RIO as a remote I/O

Hi Opto22 experts,

How many of RIOs can put as a remote I/O per one of EPIC.

My BAS system got more than of 30 RIOs at the moment.(in the conceptual designed).

Thanks for answers.

Short answer: Somewhere around 256

Long answer: It depends.

While there is no real hard limit (beyond network addressing) placed on the number of remote I/O there will be a network traffic and human sanity limit that is reached at some point.

If you set your poll rates sensibly and have a well managed network and have a well designed PAC Control strategy or CODESYS program, you can easily go beyond 30.

That said, at some point you would need to start thinking about using tables and other methods to manage the I/O.
Also at that point, before going much further we would need to know more about the I/O point density at each location, ie, would a SNAP Controller at some key locations make more sense than a RIO?

Offline I/O would also need to be taken into account. If the network does go down (a backbone switch for example), how does the I/O (RIO) control the process in this condition vs a SNAP Controller that can run 16 charts locally and keep things on track while the network is repaired / rebooted.

Just some Thanksgiving food for thought.


Thanks Beno,

How about let’s RIO to control AHU, FCU, Exhast…and the like at each location with Node-Red and data visualize/remote control by EPIC.

Would this a better options?

Customer wants something simple but more reliable.


Its hard to answer “better”… There are a lot of variables that would need to be reviewed between you, the customer and the IT / Network guys.

Sounds like a HVAC application, so it may not necessarily be a high speed application, but there still might be some interlocks (fire / smoke) alarms and dampers that need to be managed that we don’t know about for example.
Node-RED can do a lot of stuff, but each AHU might be different enough to warrant a more mindful solution.

I was just trying to have a conversation about your first question ‘How many RIOs can be used as remote I/O per EPIC?’
The number of RIOs per EPIC is not hard and fast, it depends on a lot of things that you need to keep in mind as you scope out the system end to end.

Thanks again, Beno,

For example, each AHU will drive by VFD, the required signal are temperature input (4-20 mA TT), analog output to VFD (0-10 V or 4-20 mA), Start/ Stop signal, and 0-10 out for cold water valve regulation. The power circuit will use existing start/delta panel.

This is the most complicate part of each RIO to performs.

So, if this works then should be ok.
I’ve got experience with RIO. It doesn’t performs well if we give it many task but for simple control should be ok, I guess.

The networks will use existing or implement separately from existing (TBD).


For HVAC, I want local control 90% of the time - all control logic happens at the unit and is not depending on the network being functional. If using a RIO, that means writing the logic in Node-RED or something custom. Make sure the RIO has enough IO for your application - a lot of air handlers will have more points than the RIO can handle (outside air damper, duct static pressure, return air temp, outside air temp, return air fan, building static, etc.) R1 racks work well though - but of course cost more.

Keep in mind that no HVAC control companies that come in after you will know what the thing does - totally different world - and integrating with BACnet devices is a pain. Much more reliable than the HVAC stuff though - deal with it all the time - just make sure your customer is advocating for it and understands the benefits as well as the down-sides, as every other vendor will come in and talk it down.


Thank you, Phillip for a very god advice. :slight_smile:

The customer wants to keep it simple as same as the existing system. We will just make it more a bit remote controlable, schedule on/off, visualize and trend a for tracking. We are going to use each RIO for each AHU and maybe a group of FCU, ventilation and the like. The energy will use 3rd party to keep the budget minimum as we can. If the RIO’s I/O is not enough then I must consider the diagram below. How a bout the I/O with Modbus TCP/IP?

I heard the customer already break the relationship with some of thermostat vender and wanted to implement with our low cost integration service.

Best, (Thanks for advice)!

I think Philip really zoomed in on the key point here:

Do you have a firm I/O count of each type and expected signal for each AHU?
That must be where you start and must be done very soon.
We can’t really talk about Modbus till we know the hardware I/O count for each suggested RIO location.

1 Like

Thanks Beno,

I think :thinking:. what is the maximum capability of RIO that we will need just something like a below.

  • Digital output for enable VFD
  • Analog input for temp (4-20 mA)
  • Analog output for speed control (proportional to temp input)
  • Digital input/output (TBD) for start signal (interface 24 Vdc with existing 230 Vac control voltage, remotely…maybe will use the relay output from RIO for this🤔)…

I correct this a bit

  • One digital input for auto
  • Stop signal for manually stop (remotely)
  • Analog signal to regulate the cold water valve (0-10V)

I think, we will got 3 channels left for other purpose.

