QUESTION: What signal does the BARANI FARS fan-speed cable provide?

ANSWER: The FARS MeteoShield Pro provides a two-wire, polarity-independent, normally-open reed-switch output. A magnet on the rotating fan closes the contact once during every revolution.

According to the FARS MeteoShield Pro data sheet :

The output is one pulse per revolution.
1 Hz corresponds to 60 rpm.
The contact is normally open.
The contact is in series with an internal 24 Ω resistor.
The normal fan-speed range is approximately 0–2500 rpm.
The stated maximum output speed is 2600 rpm.

RPM = frequency in Hz × 60 ÷ pulses per revolution

Because the FARS produces one pulse per revolution, its conversion simplifies to:

FARS RPM = frequency in Hz × 60

QUESTION: Is the reed contact open or closed when no magnet is present?

ANSWER: It is open when the magnet is absent and closes while the rotating magnet passes through the reed switch's actuation field.

When connected between an SDM-IO16 input and ground:

An open contact is read as logic state 1.
A closed contact is read as logic state 0.

Campbell command 92, included in the diagnostic portion of the program below, reads the present state of all 16 SDM-IO16 inputs into 16 separate variables. This is useful for a stationary test: stop the fan, hold the magnet over the reed switch, and confirm that the indicated state changes from 1 to 0.

Static state testing is not a high-speed pulse test. A reed switch can pass a slow continuity test while still producing a pulse that is too short, noisy, or intermittent for the data logger at full fan speed.

QUESTION: How should the two FARS output wires be connected?

ANSWER: For the example program, connect the FARS output to SDM-IO16 port 1:

Connect either FARS output lead to SDM-IO16 input 1.
Connect the other FARS output lead to an SDM-IO16 G terminal.
Set the SDM-IO16 hardware address to 1, matching the program.

The two FARS leads have no polarity. Do not apply a separate excitation voltage to the FARS cable when using the SDM-IO16 internal pull-up.

Electrical-rating distinction The SDM-IO16 input is internally pulled up to approximately 5 V through 100 kΩ, so the closed-contact current is only about 50 µA. This is below the FARS absolute maximum switching current, and 5 V is below its absolute maximum switching voltage. However, BARANI lists a lower preferred switching voltage for maximum contact life. Installations that must remain strictly inside that preferred long-life voltage envelope should use a suitable low-voltage interface rather than connecting the contact directly to the 5 V pull-up.

QUESTION: Does the Campbell program calculate speed from duty cycle?

ANSWER: No. Campbell SDMIO16 commands 24 through 46 measure frequency. Duty-cycle measurements use commands 47 through 69.

The original instruction:

SDMIO16(Freq(),Ret_Code_1,1,46,0,0,0,0,30,0)

does measure frequency, but the final 30 multiplies every measured frequency by 30. It is not a duty-cycle setting.

A multiplier of 30 is correct for a conventional fan tachometer that produces two pulses per revolution:

RPM = Hz × 60 ÷ 2 = Hz × 30

The FARS produces only one pulse per revolution, so its correct multiplier is 60:

RPM = Hz × 60 ÷ 1 = Hz × 60

Program item	Original sample	Correct FARS treatment
Measurement method	Frequency, command `46`	Frequency; retain command `46` or use command `24` for port 1 only
Pulses per revolution	Effectively assumes 2	1 pulse per revolution
Frequency multiplier	`30`	Read raw Hz with `1.0`, then multiply the FARS channel by `60.0`
Input mode	`2`, no debounce	Retained intentionally for the narrow FARS pulse
Destination name	`Freq()`, although already scaled by 30	Store true Hz in `FARS_Hz` and calculated speed in `FARS_RPM`

QUESTION: What debounce setting should be used for the Standex-Meder MK24 reed contact?

ANSWER: On the original SDM-IO16, begin with mode 2—digital input with no debounce filter.

The MK24 series data sheet specifies a maximum operating time of 0.25 ms including bounce and a maximum release time of 0.15 ms. The 0.25 ms value is not a separately specified bounce-only duration, and it is not the contact-closed pulse width of the assembled FARS.

In the assembled FARS, the available contact-closed pulse duration depends on the complete magnetic geometry:

magnet length and strength;
magnet-to-switch air gap;
reed-switch sensitivity and orientation;
fan radius and rotational speed; and
the time for which the contact remains inside the magnetic actuation field.

Campbell's default switch-closure mode 3 requires the input to remain closed for approximately 3.17 ms and then open for approximately 3.17 ms. That filter is far longer than an assumed 1.00% contact-closed pulse at 2500–3000 rpm.

