Barco ICMP GPO port to drive relay

[Marin Zorica]

Hi all!

I have encounter few situation to drive dc relay direct via GPO port. Pinout, diagram etc is clear to me and also available from manual. When I try to drive relay, it cannot be triggered.....since some relays have built in led for indication, led is lit but still relay cannot "pull". It seems like there is current limit resistor on GPO port, and current is ok for led to lit, but still not enought for relay to pull......even if is some small relay. It is same either GPO on ICMP or GPO on SP2K direct. To be more sure, I've have many sites running doremi showvault with IMB, and there were some automation where relays were driven direct from IMB GPO. After projector was replaced, same relay did not work either with GPO from ICMP or projector direct....same thing, led is bit lit, but relay cannot pull, while it was working with IMB prior to replacement.

I think it was 12V relays, so maybe 24V would pull, because 24V relay would need less current to pull!? Maybe they either put some current limit resistors on optocoupler (which i did not saw on ICMP pcb) or they do not drive optocoupler to full saturation!? I have solve this by adding ULN2003 IC to drive relay between GPO and relay and it works without problem.

Strange is, because they use TLP291 optocoupler which is rated up to 50mA and 80V on output, and that should be just ok to drive small relays, but in my few cases it didn't and also I have call from some collegues which had same issue. Did any of you encountered same problem or I am doing something wrong?

[Steve Guttag]

From the manual (page 38 on version 00)

• Maximum output driving voltage : Vmax = 18 V

• Maximum current : Imax = 30 mA

• Maximum power dissipation : 120 mW

So, that is a "No" on 24V and also a "No" on going above 30mA

Using Omron as my reference, typical coil current is between 41.7mA (G5V, DIP style) and 75mA (LY2-DC12),

On their G5V "H" series for low coil current, you do get to magical 30mA. Their G5V "H1" series the coil current drops to 12.5mA, which is probably where I would go if I was driving it directly from the server.

Reed relays should also be in a similar current range (10mA, give or take).

Clearly, Barco wasn't planning on the server driving things like relays directly and would rather drive something with typical TTL or CMOS logic type inputs.

[Marin Zorica]

Interesting.....on SP2K-C series, from the manual, it say 24V 30mA, 120mW.......so using small relays with 10mA should work, for example I have some Zettler AZ6991 which are under that...... I did miss part where they mentioned ICMP-X GPO ratings, but this is from the projector......and result is same, plus, projector uses same optocoupler IC

[Harold Hallikainen]

There are some amazing solid state relays available now. These use back to back opto FETs so they can handle AC or DC of either polarity. The LED current is quite low. Here are some part numbers I've been using:

G3VM-41AY1	40 V	2 A
G3VM-61AY1	60 V	500 mA
G3VM-201AY1	200 V	250 mA
G3VM-351AY1	350 V	100 mA
G3VM-401AY1	400 V	120 mA
G3VM-601AY1	600 V	90 mA

[Carsten Kurz]

Did Barco solve the sporadic GPO triggers during boot on the ICMP with some software upgrade? When I was playing with this, I bought one of these cheap 8ch relay-boards from ebay, they use optocouplers on the inputs.

image.png

[Marin Zorica]

There are many board you can use for it. I do not use it, except for things like fan, audio preset on old processor or so. Other things are done either via my oem ip automation or jnior, sometimes plc, amx, crestron depends on site. I remember last time, a year ago i set up one for fan, and during boot fan would run and stop, so back them was not solved, as projector was on latest firmware.....

[Jerry Bian]

Did Barco solve the sporadic GPO triggers during boot on the ICMP with some software upgrade? When I was playing with this, I bought one of these cheap 8ch relay-boards from ebay, they use optocouplers on the inputs.

image.png

Yes, the issue of automatic GPIO opening and closing has been resolved in versions later than a specific one

[Marcel Birgelen]

Interesting.....on SP2K-C series, from the manual, it say 24V 30mA, 120mW.......so using small relays with 10mA should work, for example I have some Zettler AZ6991 which are under that...... I did miss part where they mentioned ICMP-X GPO ratings, but this is from the projector......and result is same, plus, projector uses same optocoupler IC

Keep in mind that the ICMP needs to pull all the power from the backplane's power bus, so I guess that's where the GPIO limitations on the ICMP come from.

I've always used some opto-coupled intermediate relays with the GPIOs, simply to have some galvanic isolation between the very expensive DCI gear and all the ghetto GPIO driven stuff out there. It wouldn't be the first time some external electrical engineer ends up putting 230V/AC onto GPIO circuits...

[Marin Zorica]

Keep in mind that the ICMP needs to pull all the power from the backplane's power bus, so I guess that's where the GPIO limitations on the ICMP come from.

I've always used some opto-coupled intermediate relays with the GPIOs, simply to have some galvanic isolation between the very expensive DCI gear and all the ghetto GPIO driven stuff out there. It wouldn't be the first time some external electrical engineer ends up putting 230V/AC onto GPIO circuits...

GPO port, as well GPI is actual optpcoupler itself. So there is no power to pull from anywhere, it is just mater of output side of tranzistor is conducting or no. But, even if is under spec, on new sp series port current is limited even more than specified!