Kindly see the diagram, I try to draw from my head. It’s quite simple but honestly no time for a nice diagram.

Thanks for all helps. :slight_smile:

You are also missing 2 very important signals that will provide you alerts and reset the VFD.
A digital input reading the vfd fault indicator output, and a digital output to reset the VFD (Normally a dry contact). I have done this using Epic-PR1 because of the amount of signals and module density FOR 14 ahu with no issue at all, solid, building owner and HVAC contractor happy, no more nonsense calls, replaced a pneumatic system.
Many buildings owners are looking for solutions alike due HVAC manufacturers ridiculous software and license fees but more importantly, they are looking for reliable equipment.
Have worked with several HVAC companies and mechanics, all of them happy with OPTO, they just didn’t know of such equipment existed, I always do very simple screens for final users and more mechanical screens for techs.

1 Like

I agreed with Jelectron, but let’s find out the result after discuss with engineering director of 5+ stars hotel in Bangkok today. If enogh room for the budget then I would like to try your advice as well.


I think :thinking:, maybe I can use just one signal for start/stop "(Start Stop of one Motor from the same Push button PLC program) then will have 4 chs left. From @jelectron advice will need another 2 chs for VFD…so, I’ve got 2 chs left then…Thanks for advice @jelectron


You have several ways to do this and save, If you want temperature measurements I’ll recommend ICTD sensors instead of 4-20ma temperature transmitters (if you are using an EPIC the amount of channels per module makes it ideal) also you eliminate a possible failure point (sensor-transmitter) and ICTD comes factory calibrated, they are super reliable, over the years only one ICTD sensor died on me because a refrig. technician accidentally broke the sensor, they are simple to wire, not need of super special wire either. At the end you will have to plan EPIC vs many RIO’s or SNAP-R’s and wiring. Also consider the options you may have using Modbus TCP in the VFD, that may save you some i/o specially with the strat/stop (Hard wired start/stop is only one contact anyway in most VFD). BUT the VFD fail indicator and VFD reset dry contact work magic even having Modbus, power failures and blackouts may drive a VFD crazy and comms will get weird, having reset contact control on the UI or an auto reset timer will save you big headaches.

This is unusual. How are you proposing to control the CHWV? Fan speed is typically controlled using a duct static pressure transmitter as the input.

When using Opto equipment, these are the best to use. You can get them in a lot of different sensor configurations, such as duct probes, outside of Opto (Opto only sells the bullet style as far as I know). Look for A/592 sensor (ACI is a good source in the US). The wiring color is different on the non-Opto ones (Opto is black and white, and the ACI ones are red and black respectively, so easy to miswire)

Modbus is good at saving IO and costs, but consider the future replacement of the VFD may be more involved if not replaced with the same model/manufacturer. You can also use Modbus RTU with a serial module or USB to RS485 adapter - most VFDs will have this capability without any add-ons.

1 Like

I found a very useful for basic of HVAC.

Thank you Philip,
This hotel is about 20 yrs+ old. We cannot change much because the customer wanted to keep it as is. We will combine with the existing and make it remote controllable. To monitor temp and AHU status so, it’s a only way that I must split signal from existing sensors to RIO and use a selector to keep the existing system as is.

So, we will do pretty much the same as Diesel pump with the RIO.

Yeah, I agreed to use Modbus but the only Modbus RTU is available from some of ABB VFD. It’s already my choice after get conclusion and see which way to go.

Thanks for advice.

Thanks for advice.
We must use the existing RTD,4-20 mA output to keep the existing system as is.
Just only Modbus RTU available from some of ABB VFD.

I try to conclude this topic because my project scope was changed by customer and I found below that I cannot understand. Could you please explain?

[quote=“Beno, post:2, topic:4788”]
While there is no real hard limit (beyond network addressing) placed on the number of remote I/O there will be a network traffic and human sanity limit that is reached at some point.
[/quote] :thinking:


My comment was based on the original question:

Opto does not code one in.
How many you want to add depends on the work load for each, the network, the control engine used (PAC or Codesys), the graphics used etc etc etc.

Then you have the programmer experience (ie, the hinted sanity level).
You might be comfortable using pointers, others may not. So the code structure is going is going to do your head in, or do the next guys head in a scrambled mess of keeping track of … what… 30 RIOs?, 50? Perhaps you can keep track of 100…?? but as I said, there are just so many variables of exactly how many RIOs you can hang off one EPIC its hard to give a hard numbered answer.



Please see below.

I think I will choose CodeSys where I can adjust around then my system is more flexibility both technically/budgetary.

Thanks, Suchet