Campbell input configuration	Sampling interval	Minimum clean state without filtering	Default mode-3 debounce
Original SDM-IO16	0.244 ms at 4096 Hz	0.244 ms	Approximately 3.17 ms
SDM-IO16A, low-speed mode	0.244 ms at 4096 Hz	0.244 ms	Approximately 3.17 ms
SDM-IO16A, high-speed mode	0.06104 ms at 16,384 Hz	0.06104 ms	Approximately 0.7925 ms

For the original SDM-IO16, the smallest nonzero programmable debounce parameter gives a clean-state recognition time of:

0.244 ms + (1 × 0.244 ms) = 0.488 ms

A 0.488 ms filter would reject an exactly 1.00% contact-closed pulse at 2500 rpm, because that pulse lasts only 0.240 ms.

Recommended original SDM-IO16 setting: no internal debounce. This does not mean that the MK24 contact cannot bounce. It means that the original module cannot safely suppress bounce with its internal debounce filter while also guaranteeing acceptance of a pulse that may be shorter than approximately 0.25 ms. If bounce creates false high-speed readings, inspect the actual waveform and use appropriate external pulse conditioning.

QUESTION: Can the original SDM-IO16 reliably detect the short FARS reed-switch pulse at high fan speed?

ANSWER: It depends on the actual duration of the brief contact-closed pulse. The program measures frequency, not duty cycle, but the LOW pulse created when the reed contact closes must still remain long enough for the SDM-IO16 input sampler to detect it.

The original SDM-IO16 samples each input 4096 times per second. The time between samples is:

1 / 4096 second = 0.244 ms

Campbell specifies that an input must remain HIGH or LOW for at least approximately 0.244 ms for a state change or pulse to be counted. Campbell's detailed appendix states the conservative guarantee as longer than the interval between samples.

The FARS reed contact is normally open. It closes briefly as the rotating magnet passes the switch, pulling the SDM-IO16 input LOW. Therefore, the important quantity is the contact-closed LOW-pulse width, not a duty-cycle measurement.

The period of one fan revolution is:

Revolution period in ms = 60,000 / RPM

At an exactly 1.00% contact-closed fraction:

Closed-pulse width = revolution period × 0.01

Fan speed	Pulse frequency	Revolution period	Closed pulse at exactly 1.00%	Original SDM-IO16 requirement	Guaranteed?
2500 rpm	41.667 Hz	24.000 ms	0.240 ms	At least approximately 0.244 ms	No
2600 rpm	43.333 Hz	23.077 ms	0.231 ms	At least approximately 0.244 ms	No
3000 rpm	50.000 Hz	20.000 ms	0.200 ms	At least approximately 0.244 ms	No

At 2500 rpm, one revolution lasts 24.000 ms. One percent of 24.000 ms is 0.240 ms. This is approximately 0.004 ms shorter than the SDM-IO16's nominal 0.244 ms minimum.

At 3000 rpm, one revolution lasts 20.000 ms. One percent of 20.000 ms is 0.200 ms, which is approximately 0.044 ms shorter than the nominal minimum.

Fan speed	Contact-closed fraction required for guaranteed detection
2500 rpm	More than 1.017%
2600 rpm	More than 1.058%
3000 rpm	More than 1.221%

Important distinction A pulse that is outside Campbell's guaranteed timing limit will not necessarily be missed every time. It means reliable detection is no longer guaranteed. A pulse close to the limit may be counted during some revolutions and missed during others, depending on its timing relative to the SDM-IO16's internal sampling instants.

At an exactly 1.00% contact-closed fraction, the theoretical sampling boundary is 40.96 Hz, equivalent to 2457.6 rpm. Therefore, an exactly 1.00% or shorter pulse is outside the original SDM-IO16's guaranteed detection range at 2500 rpm, 2600 rpm, and 3000 rpm.

Measure the actual LOW-pulse width at the SDM-IO16 input with an oscilloscope while the FARS is operating near maximum speed. If the pulse is shorter than 0.244 ms, use an SDM-IO16A in high-speed mode or an external monostable pulse stretcher that produces a clean, longer output pulse.

A Schmitt-trigger input alone is not a pulse stretcher. It can clean up slow or noisy transitions, but it does not necessarily increase a 0.200 ms pulse to a longer duration. A monostable or other defined-width pulse-conditioning circuit is required when pulse extension is needed.

QUESTION: What CRBasic program should be used with a CR1000X and an original SDM-IO16?

ANSWER: The following focused program assumes:

CR1000X data logger;
SDM-IO16 address 1;
FARS output connected to SDM-IO16 port 1;
one pulse per revolution; and
a one-second measurement update.

It configures only port 1 as an unfiltered digital input, leaving ports 2–16 unchanged.

