the strategy to handle off line racks. In other words, there is no point turning on a pump if the rack that the pump digital output is on is offline.

Racks go offline for two main reasons, power cycle and network connectivity issues.

You need a chart / strategy to deal with reconnecting the controller to remote I/O. The good thing is that power and network issues can be dealt with the same way since they look the same to the controller… the I/O simply stops responding.

So, to sum it up.
Yes, you ‘can’ hot swap the same 4-channel digital module for the same digital module in the same location.
No, you can’t hot swap analog (or high density digital modules) modules.
Yes, you will either need to restart the strategy or have a chart that re-enables the I/O rack if you cycle power or re-establish network connectivity.

However, I gota ask… How often and why the need to hot swap modules?
In 25 years of using Opto gear, I have never needed to swap modules. They just don’t go bad. No really… Other than lighting, I have never had a module just go bad. (Other than during commissioning when we find people have wired 240 VAC to a 24 VAC module).

Hope that helps.

Thank you for the reply.

I agree your I/O is pretty solid and without human error it is rare to see a module go out. We have thousands of G4 points in the field and have grown to love the ability to “hot swap” them without powering down the brick or restarting the strategy. The “hot swap” is usually one of the last steps in troubleshooting an instrument. Like you said, it’s usually not the module but it helps (especially the tech in the field) to see that the module is good by taking it and putting it in another “like kind” point. Once you establish that the module is good it focuses the technician to look elsewhere.

I will say though on the SNAP IO we have more issues with “connectivity”. We have seen where the module hold down screw tension can cause problems. Also, the Snap TEX cable connections to the module…we have seen where if the “wedged keeper” (not sure what the official name is) is properly seated, the module would not be recognized. If we unsnapped the wedged keeper it allowed the connector to “straighten up” and the module was recognized.

Again thank you for your help…we love your products.

I’d disagree on the short answer, and as a Mary, I do feel obligated to be contrary, so short answer would be: DON’T DO IT.

We officially do not recommend hot-swapping of modules, although we know people do it with 4-channel digital modules w/out a problem. And I can see how that would be handy for ruling out the module as the issue.

In any case, a few other thoughts, perhaps more info than you ever wanted to know on the topic…
To the EB1 (and other SNAP brains) a digital input or empty slot look the same, but if you tell the brain (either via your strategy or directly using PAC Manager) a module/point is a digital output, then you want to write values to it, as Ben mentioned. But the brain does not automatically detect the presence of those 4-channel digital modules.

Analog, serial, and High-Density Digital modules, on the other hand, are “smarter” and let the brain know their hardware type during that ongoing “conversation” Ben mentioned. You will still need to configure them for something other than defaults for scaling, calibration, etc. You’ll notice that if your PAC Control strategy includes an analog module type that does NOT match what’s plugged into your rack, you’ll get a warning in your PAC Control message queue when the controller actually sends that module config info to the brain.

Also, while we’re on the topic, the default of [B]-32768 as “out of range” value[/B] is not carved in stone. Check out [URL=“”][U][B]this post[/B][/U] for options and opinions on that topic.

Somewhat related, if you’re interested in detecting the loss of communications between controllers and brains, [U][B]this thread[/B][/U] even has a little video at the end, and more fun suggestions from OptoFans all around the globe!

Hope that helps!


I’m a little late to the party here but I would like to put my two cents worth into this discussion.

I’ve been working with snap since it first came out over twenty years ago. While I agree that it is best to power down the rack when changing any module, in reality that just isn’t always an option. In twenty years I’ve have changed numerous analog modules, probably about 50, without powering down the rack. Most of the time I seem to be changing AIRTD or AIMA modules. Changing the modules while the rack is still powered can lead to some weird issues but you can recover from them. The most common issue I see when changing a module on a powered rack is the loss of the scaling for that module. This can be fixed by going into PAC Manager and reconfiguring the module. On a couple of occasions the scaling has been lost for all of the analog modules on a rack. Again PAC Manager can be used to reconfigure all of the analog modules. As mentioned above, you need to be aware of the possible impacts on your strategy when doing this as the program may see weird inputs during the time the module is out.

One other occasion where I have removed modules from a powered rack is when I am trouble shooting a problem. For example, in a recent case I was having trouble getting proper values from a particular module. PAC Manager showed lots of errors on that module so I powered down the rack and replaced the module. When the rack was fired up I was still getting lots of errors on the new module and incorrect values. From prior experience I new that sometimes the module that is not communicating and providing proper values is not the one that’s actually causing the problem. To troubleshoot this problem I pull out each module individually and see what happens with the module that appears bad. Eventually I will remove the true bad module and the errors will stop. The bad module is then replaced. Again, probably not the best way to do this but sometimes you just need to do what you need to do to keep things up and running.