5069-IB16 inputs not turning on: the SA return, the 1 ms filter, and what a green MOD light won't tell you

The picture is almost always the same. The 5069-IB16 sits in the rack with a solid green MOD light, no faults in the controller, 24 V on the terminal, and the input light stays dark. Most of the time the card is fine. What's missing is the path the input current takes back home, or a filter doing exactly its job while a drive dumps noise into the wiring. Below is how the 5069-IB16 wants to be wired, what its lights can and can't show you, and what I check on a surplus one before it goes in a rack.

Where the current goes back on a 5069-IB16

Sixteen sinking inputs, one group. Current flows into the terminal, so whatever drives the input has to switch the positive side: a PNP sensor, or a dry contact fed from +24 V. An NPN sensor pulls toward 0 V and this card will never see it. More on that in sinking vs sourcing on Allen-Bradley I/O.

Now the terminal block. Eighteen positions: I00 through I15, plus two at the bottom marked No Connect. No COM, no L+ or L-. Anyone coming from 1769 I/O, or from a 5069-OB16 with its LA terminals, goes hunting for a common that isn't there.

The return is internal. Every input on a 5069-IB16 goes back through the module circuitry to SA(-), which lives on the SA power terminal block of whatever starts that SA bus: the CompactLogix 5380 controller, a 5069-AENTR or 5069-AEN2TR adapter, or a 5069-FPD. SA then rides down the bus through each module's side connectors.

That one fact explains most of the dead inputs I get asked about.

Say the sensors run off a second 24 V supply, and nobody tied its 0 V to SA(-). The switch closes, the meter reads 24 V from the terminal to that supply's 0 V, and the input stays off, because the current has no way back. The reading is real, it just references the wrong point. Put the black lead on SA(-) instead. If the voltage disappears, you've found it: tie the field supply's 0 V to SA(-), or feed the devices from the SA supply.

A few more power things that bite:

·       SA has to be DC, 10-32 V. If the controller's SA terminals carry 120 V AC for a row of AC cards, this module needs its own DC SA bus, which means a 5069-FPD in front of it. Placement and the 10 A budget are in the 5069-FPD write-up.

·       Each card draws 75 mA of MOD power at 18-32 V DC and 200 mA of SA. Neither bus may pass more than 10 A.

·       Separate supplies for MOD and SA are good practice, since losing one doesn't take the other down. They also set up the classic confusion: SA dies, MOD stays up, and the module sits there green and happy, reading 16 zeros.

If the field devices run off a second supply, its 0 V has to be tied to SA(-), or the inputs never turn on.

Green MOD, dark inputs: the order I check things in

Lights first, and the MOD light is the one that actually tells you something.

MOD light	What it means	What I do next
Off	No power reaching the module	Check MOD power and that the module is seated on the bus
Steady green	Connected to its owner-controller, normal	Leave the project alone, go to the field side
Flashing green	No connection: configuration missing, incomplete or wrong	Check the I/O tree, the slot and the keying
Flashing red	Firmware update running or failed, recoverable fault, or connection timeout	Cycle power or reset from Module Info, then check the Connection page
Steady red	Nonrecoverable fault	Cycle power; still red, replace the card

 

The 16 point lights tell you less than they seem to. On a plain 5069-IB16 a point is off (input off, or no module power) or steady yellow (input on). No red state. Flashing red for field power loss exists only on the IB16F, so a card that has lost SA can't say so. It just shows dark points, same as 16 open switches.

The order I go in:

1.    MOD steady green? Then the connection and the keying are fine. Put the laptop down.

2.    Measure SA+ to SA(-) at the terminal block that feeds this bus. Anything from 10 to 32 V DC is in range.

3.    Jumper SA+ to I00 with a short piece of wire. Yellow means the input circuit, the bus and the return are good, and the problem is in the field wiring or the device. No light means the card isn't getting usable DC SA (an AC bus with no FPD, a damaged side connector) or the module is bad.

4.    Measure the field side from the input terminal to SA(-), never to the nearest 0 V bar. The input needs at least 10 V to turn on and draws about 6 mA at 24 V.

5.    Stuck on instead of off? A 2-wire sensor that leaks more than 1.5 mA when off, or leaves more than 5 V on the terminal, is past what the card guarantees as off.

About that jumper: landing a wire on a live terminal can arc, so in a hazardous location kill power first. And anywhere, don't pull the terminal block or the module live to "reset" it. Neither is rated for removal and insertion under power, and a broken backplane connection can cause unexpected machine motion.

The input that freezes while the drive runs

This one fools good people. The sensor works, the input follows it, then a conveyor drive starts and the bit stops following anything. It sits in its last state. Stop the drive and it all comes back. Look first for an unshielded motor lead sharing a duct with the input wiring.

Inside the card, every 5069-IB16 input goes through a 50 µs hardware delay plus a filter time set on the Points page of Module Properties. Default 1 ms, range 0 µs to 50 ms, and it works on both edges, Off to On and On to Off. On the fast IB16F, and on IB16 cards running firmware 2.011 or earlier, a transition only counts if the signal holds its new state for the entire filter time. Newer IB16 firmware got a reworked filter, so the exact timing on a newer card can differ. Where the strict rule applies, chop the signal with noise every few hundred microseconds and no state ever lasts 1 ms. Nothing qualifies, so the bit stays where it was.

The card isn't broken. The filter is working on a signal that is mostly garbage.

With the strict filter rule, noise that keeps chopping the signal leaves the input stuck in its last state.