[Marcel Birgelen]

There might be optocouplers and other protection circuitry in place inside both the projector and the IMB, but for the couple of bucks it costs me to put some extra layer of isolation between someone else's high-powered stuff and a potential $10K+ ticket item, I'm not going to cheap out.

[Bruce Cloutier]

There are some other advantages to the JNIOR.

First there is the logging. The IOLOG can tell you precisely when a relay closed and opened. So if the rumor is that something didn't seem to go as planned you can look back to confirm.

The inputs are isolated. So if you tied them one on one with the GPO the expensive equipment is protected. You can slave relays to inputs and while that seems ridiculously trivial, there is a lot of protection afforded and you do have the logging. You can also interpose logic. That might be something as simple as enforcing minimum on and off times for the relay. That is something critical in the HVAC industry. That would keep things stable so if the ICMP reboots or glitches the lights/audio/fans can ride through the event in a professional manner.

The JNIOR lets you connect remotely and monitor the automation. Techs don't take advantage of this. From the command line (Telnet or SSH) you can use the JRMON command to watch digital inputs and relay outputs in real time (not to mention manually control). If your automation is more complex the NETSTAT -S command provides a sniffer and you can watch in real time your equipment communicate with the JNIOR over the network. Even when you are using the AUX serial port you have the IOLOG to look back on to see if it sent what you expected to be sent and when. The IOLOG doesn't have a real time monitoring mode but I can easily be persuaded to make it so.

Of course, the JNIOR can do several different things simultaneously. You could add a temperature sensor and easily enable a relay to kick on booth exhaust for instance. Those of you with JNIORs now can add that easily. Or, you can just have the JNIOR send you an email that your sauna is ready. Well, I guess not all JNIORs have access to the outside email server.

Then there is the whole thing in not having a random bare circuit board hanging with a wire mess off the back of the ICMP. So when you leave your work behind, the next person doesn't panic. But we've seen some pretty scary JNIOR installs. Nothing is perfect.

And the JNIOR does come with some technical support.

I know that there is a monetary difference. And there is a pretty valuable educational aspect to those inexpensive relay boards (and Arduino processors).

[Ioannis Syrogiannis]

After I liked the commend before me, my two cents:
Last I needed to use a relay to open/close door magnets with an ICMP-X, I made sure to power the relays separately, not to rely on the GPOs' current. I chose 5V powered ones, so I could turn to USB, if I had to troubleshoot power supply.
It went well, and -temporarily- so did when I connected in parallel the tungsten lights used for cleaning and emergencies (that is 220V) on the other side.
So, if you have the option, go for powered relays.

Powering on the LEDs but not switching the relay on seems really weird to me. I would imagine that the same circuit closes for both of them.

[Marin Zorica]

After I liked the commend before me, my two cents:
Last I needed to use a relay to open/close door magnets with an ICMP-X, I made sure to power the relays separately, not to rely on the GPOs' current. I chose 5V powered ones, so I could turn to USB, if I had to troubleshoot power supply.
It went well, and -temporarily- so did when I connected in parallel the tungsten lights used for cleaning and emergencies (that is 220V) on the other side.
So, if you have the option, go for powered relays.

Powering on the LEDs but not switching the relay on seems really weird to me. I would imagine that the same circuit closes for both of them.

GPO is not powering anything, it is just collector-emitter side of transistor which goes to saturation when GPO is on, eg it should conduct. Led or relay are powered from external power supply, and there is seems like current is limiter because led can lit, but relay which needs 15mA will not pull. It was strange for me because on doremi IMB it was working without problem.

[Ioannis Syrogiannis]

I don't quite follow you, Marin. The relay needs 15mA from where, if GPO is not powering anything?
Two more cents: If you haven't yet, try to reverse polarity of the GPO to the relay.

[Bruce Cloutier]

A point worth mentioning regarding the OP is that relay coils are an inductive load. If you switch those and do not provide back EMF protection of some kind, circuits will fail. Even the small signal relays should be addressed. So if you want GPO optoisolators to last, you need something. Typically a diode in a reverse orientation across the relay coil. When the relay is being activated that diode won't conduct. It is there so when you break the circuit and energy stored in the coil needs to disapate, it doesn't pierce the drive transistor in the opto (or arc someplace). It just winds down through the diode. The GPO should have those designed in but, no.

Automotive relays have snubbers built in so they don't affect your quality AM radio listening experience. Nothing is small signal in that category.

That is even a concern when your small signal relays then drive more serious relays. If you don't deal with the back EMF there will be an arcs at the relay contacts. Eventually they will fail.

I believe that the back EMF diodes are clearly visible on the 8 relay board between the relay and the opto isolator.

Also if you place an LED in parallel with the optoisolator input (looking for an indicator), if there is enough of a threshold difference between the LEDs, the external LED can prevent the opto from triggering. Similarly if you put the LED in series you won't get exactly what you might expect either.

The relay LEDs on the JNIOR are switched by one side of the DPDT relay. So if the LED lights, then the relay has physically activated. If at that point your ohm meter doesn't show nearly 0 ohms across the relay contacts, then you've burnt out the relay. It happens.