Network 8000 GCS Question

[RC20]

I could find no info running a search so.......

I am working on a DI signal issue into a GCS module.

Two separate DI inputs are not showing as closed, confirmed the switch is closed, xformer working.

I am baffled by the lack of a common (or one I can identify) from the xformer to connect to the module as well.

I can go from the common side of the xformer and read 19 volts on the input, but no ref seems to be connected directly to any point.

I can do the same on another GCS that I know has the same two switches closed and read 24 volts, but again, using the xfomer common side for reference, nothing in the GCS connections gives that.

Two Questions:

1. Bad points or what would cause the GCS not to see the signal ? (related to the 19 volt level I think but.....) Seems odd two points would go bad.

2. How does a GCS references a common or interpret a 24 volt DI signal without it if that is not hooked tied into the module?

I am missing something obviously but baffled as to what.

Will try to check back before day is done, but emails good for reply

[jurban10]

A Digital Input is simply looking for a closed set of dry contacts. Short the DI at the GCS and it should show closed. If it dosn't then the GCS is the problem. If it does then the device or wiring is the problem. There should be no voltage introduced to a DI on a GCS.

[RC20]

Got it on the dry contact.

There is no wire from the common that’s between the points (7-5 and 7-6) to the switches.

They ran 24 volts to the common terminal on the switches (float system) that has the two switches, then ran it back to the GCS to points 7-5 and 7-6.

I know that’s totally weird and against everything I have ever seen.

Its worked for 15 years, and there are 24 of these in 7 GCS modules in the same area.

As its fuel tank refill setup for 6 tanks so the whole thing gets dicey.

Am I correct in that a normal setup would have a wire from C terminal to the switches, they would then close and feed the signal back to the GCS to the 7-5 and 706 points (if they were closed)?

Not the first time I have seen someone do something that should not work and did, the quite working down the road.

Just need to fully understand it as its spills over (fuel pun) onto the rest of them.

In this case I would assume I could remove the wires, jump the common to 7-5 and have that register an on input.

[Control Man]

I have had a few GCS's ( doing Net 8000 since 1986 ) with DI's wired like that and it was to save on wire .

18/3 TSP used White Com for PT1 ( small jumper to 2nd DI device in this case AIR FLOW switches ) , Black to PT1 , RED to PT 2 in the GCS.

The resistance on DRY CONTACT signal back to GCS / LCM should be 800 OHMS or less and wiring should never be over 1000 '

PC1 PC2 on the GCS / LCM can wired like this for PULSE imputs as well

[RC20]

I assume you have not seen a 24 VAC xformer wired per the description ? Seems like it makes things register in the GSC even if no reference (no idea what the GCS voltage to its dry contact using its normal internal setup ( DI C to Pt 5 for example is).

So being perfectly clear.

If I remove existing wires put jumper from the C terminal associated with Points 5 and 6, and jump C to 5 it would register as a closed contact? I.e. duplicating the current switch position.

I know I am being overly repetitive, but sans an explanation why they used the setup with the 24 system, I am being ultra cautious.

[jurban10]

So this is two seperate switches on a float assy. perhaps indicating 2 levels in the tank like high and low or fill and stop?

I have seen this. Keep in mind that the common terminal on a GCS input is the supply voltage for the completed circuit. I thought this was a 5.5 VDC signal. I know the maximum short circuit across the input is 10mA. So nothing on the common and inject current from an external source on the other terminal and it should read as a completed circuit id everything is properly drained(grounded).


If you use the method of shorting the 7-5 or 7-6 to common to test the input make sure there are no wires carrying voltage connected to any of them. This will most certainly damage the board.

I don't know why someone would go through the extra expence of the external source voltage unless it is a distance thing. The contact resistance for a closed DI on a GCS is 800 Ohms or less. 1000 ft of wire is the max.

[RC20]

Correct for float levels in a tank. Only two of them I am currently (pun) concerned with, but there are a total of 4. The two I am dealing with that are causing the immediate problem are the Full and Fill. Other two that need to work are the Low Alarm Level and the Overfull (pretty safe as we have visual gauges which is why this was caught on the auto fill not working, and overfull returns to the source tank).

The transformer is not grounded.

Distance to the furthest tank and the one that is affected is 50 ft, so add another 50 feet for routing wire runs and for a total of 100.

I am just starting in on this, but I believe they used the SAME transformer for all 6 tanks.

For test purposes to see if the input is working, I would remove the wires coming back from the tank to Points 5 and 6, and then jumper one at a time.

If that registers as a closed input(s), then have to consider correcting wiring.

Let me know if you think I am of base, but it almost seems like they used a bloom affect in that if you put enough voltage from the 24 volt xfomer to the input via a switch closure, it would register as if you ran a common to the switches common side and the rest was per a normal wire return.

If that works (at least until something goes a bit off ) then they would have saved quite a bit of wire as each set of switches (two per tank to use the common between the inputs) would have had to have two more wires running to each tank.

That would have been a total of 12 more wire to the 6 tanks, instead of the single one from the xformer.

Is it possible to use just run one common to the float assembly and have it work or does each one need its own common for a pair of inputs?

[RC20]

I should add this was a last minute change in the project, so it was done on the fly and not part of the original setup and layout.

[jurban10]

I think you would be taxing the VA of the GCS. Better to run a pair from each switch to the input terminals. I think trying to use one common from the GCS to supply all the switches will lead to false or intermittant reads down the road. It's a repair, go for the gusto! The results will be better then the original install and you'll be illiminating one component from possible failure. The Xformer.

I agree though that is seems that the entire design was a value engineer.