Complete focused CRBasic program

'CR1000X + Campbell Scientific SDM-IO16
'BARANI FARS fan-speed reed output
'
'Wiring:
'  Either FARS lead -> SDM-IO16 port 1
'  Other FARS lead  -> SDM-IO16 G
'
'SDM address: 1
'FARS output: 1 pulse per revolution
'Measurement method: frequency, not duty cycle

Const FARS_PULSES_PER_REV = 1
Const FARS_TEST_CEILING_RPM = 3000

Public FARS_Hz
Public FARS_RPM
Public FARS_Overrange

Public IO16_State(16)

Public Ret_Setup
Public Ret_Frequency
Public Ret_State

BeginProg

  'Configure SDM-IO16 port 1 only.
  '
  'Command 89 configures ports 4 through 1.
  'Mode digits are ordered as port 4, port 3, port 2, port 1.
  '
  '9992 means:
  '  Port 4 = 9, no change
  '  Port 3 = 9, no change
  '  Port 2 = 9, no change
  '  Port 1 = 2, digital input with no debounce
  '
  'Perform setup during two startup scans.

  Scan(1,Sec,0,2)

    SDMIO16(Null,Ret_Setup,1,89,0,0,0,9992,1.0,0)

  NextScan


  'Continuous measurement scan.

  Scan(1,Sec,0,0)

    'Command 24 reads frequency from SDM-IO16 port 1.
    '
    'Multiplier = 1.0:
    'Store the actual raw frequency in Hz.
    '
    'Offset = 0:
    'No value is added to the measurement.

    SDMIO16(FARS_Hz,Ret_Frequency,1,24,0,0,0,0,1.0,0)


    'FARS provides one pulse per revolution.
    '
    'RPM = Hz x 60 / pulses per revolution

    FARS_RPM = FARS_Hz * 60.0 / FARS_PULSES_PER_REV


    '3000 rpm is a program diagnostic ceiling and test margin.
    'The published FARS maximum operating speed is 2600 rpm.
    '
    'Preserve the measured value so bounce spikes remain visible.

    If FARS_RPM > FARS_TEST_CEILING_RPM Then
      FARS_Overrange = 1
    Else
      FARS_Overrange = 0
    EndIf


    'Diagnostic input-state reading.
    '
    'With the fan stopped:
    '  IO16_State(1) = 1 when the reed contact is open.
    '  IO16_State(1) = 0 when a magnet closes the contact.
    '
    'Do not use this once-per-second state value to evaluate
    'the pulse waveform while the fan is rotating.

    SDMIO16(IO16_State(),Ret_State,1,92,0,0,0,0,1.0,0)


    'Add FARS_Hz, FARS_RPM, FARS_Overrange,
    'Ret_Frequency and any desired diagnostic variables
    'to the project's existing DataTable.

  NextScan

EndProg

Why command 24 instead of command 46? Command 24 reads only port 1 and stores one result. Command 46 reads all 16 ports and requires a 16-element destination array. Either is valid; command 24 makes the single-FARS example easier to understand and does not change the other channels.

Compile the program in Campbell Scientific CRBasic Editor for the operating-system version installed in the CR1000X before deployment.

QUESTION: What does each SDMIO16 parameter mean?

ANSWER: Campbell defines the instruction as:

SDMIO16(Dest, IO16Status, SDMAddress, IO16Cmd,
        Mode16_13, Mode12_9, Mode8_5, Mode4_1,
        Multiplier, Offset)

Parameter	Value in this program	Meaning
`Dest`	`FARS_Hz`	Variable receiving the measured port-1 frequency
`IO16Status`	`Ret_Frequency`	Communication result; zero indicates a successful SDM command
`SDMAddress`	`1`	Hardware address selected on the SDM-IO16
`IO16Cmd`	`24`	Read frequency from port 1
`Mode16_13` through `Mode4_1`	`0,0,0,0`	Ignored for frequency command 24; enter zero
`Multiplier`	`1.0`	Preserves the result as true frequency in Hz
`Offset`	`0`	Adds nothing to the measured frequency

Campbell's complete instruction reference is available in the SDMIO16 CRBasic documentation .

QUESTION: What is the smallest change to the original all-port sample?

ANSWER: Read all channels as unscaled frequency in Hz, then apply the appropriate pulses-per-revolution conversion separately to each sensor.