You'll hear the advice to set every filter to 0 µs. It does get the frozen bit moving again. I use it as a test and nothing more: if dropping the filter changes the behavior, you've proven there's noise on that wire. Leave it at 0 and any spike longer than the 60 µs minimum pulse width lands straight in your logic. A longer filter buys noise immunity, a shorter one gives it away.

The real fix is in the wiring:

·       Keep control wiring at least 0.3 m (1 ft) from motor and power leads.

·       Where a signal run has to cross a motor lead, cross at 90 degrees.

·       Run the motor in shielded cable with a braided shield of at least 75% coverage, and actually land that shield instead of cutting it back.

·       Shielded cable on the input runs helps too.

The VFD cable article covers when shielded motor cable is worth paying for.

5069-IB16, IB16K or IB16F: what actually differs

Three cards share the same housing, and one letter in the catalog number is the whole difference at a glance. A lot is shared: 16 sinking inputs in one group, on at 10-32 V DC (4 mA at 10 V up to 7.4 mA at 32 V, 6 mA at 24 V), off at 5 V and 1.5 mA or less, 75 mA MOD and 200 mA SA draw. The differences:

	5069-IB16	5069-IB16K	5069-IB16F
Input delay, screw to backplane	100 µs max	100 µs max	10 µs max
Minimum pulse width	60 µs	60 µs	6 µs
Default input filter	1 ms	1 ms	0 µs
Counters	up to 8, 500 Hz	up to 8, 500 Hz	up to 8, 30 kHz
Timestamps, events, pulse latching, pattern matching	No	No	Yes
Field power loss detection	No	No	Yes
Conformal coating	No	Yes	No

 

The K is the plain card with conformal coating for dirty or damp air. The F carries the extras, and it's the only one of the three that flashes a point red when field power drops.

Counters are where people get caught. A 5069-IB16 runs up to eight counters at 500 Hz, and they take terminals from I00 up: set four counters and only points 4 through 15 are left as plain inputs. Do the math before you hang an encoder on it. A 100 PPR encoder at 300 RPM is already 500 pulses per second, right at the ceiling. Anything faster goes on an IB16F.

Moving a 1769-IQ16 over to a 5069-IB16

The 1769-IQ16 splits its 16 inputs into two groups of eight, each with its own DC COM terminal, and each group is sinking or sourcing depending on where that COM is tied. For the sinking case, the listed Compact 5000 replacement is the IB16. Its single group only sinks, and there's no COM to land anything on.

So before ordering, look at where both COMs on the old card went. If a COM was on DC negative, that group was sinking, and the IB16 takes those devices as they are. Find one tied to DC positive and you've got a sourcing group running NPN sensors, which the new card will never see. For that group the listed option is the 5069-IV16F-SC, which comes from Spectrum Controls rather than Rockwell. One group of each means both modules, or new sensors.

Every conductor gets moved by hand, so check each one reaches its new terminal and replace the ones that come up short.

Checking a surplus 5069-IB16 before it goes in the rack

·       Read the catalog number on the module label, not on the box, and check the last letter. IB16, IB16K and IB16F are easy to mix up.

·       Plan on a terminal block, because the module ships without one: a 5069-RTB18-SCREW or 5069-RTB18-SPRING. 22-16 AWG, one wire per terminal, strip 12 mm for screw and 10 mm for spring, screw torque 0.4 N·m.

·       Look at the side bus connectors before snapping it onto a neighbor. A bent contact there can stop MOD or SA from reaching the next module.

·       Match the keying. Compatible Module wants the same catalog number and a major revision equal to or higher than the one configured, so an IB16F in an IB16 slot means changing the module in the project first. Disable Keying exists. I don't use it.

·       If the firmware needs an update, get it from Rockwell's own download site and nowhere else.

·       Bench it with a controller or adapter and a throwaway project until MOD goes steady green, then jumper SA+ to each of the 16 inputs and watch every point go yellow and every Data tag go to 1.

Both RTB18 styles take 22-16 AWG, one wire per terminal. Switch MOD and SA power off before wiring.

FAQ

Does the 5069-IB16 need a common wire?

No. Its terminal block has no COM. The inputs return inside the module to SA(-) of the bus they sit on, so a separate field supply needs its 0 V tied to SA(-).

Will a 5069-IB16 work with NPN sensors?

It won't. The card only sinks, so it needs PNP sensors or contacts switched from +24 V. For NPN devices the sourcing option is the 5069-IV16F-SC from Spectrum Controls.

What is the default input filter on the 5069-IB16?

1 ms, adjustable from 0 µs to 50 ms on the Points page. The IB16K also defaults to 1 ms, the IB16F to 0 µs.

Why is there no red light when SA power is lost?

The plain card can't detect it. Field power loss detection, and the flashing red point that goes with it, exist only on the IB16F. An IB16 with dead SA shows a green MOD light and 16 dark points.

Can a 5069-IB16K replace a 5069-IB16?

Electrically, yes: same ratings and the same firmware package, plus conformal coating.

When it really is the card

If MOD is steady green, SA checks out at the terminal block feeding the bus, a DC card further down the same bus works fine, and a jumper from SA+ still lights nothing, the card is done. We carry the 5069-IB16, the coated 5069-IB16K, and the 5069-IB16F for jobs that need counters above 500 Hz or a card that reports field power loss. The module ships without a terminal block, so the 5069-RTB18-SCREW and 5069-RTB18-SPRING are listed on their own. And if you'd rather keep a 1769 line running than migrate it, the 1769-IQ16 is listed too. Everything ships with our 2-year warranty.