T100MD2424+ OUTPUT DIODES NEEDED FOR SOLENOIDS??

[cch1955]

Oh great one:

I have a machine which uses 24vdc solenoids sourcing current directly through the outputs on the T100MD2424+. It has worked well for many months, but has recently developed very rare output transitions. The only items being effected are the outputs driving the solenoids.

My thoughts are back EMF problems killing the chip or tripping it (just swapped it out so will see if helps).

Second the machine handles plastic tubing and can generate static build up. Possibly the static is causing this, but i would think I would see other problems too. I have added static discharge brushes to drain this off to ground.

The only way the solenoids change stae is through a CuFunction which toggles the output bit each time the CuFunction is executed. It is only executed by a manual foot switch with 1 second time eliminating false foot pedal trips.

Ideas, comments, great wisdom

Thanks
CCH

[support]

What do you mean by rare output transition? Do you mean the output turned ON or OFF when they are not supposed to? If you connect the computer and observe online monitoring, does the online screen correspond to the physical state of these affected outputs?

You mentioned sourcing current - I assume you are using the two 10A outputs (7 & on the T100MD2424+? These two outputs are controlled by the IRL9530 MOSFET. There is an intrinsic diode in each MOSFET driver that clamps the backemf to no more than 100V. However, for high inductive load current it is still better to use a diode to clamp the backemf as it is possible that the high avalanche voltage and high current may affect the other electronic components.  However, if the online monitoring shows the same output state as the physical output then it is not a case of inductive noise but more likely a logic problem.  I know you said the program has worked well for many months but sometime certain bugs don't show up until I/O condition has changed.

[cch1955]

Ah you point out my poor use of words. I am using regular outputs as SINKING outputs. My mistake. The other side of the solenoid coil is tied to +24vdc on same PS as PLC. I will try to see if states are same with online monitor, but hard to catch.

Rare output means when the machine is coming to the end of the cycle it will infrequently  and without pattern (once in 40 times as an example with no setting changes) cause the output which have state changed through a toggleio in CuFunction to change state and stay in the changed state. The only pattern which seems more common is that the bit changes from low to high. It has done the opposite, but not often.

The only section of code to change solenoid bit state is as follows:

'Toggles i/o 9 to change clamp state. for supply spool clamp 1. A low output=clamped so if have bad solenoid or bad wire or loose power , part stays clamped.

'SOL_SU1_CLMP IS THE SOLENOID

TOGGLEIO SOLSU1_CLP         'CHANGE OUTPUT BIT FROM CURRENT SETTING EACH TIME FOOT SWITCH  IS PRESSED.

IF TESTIO (SOLSU1_CLP) = 0 THEN   'Check to see SOLSI1_CLP current state after toggle i/o bit

   setio Su1clamped      'i/o bit is OFF to SOLSU1_CLP, clamp flag relay set indicating SU1 clamped.
ELSE
   CLRIO Su1clamped      'i/o bit is ON to SOLSU1_CLP, clamp flag relay cleared indcating SU1 not clamped.
ENDIF

The only way the above section of code is executed is by the pressing of a manual foot switch. The machine has four of these clamps and all have the same problem. All are on adjacent outputs.

I swapped the output IC as a test, same results. Added static brushes to ground product same results.

Ideas?

[cch1955]

Also solenoid specification is 24VDC unit drawing 85mA for both inrush and hold. Machine has 6 solenoids on outputs 9,10,11,12,13,14 on the T100MD2424+ main board.

Thanks
CCH

[support]

Again online monitoring is the best way to find out if the physical output correspond to what's in the CPU. If they are the same, then the problem lies in the software. Of course software may change the bit because it is affected by an input that has been false triggered. But we know it is not the PLC hardware issue.