FP TC-120 readings freeze after about 30 minutes

[TRCErin]

I have a very simple VI that controls heaters in a chunk of steel.   The steel has four thermocouples(one for control and three for monitor).

The VI loops and looks at the setpoint and the control TC reading.  A boolean determines if heater is turned off and on.

Everything runs great for awhile (usually less than an hour) and then all the TC's "freeze" into a reading that is just slightly higher than the setpoint.  Consequently the heater stays off and everything cools to room temperature.

I'm using a TB-3 and V/C are not hooked up.   This problem happens if one thermocouple or all of them are hooked up, it makes no difference.

I'm at a loss for ideas.  I do very similar things with other TC-120s at my facility, but I can't swap them since they are all currently in use for testing.

Any ideas?

[TRCErin]

Here's some more information.   Once this "freeze" occurs, I go to MAX  and look at TC-120 in Data Neighborhood and it says Bad Channel status "Unconfigured" reported for all channels.  When I started the application , MAX reported the channels as "Successful"

All the other devices (RLY-420, AIO-610) show "Successful" in MAX  both before and after the "Freeze" occurs.


Sorry for the double post!

[tst]

Hello TRCErin

We have a similar problem with cFP-AI-110 which freeze all channles, but after a longer, and not constant, time.

In MAX we read the same error, but it blinks from "unconfigured" to "successful" all the time, until a reboot of the unit.

That problem stays also after replacing all the hardware with a new one.

We recently improved the issue (the problem return after a longer time), by removing all power supplies from the cFP area,

but we not really know if the too high temperature made the problem, the Electricity noise from the suuliers, or Electricity shocks from 3 phase contactors, which the controler operate.

 (we use the recommended "flay back diodes" on the outputs, and operated a small 24 vdc relay to operate the contactor.)

In summary, we have some ideas, but looking with you, for an authorized answer from NI.

[TRCErin]

I've now isolated the "freeze" to initiate when the contacts for the heater control relay close.    This occurs when the process reaches the setppint.   Apprarently there is a voltage spike or something occurring that the TC-120 doesn't like.   It gets its power from the BUS so I'm surprised that the other FP modules aren't doing the same thing.     However, I do have the TC-120 at the end of the BUS strip.

Does this sound logical?

[Grub]

TRCE,

Fieldpoint modules are isolated from one another, so a spike on one will probably not affect the other.  You, by chance, don't happen to have a themocouple extension wire connected to the bank that isn't connected on the other end with a probe?  I have seen, especially around highly inductive loads (heaters coils), where an unterminated tc wire acts as an antennae and transmits noise to the Daq device.  If this is the case, try removing these wires or terminating them with a tc probe.

Hope this helps,

Grub

[TRCErin]

Thanks for all the tips....I think I got the problem solved.   We were using a mechanical relay for the heater control and believe this was causing a voltage spike when the contacts opened.    The TC-120 didn't like this.     Today I switched to a solid state relay and the "freezing" problem has not reoccurred.

Just don't ask why we had a mechanical in there in the first place  

[jfmorneau]

FINALLY!!!  Thanks to you guys, we finally fixed our problem!! We were experiencing the exact same problem as described here... This problem was haunting us for the last 5 months...

In our case, 24V solenoids valves were causing the problem... We fixed it by installing a diode and a capacitor directly on the valves coil to prevent voltage return.

Seems like Field Point modules have a weakness.. Hope NI will pay some attention to that problem to prevent this.

Thanks to all of you!

[Conseils]

In any electrical system it is necessary to engineer all aspects of the circuit and possible electrical interference between equipment.

Not fitting the diode was the problem, not the Filedpoint module per say. The Flyback diode is a fairly standard requirement; I hope that you have selected the appropriate size, otherwise they will probably fail randomly in the future and no one will be around to know what's going wrong. Of course they could just be way too big.

The lack of suppresion would have been just as likely to trash completly somebody elses system.

Just so you know; I have FieldPoint modules in panels with hundreds of 24V relays, solenoids and fitted within feet / meters of 90KW motors and associted switch gear.

The thing to ask yourself now is.... In what other ways have we failed to correctly manage the system e.g. screening, earthing, electrical noise suppression (ferite rings) etc.....?

My bet is that you used the same power supply for the FieldPoint unit and the relays...?

Wow it's amazing you didn't trash the lot with the spikes that were thrown into the system. I would put money on random early failures. System engineering takes a bit of time and effort but its generally worth it, It will also help reduce noise on you thermocouple channels / analogue inputs.