Minimal all-port correction

Public Frequency_Hz(16)
Public FARS_RPM

'Keep the existing no-debounce setup.
SDMIO16(Null,Ret_Code,1,90,2222,2222,2222,2222,1.0,0)

'Read all 16 channels as actual frequency in Hz.
'The multiplier is 1.0, not 30.
SDMIO16(Frequency_Hz(),Ret_Code_1,1,46,0,0,0,0,1.0,0)

'Example: FARS connected to port 1.
'FARS has one pulse per revolution.
FARS_RPM = Frequency_Hz(1) * 60.0

Other fan tachometer channels that produce two pulses per revolution can still be converted with a multiplier of 30:

Conventional_Fan_RPM = Frequency_Hz(2) * 30.0

Do not multiply every channel by 60 unless every connected device produces one pulse per revolution. Reading raw Hz first allows each channel to use its own documented pulses-per-revolution value.

QUESTION: What should be changed when an SDM-IO16A is installed?

ANSWER: The SDM-IO16A supports command 104, which enables a 16,384 Hz high-speed sampling mode. Execute this command during every scan, before reading the frequency.

SDM-IO16A high-speed addition

'Declare with the other public variables:
Public Ret_HighSpeed

'Only for an SDM-IO16A.
'Place inside every measurement scan before command 24 or 46.

SDMIO16(Null,Ret_HighSpeed,1,104,0,0,0,0,1.0,0)

In high-speed mode, an unfiltered state needs to last approximately 0.06104 ms rather than 0.244 ms. A clean 0.200 ms pulse at 3000 rpm is therefore longer than the high-speed sampling requirement.

Command 104 is for the SDM-IO16A only. Do not add it to a program controlling the original SDM-IO16. Also note that high-speed sampling improves pulse detection but does not remove mechanical contact bounce. Retain mode 2 initially and inspect unexpected high-speed spikes before adding any filter.

See the Campbell Scientific SDM-IO16A manual for the high-speed command and timing specifications.

QUESTION: How can a reed-switch, wiring, pulse-width, or scaling problem be distinguished?

ANSWER: Use the following symptoms as a diagnostic guide:

Observed result	Most likely explanation	Recommended check
Displayed speed is exactly one-half of the expected value	The program uses multiplier 30, which assumes two pulses per revolution	Read raw Hz and multiply the FARS channel by 60
Correct reading at moderate speed, but zero or intermittent near maximum speed	The closed pulse may be shorter than the original SDM-IO16's 0.244 ms sampling interval	Measure pulse width at the IO16 terminal with an oscilloscope
RPM occasionally jumps above the physical fan-speed limit	Contact bounce, cable interference, or multiple transitions are being counted	Inspect the waveform; use an appropriate pulse-conditioning circuit if necessary
`IO16_State(1)` remains 1 with a stationary magnet directly over the switch	Contact is not closing, magnet position or orientation is incorrect, or wiring is open	Test continuity directly at the FARS cable and then at the IO16 terminal
`IO16_State(1)` remains 0 with the magnet removed	Short circuit, stuck contact, or continuous magnetic actuation	Disconnect the FARS cable and compare the port state
`Ret_Frequency` or another return code is nonzero	SDM communication, address, power, or cable problem	Verify the hardware address, 12 V supply, ground, and SDM-C1/C2/C3 wiring
Frequency remains zero only at very low speed	The frequency command has received fewer than two complete high-to-low transitions during the measurement interval	Increase the interval if low-speed measurement is required

With a one-second frequency measurement interval and one pulse per revolution, the SDM-IO16 reports 0 Hz when fewer than two high-to-low transitions occur between reads. This can affect readings below about 120 rpm while the fan is starting or stopping, but it is not relevant near the normal operating speed.

QUESTION: What is the final recommended configuration?

ANSWER:

Measure frequency, not duty cycle.
Use 60 rpm per hertz because the FARS has one pulse per revolution.
Use SDM-IO16 input mode 2 with no internal debounce as the initial configuration.
Do not use the default 3.17 ms switch-closure filter for an expected narrow perimeter pulse.
Confirm that the stationary input changes from 1 to 0 when the magnet operates the reed contact.
Measure the dynamic contact-closed pulse width near 2500–2600 rpm if the original SDM-IO16 misses pulses.
Use an SDM-IO16A in high-speed mode or a defined-width external pulse conditioner when the pulse is shorter than 0.244 ms.
Treat RPM spikes above the physical limit as possible contact bounce or electrical noise rather than valid fan speed.

Do not diagnose the MK24 reed contact as faulty from a missing high-speed reading alone. A contact that works correctly during a slow continuity test can still produce a pulse that falls below the original SDM-IO16's input timing limit. The decisive test is the waveform measured at the data-logger input while the fan is operating.

FAQ

FAQ Search

How do I connect a BARANI FARS reed-switch fan-speed output to a Campbell Scientific CR1000X data logger?